New Directions in International Advertising Research (Advances in International Marketing)

LIST OF CONTRIBUTORS Hae-Kyong Bang Villanova University, USA P. Greg Bonner Villanova University, USA Les Carlson...

Author: Charles R. Taylor

113 downloads 1873 Views 13MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

LIST OF CONTRIBUTORS

Hae-Kyong Bang

Villanova University, USA

P. Greg Bonner


Les Carlson

Clemson University, USA

Albert Caruana

University of Malta, Msida, Malta

Aruna Chandra

Indiana State University, USA

Patrick De Pelsmacker

Universiteit Antwerpen Management School and Ghent University, Belgium

Kanwalrooop Kathy Dhanda

University of Portland, USA

Michael Dolezal

Masaryk University, Czech Republic

Michael T. Ewing

Monash University, Australia

Maggie Geuens

Vlerick Leuven Gent Management School and Ghent University, Belgium

David A. Griffith

University of Hawaii, USA

Ronald Paul Hill

University of Portland, USA

Cameron James Hughes

Millward Brown International, United Kingdom

Chad M. Johnson


Dimitri Kapelianis

Arizona State University, USA ix

X

LIST OF CONTRIBUTORS

Stephen Keysuk Kim

Oregon State University, Corvallis, OR, USA

Wei-Na Lee

The University of Texas at Austin, USA

Jong Won Lim

Seoul National University, South Korea

Dominika Maison

University of Warsaw, Poland

Gordon E. Miracle

Michigan State University, USA

Changho Oh

Hanshin University, Osan, South Korea

Sak Onkvisit

San Jose University, USA

Michael Jay Polonsky

Victoria University, Australia

Andrea Prothero

University College Dublin, Ireland

Mary Anne Raymond

Clemson University, USA

John K. Ryans, Jr.

Kent State University, USA

Trina Sego

Bosie State University, USA

John J. Shaw

Providence College, USA

Charles R. Taylor


Andy Teo

PSB Corporation, Singapore

Fei Xue

University of Alabama, USA

TaiWoong Yun

The University of Texas at Austin, USA

Shuhua Zhou


Peiqin Zhou


PREFACE

This special volume of Advances in International Marketing is devoted to exploring new perspectives on international advertising - a generally underresearched area. It is guest edited by Charles R. Taylor of Villanova University. The idea for devoting a separate volume on international advertising came from Professor Taylor. He issued a call for papers, which then attracted a variety of submissions of high quality. We owe gratitude to him for screening and evaluating these submissions, and for preparing the final set of chapters. We are also indebted to many colleagues who assisted in the review process. The resulting selections draw from a variety of perspectives and offer rich insights on international advertising issues. Our thanks to Dr. Taylor for his efforts in creating this volume. Finally, we express our appreciation to Ms. Alison Evans and other staff at JAI/Elsevier Science who saw the volume through the production process. S. Tamer Cavusgil Series Editor

xi

I N T R O D U C T I O N - NEW DIRECTIONS IN INTERNATIONAL ADVERTISING RESEARCH Charles R. Taylor International advertising is an area in need of more quality research. While this need is somewhat surprising given the large amount of attention focused on advertising in many parts of the world, one needs only to look at chapters on international advertising in International Marketing or Advertising textbooks to see the problem. In comparison to well-researched areas such as foreign market entry modes, country of origin perceptions of products, pricing strategies, and global branding strategies, there is far too little that is managerially actionable in international advertising discussions. Too often, we are left to give our advertising students vague advice, such as "you need to search for multi~market target audience similarities," in spite of not knowing a lot about how it can be done due to a paucity of research. Another example of a failure to deliver value to managers, (and our students), stems from being preoccupied with the "debate" over standardization vs. specialization. Too often, the discussion is positioned as a controversy rather than providing managers with specific advice on how to deal with the decision. We also know too little about setting advertising budgets and measuring advertising effectiveness in foreign markets. Research focusing on allowing us a better understanding of the impact of culture on advertising effectiveness and on client-agency relationships is also urgently needed. While the work of early international advertising scholars such as Gordon Miracle and John Ryans, who are among the contributors to this volume, did an excellent job of setting the

New Directions in International Advertising Research, Volume 12, pages 1-5. Copyright © 2002 by Elsevier Science Ltd. All rights of reproduction in any form reserved. ISBN: 0-7623-0950-4

2

CHARLES R. TAYLOR

stage for additional research, too often we have fallen short. Some recent key studies suggest promising things are ahead, but more is needed. There are some understandable reasons why international advertising research has not advanced as far as it might. As noted by Zinkhan (1994), it is simply more difficult to collect data from multiple countries. Taking on multiple co-authors is not viewed as desirable at many universities though it may often be necessary in a large international study. Another issue related to the difficulty of collecting data in multiple markets is that a disproportionately high number of content analyses have been published, perhaps to the neglect of other research techniques such as surveys and experiments. It will be especially important for researchers of the future to build on descriptive research with experiments that focus more on the underlying reasons why various executional techniques are effective in a market as opposed to simply describing the existing content of ads. Another issue facing international advertising researchers is that elite journals frequently frown on studies from countries other than the United States that do not collect comparable data from the U.S. Even when the U.S. is included, the second country in a study had better be of interest to the "general readership" of the journal if the author hopes to publish it. An additional problem for researchers is that studying multiple countries adds noise to data sets, yet international studies are held to the same reliability levels as domestic studies. Collectively, all of these problems associated with conducting research in international advertising have led to a situation where we have only a very limited number of scholars pursuing programmatic research, and even fewer who have been able to publish their work in the top journals in the field. While the above problems do, indeed, make it difficult to conduct international advertising research they are not insurmountable, as illustrated by the articles in this volume. Several different methodologies, including experiments, surveys, qualitative research, and content analysis are used in the papers offered here. Moreover, the studies are conducted in several countries from various parts of the world and several delve into neglected areas of inquiry. The opening section provides some guidance in terms of implementing procedures to ensure that survey data compared in multiple countries is equivalent. The two articles in the Methodological Advances section build on prior work on ensuring that equivalent data are being compared in cross-cultural studies. In the first article, Miracle and Bang discuss procedures for making sure equivalent measurement instruments are produced prior to data collection, including guidelines for effective translations and procedures for training coders when content analysis is used. In the second article, Ewing, Caruana, and Teo build on Steenkamp and Baumgartners's (1998) analytical procedures for

Introduction

3

assessing whether data are equivalent by developing the eric Multicultural Advertising Response Scale (MARS). Collectively, these studies demonstrate how it is now possible for researchers to eliminate much of the noise in crossnational research. Employing the types of procedures illustrated in these studies, both before and after the data is collected, will enhance the chances of an international advertising study being accepted in high-level journals. The second section of the volume is titled "Standardization vs. Specialization Issues." These studies examine the advances that have been made in recent advertising research on standardization and reflect the need to see the issue from a managerial perspective, as opposed to becoming preoccupied with the "debate." In a review of the literature on standardization vs. specialization in the 1990s, Taylor and Johnson stress the growing consensus that brand positioning or the main selling message can often be standardized while individual country executions must be carefully analyzed when making a decision whether to standardize. Thus, at a general level, the decision to standardize should become a matter of "what"and "how," as opposed to whether. The authors also call for more research on the client-agency relationship vis-a-vis standardization and on inter-market segmentation. The second article in this section, by Chandra, Griffith, and Ryans, investigates the relationship of process standardization to program standardization using results from a survey of U.S. multinationals operating in India. This study provides an excellent country-specific example of the level of analysis recommended by Taylor and Johnson for determining whether specific programs (or executions) can be standardized. The third contribution to this section is a fascinating piece by Onkvisit and Shaw in which the universality of theories and concepts contained in U.S. textbooks is examined and questioned. As textbook authors, their call for more research evaluating and replicating advertising concepts in other cultural contexts should be taken especially seriously, as they have first hand the need to summarize the state of knowledge on international advertising. In the final article in this section, Raymond and Lim provide a detailed case study that, again, illustrates the level of depth of analysis needed in making the decision of what can be standardized. Their case study of Hyundai's Santa Fe SUV shows how country of origin effects and other factors need to be considered in deciding whether adaptation is necessary. The third section of the volume is entitled "Advertising in Central/Eastern Europe." The two articles here focus on two transitional economies, Poland and the Czech Republic, that have seen substantial growth in advertising over the past decade. De Pelsmacker, Maison, and Geuens conduct an experiment which examines how emotional vs. rational advertising appeals work in positive versus negative media contexts. Experimental studies of this type are too rare

4

CHARLES R. TAYLOR

in advertising research in general, but especially in transitional economies. The second article, by Taylor, Bonner, and Dolezal provides evidence on two issues. First, it reports on a survey of Czech perception of advertising clutter in various media. Second, it analyzes Czech consumers' perceptions of what types of advertising executions are effective, using data from both a survey and personal interviews. The fourth section is rifled, "Global Advertising, Social Responsibility, and Country of Origin Effects." Polonsky, Carlson, Prothero and Kapelianis start the section off with a study of environmental information contained on packaging in four countries - Australia, South Africa, the United Kingdom, and the U.S. Using a new methodology in which judges in each country determine the categories for coding as well as performing the actual coding, the study finds that some types of information not traditionally classified as environmental were identified. The authors also provide implications for advertisers. The next study, authored by Hill and Dhanda, analyzes secondary data in order to assess the impact of differential access to the internet on the ability to promote goods and services worldwide. Macro-level implications for advertisers are provided. The final paper in this section, by Yun, Lee, and Sego, conducts an experiment in which the impact of hybrid vs. non-hybrid country of origin effects on response to an ad are analyzed. The study, conducted over the Intemet, also examines the effect of indirect foreign language cues in the form of foreign language slogans. Advertising in Asia Pacific is the title of the final section of the volume. Oh and Kim lead off the section with a study of how power impacts on clientagency relationships. Drawing on theory that is often applied in a channels context but seldom in an ad agency context, they find that differential power between clients and agencies can actually play a constructive role. The second paper, by Zhou, Xue, and Zhou, examines advertising and its relationship with materialism, self-esteem and life-satisfaction. Their results, from a student sample, clearly demonstrate that Chinese students are optimistic in terms of their future and that the relationship between the tested variables is complex. In the final paper, Hughes and Polonsky compare the information content in the advertising of Australian, Japanese, and U.S. MNEs in Australia. Their findings suggest that Japanese and U.S. firms use different levels of information in Australia than they do in their home markets. Collectively, these studies provide a basis on which to build new directions in advertising research, or to continue to build on existing trends in areas in need of research. I would like to thank S. Tamer Cavusgil, the Series Editor, for his encouragement and for the opportunity to edit this issue, and Kathy Waldie of Michigan State University for her support throughout the process.

Introduction

5

Special thanks are also due to Alison Evans of Elsevier who served as Administrative Editor of the Volume and Kyong Ryul (Kevin) Koo, my graduate assistant at Villanova University who worked hard on this project. Finally, I very much thank the Editorial Board members who reviewed the papers for this issue, thereby ensuring its quality.

REFERENCES Steenkamp, J. E. M., & Baumgartner, H. (1998). Assessing measurement equivalence in crossnational consumer research. Journal of Consumer Research, 25(3), 78-90. Zinkhan, G. M. (1994). International advertising: A research agenda. Journal of Advertising, 23(1), 11-15.

ACHIEVING RELIABLE AND VALID CROSS-CULTURAL RESEARCH RESULTS IN CONTENT ANALYSIS Gordon E. Miracle and Hae-Kyong Bang

ABSTRACT

Cross-cultural research presents unique challenges due to language and cultural differences. Based on a large-scale research project involving television advertisements from Japan, Korea, and the U.S., the article identifies a number of problems encountered in developing equivalent research instruments in, and discusses ways to achieve reliable and valid results in cross-cultural content analysis research. The importance of understanding the languages and culture, selecting, training and supervising coders in a consistent manner and the practical value of back-translation process in cross-cultural research in achieving reliable and valid research results are emphasized. Specific guidelines for ensuring equivalent research instruments are provided. INTRODUCTION

Many products and services are produced in one country and sold in distant and culturally different markets. Often international marketers depend mainly on judgment to adjust advertising for differences in foreign markets. Some even

New Directions in International Advertising Research, Volume 12, pages 9-23. Copyright © 2002 by Elsevier Science Ltd. All rights of reproduction in any form reserved. ISBN: 0-7623.0950-4

10

GORDON E. MIRACLE AND HAE-KYONG BANG

attempt to find globally acceptable advertising that can be used in all markets. Cross-cultural research methods guide such decisions. Such research requires specialized techniques to overcome cross-national problems of functional, conceptual and linguistic equivalence (Miracle, 1988). In fact, ensuring equivalence is essential to producing useful research (Douglas & Craig, 2000; Kumar, 2000). However, results of a meta-analysis of major studies of the information content of advertising suggest that, all too often, appropriate measures are not taken to ensure that the results of research conducted in two countries are equivalent (Abernethy & Franke, 1996). Notably, the Abemethy and Franke study suggested that studies that followed training procedures for coders produced different, and one can surmise, more accurate, results than those that did not. The article describes selected research design ideas, techniques, and lessons that have been found to be useful in doing cross-cultural research in Japan, Korea and the USA. These lessons may also be useful to those who would do research to guide advertising in other cultures. Many o f the ideas are not expensive to implement, and can be helpful to both academics and practitioners conducting research in international markets. The study that provided the experience for this methodology article was a large content analysis of television commercials in three countries: Japan, Korea and the USA. A number of articles and papers have been written, and some are already in print or in press (Miracle et al., 1992a, b; Taylor et al., 1992; Miracle et al., 1992). Subsequent data collection and analysis also resulted in articles (Taylor et al., 1997; Taylor et al., 1994). Cross-cultural research encompasses not only the entire range of problems encountered when doing research within one culture, but also problems that arise from differences in cultures. Although this article is based on lessons learned from content analysis, some of these methods may also be useful for survey research. Indeed some of the interpersonal communication problems encountered and resolved when preparing data coding forms, when training and using translators, and when training supervisors and coders, may be similar to those encountered when preparing interviewers, and training them to handle telephone or personal interviews.

The Concept of Culture The concept of culture is elusive. Although culture has been defined in many ways, it is not appropriate here to deal comprehensively with the many definitions and components of culture. For our purposes it seems suitable to use Hofstede's (1984) definition: " . . . culture is the collective programming of the

Achieving Reliable and Valid Cross-Cultural Research Results

11

mind which distinguishes the members of one group or society from those of a n o t h e r . . , culture is reflected in the meaning that people attach to the various aspects of their life; their way of looking at the world and their role in it; in their values, that is, in what they consider as 'good' and as 'evil'; in their collective beliefs, what they consider as 'true' and as 'false'; in their artistic expressions, what they consider as 'beautiful' and as 'ugly'." Advertising reflects culture. Differences in culture between societies suggest measurement problems. The central issue addressed in this paper is how to define and measure culturally influenced characteristics of Japanese, Korean and U.S. television advertising systematically, objectively and quantitatively so as to achieve reliable and valid research results. As alluded to earlier, many of these lessons are applicable to advertising research in other cultures.

The Emic-Etic Distinction An emic generalization is one that is valid within a single culture. An etic generalization is the one that is valid in two or more cultures. Cross-cultural research is concerned with identifying and distinguishing emic from etic generalizations. Etic generalizations depend on the equivalence of culture between two or more societies. Equivalence There are differences and similarities between cultures and the ways in which language is used to express ideas and information. These differences and similarities can be grouped into functional, conceptual or linguistic characteristics of a culture and language. These similarities or differences can be important influences on setting advertising objectives, deciding advertising strategy, or doing advertising creative executions. Functional differences and similarities of cultures relate to behavioral phenomena. From the standpoint of advertising, an important question is whether or not habits and customs with respect to consumer purchasing behavior or product usage are the same or different in the culture under study. For example, it may or may not be appropriate to discuss business at meals, to eat a particular food for breakfast, or to use a bicycle for the same purpose (e.g. for pleasure or for basic transportation). Conceptual differences and similarities of culture relate to the existence or non-existence of certain ideas or ways of thinking in a culture. Of importance to advertising is whether certain concepts of consumer purchasing behavior or product usage are the same or different in the cultures under study. For example, the meaning of certain colors may or may not have the same

12


connotations; white may be the color of purity in one country or the color of death in another. The concepts of "ethnic" and "foreign" may be different in the USA, but virtually identical in Korea. What is humorous in one nation may not be at all humorous in another.

Linguistic differences and similarities relate to the translatability of function and concepts from one language into another. For example, Eskimos may have many more words to express different kinds of snow than are available in English. Sometimes idioms or metaphors are meaningful in one language but not in another. An example is the U.S. expression: it is as easy as rolling off of a log. This expression, if translated, may be entirely meaningless in many languages. Such words as "marketing," "computer" and "television" have been incorporated into many languages in which there is no word that is exactly equivalent. When such a word is adopted into a language, there may remain a difference in the meaning for an extended period of time. Additionally, some words are translatable but require lengthy explanation to express the full idea.

Determining Reliability In its simplest form, reliability is the "degree of correspondence between two sets of frequencies of classified symbol data when the analysis of the same communication is done by two independent analysts or groups of analysts." (Janis et al., 1943, p. 293) Three types of reliability have been identified in the literature, First, stability: the degree to which a process is invariant or unchanging over time when measured at different times by one observer. It is also known as consistency or intraobserver reliability. This is also the weakest form of reliability. Secondly, reproducibility: the degree to which a process can be recreated under varying circumstances at different locations by different coders. It is also known as intercoder reliability. Finally, accuracy: the degree to which a process functionally conforms to a known standard, or yields what it is designed to yield. This is the strongest reliability test available, surpassed only by a measure of validity that assumes the known standard must be true (Krippendorf, 1980). In this project, we were concerned with reproducibility and accuracy. In creating equivalent coding or data collection instruments across two or more cultures, the translation process may pose a problem either by decreasing reliability or by increasing reliability. Unlike the common belief that the reliability level may only diminish in cross-culture studies, several factors may actually create seeming equivalence: (1) translators may have a shared set of rules for translating certain non-equivalent words and phrases; (2) some back-translators may be able to make sense out of a poorly written target


13

language version; or (3) the bilingual person translating from the source to the target may retain many of the grammatical forms of the source (Brislin, 1970). Should this take place, the result may be higher reliability but lower validity. In order to secure consistent measures, translations and back-translations should be done by those who are knowledgeable not only about the languages, but also about the subject of research, in this case, advertising. While a full discussion of the best measures for reporting reliability is beyond the scope of this paper, it should be noted that there is now a widespread consensus among marketing and advertising researchers that reported reliability figures should be corrected for chance agreement (Taylor & Stem, 1994). Perrault and Leigh (1989), Hughes and Garrett (1990), Franke (1992), and Rust and Cooil (1994) all provide excellent discussions of this issue. Determining Validity As Holsti (1969) points out, the meaning of validity can vary from study to study, depending on the investigator's purposes. According to the American Psychological Association Committee on psychological tests, four types of validity can be distinguished such as content validity, predictive validity, concurrent validity, and construct validity. Content validity, also sometimes referred to as face validity, has most frequently been relied upon by content analysis. If the purpose of the research is a purely descriptive one, content validity is normally sufficient, and this type of validity is usually established through the informed judgment of the investigator. Since the research on which this article is based was descriptive, it was deemed sufficient to rely on content validity. The researcher's judgments on validity depend on the research methods used and the care with which they are employed. The main thrust of this article is to illustrate such methods.

RESEARCH METHODS The Translation~Back-Translation Process

The data coding instruments for this study were prepared initially in English. They were based largely on the review of the U.S. literature (see references in Miracle, 1988), but also on several cross-national and non-U.S, studies. Additional adaptations were based on the language and advertising experiences of the researchers in Japan and Korea. Many uncertainties were encountered and resolved as to which items on the data coding instrument were emic or eric. In view of the importance of problems of equivalence, it was important to devise a method to: (1) identify differences when they exist; and (2) resolve

14


them when possible, especially to discover if they are an artifact o f a particular translator rather than a genuine difference. One technique to deal with problems of equivalence is to: (1) use one or more persons to translate material from the original language into a second language; (2) translate it back into the original language; and (3) compare the original with the back-translated version. This study involved three languages: Japanese, Korean and U.S. English. Thus it was more complex than if only two languages had been involved. The scheme is shown in Fig. 1. In each language, it is important for a competent translator to translate into his or her own native language. It is not only easier and quicker to translate into o n e ' s own native language, but more accurate as well. A translator should have a current knowledge of his or her own native language from recent experience, since languages evolve continuously and the translator must be up to date. The cultural, economic and social backgrounds o f translators are also important regarding their perspectives on functions, concepts and language. Additionally it was important for this particular project that the translators should know the technical language o f advertising. Miracle (1988) reported more fully on the requirements for effective translations o f advertising and related material. W h e n each back-translation was compared with the original in that language, many differences were discovered. A few apparently genuine emic concepts were identified, as well as a number of differences that arose from what appeared

E ..............

E ..............

~

K ...............

~

K ...............

" ~ J, and compare with J from below

"~ J J

Notes: 1. E = English,

.

.

.

.

.

.

.

.

.

.

.

...............

.

.

.

.

E, compare E

"liP" E, compare E " ~ K, and compare with K from above

K = Korean, J Japanese 2. The direction of the arrows shows the translation from one language into the other. 3. In each case the end product back-translated must be compared with the original in that language - for purposes of identifying and resolving problems of functional, conceptual and linguistic equivalence. 4. At each stage, one or more independent translators must be utilized.

Fig. 1.

=

Translation and Back-Translation Performed on the Data Coding Instruments and Codebook.


15

at first to be linguistic non-equivalence. These were subsequently resolved by improving the translations.

Special Problems of Translations Between Japanese and English Written Japanese is a combination of three symbol systems: (1) Chinese characters or symbols known as kanji; (2) indigenous Japanese symbols known as hiragana; and (3) indigenous Japanese symbols known as katakana. The latter are used primarily for transliterating foreign words so they can be pronounced in Japanese. All three symbols may be used in the same sentence. If a kanji or hiragana symbol is not available to express a foreign concept, the Japanese can write a word in katakana so that they can pronounce it. For example, the word "television" becomes a three syllable word "te-re-bi," "beer" becomes a two syllable word, "bi-ru," and the word "Miracle" becomes a four syllable word, "Mi-ra-ku-ru." The interesting feature of this practice is that someone reading the transliterated katakana can render it back into the original language, without understanding the meaning of the word. In our study, such a problem arose when a translator used katakana to translate a few "difficult" words from English into Japanese. Some of the back-translations from this Japanese into English showed remarkable agreement with the original English, suggesting the false conclusion that there were no problems of equivalence. The problem was compounded by the fact that even though such transliterated words have become a part of the Japanese language, these words often take on new connotations, or lose part of the original meaning. For example, the katakana rendering of skeleton (se-ke-ru-ton) carries the meaning of "outline" (as used when referring to an outline of an article). With all languages, virtually all translations suffer at least a little from a partial rather than full correspondence in the meaning of words. For example, the Japanese expression for "computer graphics" carries the meaning of a picture a drawing, an illustration or sketch, but not a graph. Another example is the Japanese word for "animation" which in translation carries the meaning of "comics" or "cartoons". Although the items on the data coding instrument originated largely from U.S. literature and experience, Japanese and Korean literature searches provided a few information and executional variables that had not arisen in English. However, the review of the Japanese literature was necessarily not as comprehensive as the review of the U.S. literature, and some Japanese emic concepts were probably missed. -

16


Special Problems of Translations Between Korean and English An example of a problem with the translation of "English to Korean and back to English" occurred with the use of the words "ethnic and "foreign". Although they have two separate meanings in English, there is only one word for both of them in Korean, and ethnic is considered to be identical to foreign. Koreans believe they are one ethnic group, so whoever comes from a different ethnic group is considered by them to be foreign.

Training and Using Translators In this study there were four main types of items to be translated: (1) data coding instruments; (2) code books of operational definitions of the items on the data coding instruments; (3) instructions for coders; and (4) sampling specifications and procedural instructions. The initial translations from English into Japanese and from English into Korean were done in the USA. These native Japanese and Korean translators were a part of the research team and therefore met regularly with the entire research team during the planning and discussion of all phases of the project. They participated first in the pilot study, learning from this early testing. They participated in the early discussion of probable emic and eric concepts, and the discussion of probable difficulties with equivalence. They helped to develop the research design. Therefore they were prepared to translate the real meaning of the concepts on the data coding instruments and the definitions in the code books. After these translations were followed by back-translations, the research group discussed in detail the problems of equivalence that were found. The data coding instruments and code books in English, Japanese and Korean were then revised. At this point, a pilot study was done to test the data coding instruments and code books. This experience led to further improvements in the data coding instruments, and in the code books, as well as in the plans for training supervisors and for recruiting and training data coders. These improved data coding instruments and code books, after the pilot test, were then put though the back-translation processes from the English to Japanese to English, and from the English to Korean to English. Translators were used who had had no prior exposure to this research. Both comparisons of the back-translated English with the original English indicated that equivalence had been achieved. The next step in preparing the data coding instruments and code books was to: (1) translate the revised Japanese version into Korean and then back into Japanese; and (2) translate the revised Korean version into Japanese and back into Korean. These two tasks were accomplished in Japan and Korea respectively. These back-translations were then compared with the original


17

versions by members of the research team. These comparisons provided additional input into the final versions of the research instruments and especially to guide the training of the native Japanese and Korean coders who did the coding in their respective countries. Sampling In order to obtain random samples that are comparable and representative, carefully designed specifications were followed in each country. In selecting a sample of television commercials, it was important at the outset to learn about differences in government and media policies with regard to program times, number and timing of airing of commercials, and so forth. The sampling method led to samples in each country that were comprised of a comprehensive range of product categories normally advertised on national television. These samples of national brand commercials from each country contained a similar proportion of products advertised. Therefore, differences in research results could not be attributed to differences in the types of products advertised. Recruiting Coders For the pilot study, six each of Japanese, Korean and U.S. coders were recruited from students at a U.S. university. From this pilot study it became clear that it was important to find coders with an interest in and knowledge of advertising. It was not possible to train those with little knowledge of advertising to code as accurately as those who had substantial knowledge of advertising. In the pilot study it also became clear that native Japanese and Koreans who had lived in the USA for several months had become knowledgeable about the U.S. culture. They apparently internalized some of this knowledge in ways that caused them to have difficulty in coding commercials in their native language in the same way that "uncontaminated" natives of each country would do the job. They had a tendency to judge commercials in their native languages according to characteristics or standards of U.S. commercials. Moreover, after even a few months, there were brands, products and commercials in their native countries that were unfamiliar to them - those that had been introduced since they had left. Some of these commercials had new expressions or ideas, and coders said they had difficulty in understanding or interpreting them accurately. Therefore it became clear that the best alternative was to recruit uncontaminated coders living in their native countries, i.e. in Japan, Korea and the USA. Training Coder Supervisors and Coders It was recognized at the outset that the reliability and validity of the research findings would depend directly on comparable coding in all three countries.

18


Therefore coder training, supervising and control necessarily had to ensure that the coding would be done in the same way and according to the same rules and standards in each country, while still consistent with the culture of all three nations. Therefore every effort was made to train, supervise and control the coders so that their output would be comparable. Many of the procedures followed were consistent with the recommendations of Kolbe and Burnett (1991), an article that provides excellent guidelines for training coders. Sixteen U.S. coders were trained by the principal investigator. Three other members of the research team, one American, one Japanese and one Korean, were also present. The American was the one who would later supervise and control the coders and the coding process for U.S. commercials. The Japanese and Korean researchers were the ones that would go to Japan and Korea to train the Japanese and Korean coders respectively. After each training session, the principal investigator discussed the training with the Japanese, Korean and U.S. research assistants who would later train and supervise coders in their respective countries. The discussion continued until it was clear that the Japanese and Korean trainer/supervisors felt they would be able to train coders in their countries equivalently to the way U.S. coders were trained.

Reliability Testing Procedures The next step was to conduct a reliability test. All U.S. coders working in one room coded 30 commercials, with no discussion among them, except that any coder could ask to see or hear part or all of the commercial as often as needed. This reliability testing process required nine sessions totaling a little more than 13 hours. This same process was later followed by the Japanese and Korean coders in their respective countries.

Coder Supervision and Control After the training and reliability testing was completed in the USA, written instructions, forms for weekly productivity reports, and a schedule of weekly meetings of coders with their supervisor were prepared. The U.S. supervisor collected the completed data coding forms each week, answered coder questions, resolved difficulties, informed other coders of relevant details, and prepared a log of all activities. This log was then used as the basis for a sununary report at the end of the coding to evaluate the entire process. This experience was then shared with the Japanese and Korean trainers/supervisors so that they would supervise and control their coders using the same process in a comparable manner. After the U.S. coding was completed, the native speaker Japanese and Korean trainer/supervisors traveled to Tokyo and Seoul respectively. There they met with


19

local Japanese and Korean research colleagues respectively and together they trained the coders. Reliability Calculation Methods

A review of previous communication studies revealed that a reliability coefficient is often expressed as correlations among coders' classifications (Hughes & Garrett, 1990). In many cases, a correlation among the coders' responses is interchangeably used to mean an agreement among the coder responses. However, it is important to note that it is the degree of agreement, not of correlation, that should be measured in order to achieve both reliable and valid results. An important distinction was made by Robinson (1957) regarding the validity of this kind of interchangeability. He stated that agreement requires that paired values must be linked by a linear relationship, or if one defines correlation more broadly, that the paired values be linked according to some mathematical function. Thus, agreement is a special case of correlation, or a narrower form of correlation, in that two variables that agree must be correlated, but variables which are correlated need not necessarily agree. Determination of an acceptable reliability level depends on the complexity of the research task (Hornik, 1988). Those who best know the whole process of research should decide which level is acceptable because the problems of reliability and validity involve every phase of the content analysis process. Therefore, the analyst must determine what level of agreement is satisfactory to him (Budd & Thorp, 1963). The current consensus among many researchers is that 0.85 is an acceptable level of reliability (Kassarjian, 1977; Krippendorff, 1980). In our study, it had been determined that the 85 percentage level of agreement is sufficient. In our study, the intercoder percentages of agreement exceeded 90% within each country. Additional tests to correct for chance agreement were undertaken, and, again, the categories exceeded the Kassarjian standard. EVALUATION

OF RELIABILITY

RESULTS

Some possible explanation for the high mean percentages of agreement achieved in this study are: (1) adequate coder training, and (2) accurate and precise definitions of variables. Since the analysis was done by groups of 10 Japanese, 12 Korean and 16 U.S. coders (not just by two or three coders as in many previous content analysis studies), the high percentages of agreement can be considered as especially noteworthy. The achievement of equivalence through translation and back-translation in this study was designed to identify both emic and eric characteristics of the

20


three countries' advertising. The research was designed to measure the same thing across the three countries. Therefore, the satisfactory level of percentage of agreement demonstrated in our study supports the conclusion that our research methods and measuring instruments yielded valid results. CONCLUSIONS Cross-cultural research encompasses not only the entire range of problems encountered when doing research within one country, but also problems that arise from differences across cultures. This study reports methodology lessons from cross-cultural research in three countries. It is hoped that these lessons are applicable also to cross-cultural research conducted in other countries. The major methodological lessons from this study are as follows: (1) The translation/back-translation process is a practical, useful, indeed necessary technique to identify emic and etic distinctions to overcome problems of equivalence and to achieve reliable and valid results. This process is useful not only for data coding instruments and code books, but also for coder training materials and other documents that must be strictly equivalent in two or more languages. Any cross-cultural or cross-national study that does not employ the back-translation technique should be an immediate candidate to be judged unacceptably flawed. (2) Translators should translate into their native languages. Someone with many years of residence in a country, with substantial professional experience, can probably substitute reasonably well for a native speaker; but in this case it is important to use more than one translator on the same material to serve as a check. Less stringent requirements would make the results of a cross-cultural research study suspect. (3) Special problems can arise when languages employ different symbol systems and types of alphabets. Transliteration is a special problem with Japanese katakana that can confound the translation/back-translation process, and lead to serious problems of equivalence in research instruments. Researchers should be alert to the possibility of similar types of problems with other languages. (4) Literature reviews of appropriate theoretical and methodological topics should be done not only in the originating nation, but also in each nation in which the cross-cultural research is done. If not done properly, the research instruments will likely be biased toward the culture of the nation in which the literature was reviewed. The consequence may be coder


(5)

(6)

(7)

(8)

(9)

(10)

21

misunderstandings, lack of comparable coding, large numbers of "other" answer categories, and unreliable and invalid results. There are likely to be special problems of equivalence associated with doing cross-cultural research in just about every country. These problems probably are not unique to any country, and they are also not the same in all countries. Therefore they need to be identified separately in every crosscultural research project. Coders who are "uncontaminated" by foreign experience are necessary to achieve accurate work. After only a few months in another country coders begin unknowingly to adopt some of the standards of the host country, and become incapable of doing as accurate work in their own language as can be done by uncontaminated coders. In a three-country study, translations and back-translations should be done between all three languages. It is not adequate to design research instruments in one language and only do the back-translation process between that language and the other two languages. Additional problems of equivalence, and solutions to resolve them surface when doing the process between the second and third languages. In each country it is important to recruit coders who are as similar as possible to the coders in the other countries. When research instruments are designed that are appropriate for coders with certain characteristics, problems of equivalence can arise if these research instruments are used by others, i.e. by those with different socioeconomic backgrounds, different educations, and so forth. They are therefore different in the way they are likely to be able to use the research instruments. In this study we recruited coders of similar age, educational level and knowledge of advertising in their countries. Although small differences in coders were detected, they were deemed acceptable. However, if we had used coders that were substantially different, our experience suggests that the research results may have been much less reliable and valid. Always do a pilot study to pretest the sampling plan, research instruments, coder training/supervision/control, and other aspects of the research process. This need is even greater in cross-cultural research than it is in one-country research. Train the multi-country research staff together in one country before they work in another country. The reason is that some of the modifications in instructions to coders that are necessary to achieve comparable results, may have to be made in the first country rather than only in second or third countries.

22


(11) Special procedures are necessary to code both verbal and visual cues accurately. While this problem is especially acute for analysis of television commercials, it probably is necessary also for the differences in the verbal and visual content and executions in other advertising media, especially magazines. Do not make the assumption that most or all o f the information in the advertising is in the verbal part o f the advertising. This point is particularly important in cross-cultural research, since some cultures are more visually oriented than others, for example, Japan compared with the USA.

ACKNOWLEDGMENTS The authors gratefully acknowledge research grants and services contributed by the following organizations: the Michigan State University Development Fund, the Hoso Bunka Foundation (Japan), the Korea Broadcast Advertising Corp., A S I Market Research (Japan), the Yoshida Hideo Memorial Foundation (Japan), and the Youngshin A c a d e m y (Korea).

REFERENCES Abernethy, A. M., & Franke, G. R. (1996). The Information Content of Advertising: A MetaAnalysis. Journal of Advertising, 2(2), 1-17. Brislin, R. W. (1970). Back-translation for Cross-Cultural Research. Journal of Cross-Cultural Psychology, 1(3), 185-216. Budd, R. W., & Thorp, R. K. (1963). An Introduction to Content Analysis. A publication of the University of Iowa School of Journalism. Douglas, S. P., & Craig, C. S. (2000). International Marketing Research: Concepts and Methods. Chichester, U.K.: Wiley & Sons. Franke, G. R. (1992). Reliability and Generalizability in Coding the Information Content of Advertising. Asian Journal of Marketing, 1, 7-25. Hofstede, G. (1984). Cultural Dimensions in Management and Planning. Asia Pacific Journal of Management, January, 81-99. Holsti, O. R. (1969). Content Analysis for the Social Sciences and Humanities (pp. 94-149). Addison-Wesley Publishing Company. Hornik, J. (1988). The Delta Binomial Test for Interjudge Reliability. Current Issues & Research in Advertising, 11, 67-74. Hughes, M. A., & Garrett, D. (1990). Intercoder Reliability Estimation Approaches in Marketing: A Generalization Theory Framework for Quantitative Data. Journal of Marketing Research, 27(May), 185-195. Janis, I. L., Fadner, R. H., & Janowitz, M. (1943). The Reliability of a Content Analysis Technique. Public Opinion Quarterly, 7, 293-296. Kassarjian, H. H. (1977). Content Analysis in Consumer Research. Journal of Consumer Research, 4, 8-18.


23

Kolbe, R., & Burnett, M. S. (1991). Content Analysis Research: An Examination of Applications with Directives for Improving Research Reliability and Objectivity. Journal of Consumer Research, 18(September), 243-250. Krippendorf, K. (1980). Content Analysis: An Introduction to Its Methodology. Sage Publication. Kumar, V. (2000). International Marketing Research. Upper Saddle River, NJ: Prentice Hall. Miracle, G. E. (1988). An Empirical Study of the Usefulness of the Back-Translation Technique for International Advertising Messages in Print Media. In: J. D. Leckenby (Ed.), Proceedings of the 1988 Conference of the American Academy of Advertising (RC51-56). Austin: The University of Texas at Austin. Miracle, G. E., Taylor, C. R., & Chang, K. Y. (1992a). Culture and Advertising Executions: A Comparison of Selected Characteristics of Japanese and U.S. Televisions Commercials. Journal of International Consumer Marketing, 4(4), 89-113. Miracle, G. E., Taylor, C. R., & Chang, K. Y. (1992b). Culture and Advertising Executions: A Comparison of Selected Characteristics of Korean and U.S. Television Commercials. International Marketing Review, 5-17. Miracle, G. E., Maenaka, H., & Chang, K. Y. (1993). Information in Japanese and U.S. Television Advertising. In: E. Thorson (Ed.), Proceedings of the 1993 Conference of the American Academy of Advertising (pp. 245-246). Columbia, MO: American Academy of Advertising. Parmeswaran, R., Greenberg, B. A., Bellenger, D. N., & Robertson, D. H. (1979). Measuring Reliability: A Comparison of Alternative Techniques. Journal of Marketing Research, 16, 18-25. Perreault, W. D., & Leigh, L. E. (1989). Reliability of Nominal Data Based on Qualitative Judgments. Journal of Marketing Research, 26(May), 135-148. Robinson, W. S. (1957). The Statistical Measurement of Agreement. America Sociological Review, 22, 1717-1725. Rust, R., & Cooil, B. (1994). Reliability Measures for Qualitative Data: Theory and Implications. Journal of Marketing Research, 31(February), 1-14. Taylor, C. R., Miracle, G. E., & Wilson, R. D. (1994). The Impact of Brand Differentiating Messages on Effectiveness in Korean Advertising. Journal of International Marketing, 2(December), 31-52. Taylor, C. R., & Stern, B. B. (1997). Asian-Americans: Television Advertising and the Model Minority. Journal of Advertising, 26(Summer), 1-15. Taylor, C. R., Wilson, R. D., & Miracle, G. E. (1997). Culture's Consequences in Advertising: The Impact of Information Level Strategies on the Effectiveness of Korean vs. U.S. Television Commercials. Journal of Advertising, 26(Spring), 1-18.

TOWARDS THE DEVELOPMENT OF A SCALAR EQUIVALENT ETIC MULTICULTURAL ADVERTISING RESPONSE SCALE (MARS) Michael T. Ewing, Albert Caruana and Andy Teo

ABSTRACT While considerable prior research has focused on the development of standardized viewer response scales in advertising, such studies have, without exception, taken an emic approach. In other words, the scales have first been developed in one country, often the U.S., and then validated or replicated in other countries. Emic approaches have obvious limitations in an increasingly multicultural environment. By contrast, we offer a simple uni-dimensional advertising response scale developed following an etic approach, in which a universal measurement structure across cultures is sought using multiple cultures simultaneously. Psychometric tests demonstrate that the new scale is reliable, valid, parsimonious and generalizable across cultures and product categories. Theory-building and managerial implications of the approach are discussed, limitations noted and future research directions outlined.

New Directions in International Advertising Research, Volume 12, pages 25-41. © 2002 Published by Elsevier Science Ltd. ISBN: 0-7623-0950-4

25

26

MICHAEL T. EWING, ALBERT CARUANA AND ANDY TEO INTRODUCTION

Advertising response scales offer a useful procedure whereby consumers' aggregated responses to television commercials can be quantitatively compared and meaningful interpretations made. A number of popular advertising response scales have been developed over the past 30 years. However, most, if not all, have North American origins. The advertising function is increasingly being confronted with new and varied challenges resulting from the inexorable forces of globalization. Cross-cultural research and instruments that can be used with confidence across cultures are thus highly relevant to both academics and practitioners of advertising. Knowledge of similarities or differences in the ways that consumers within a multi-cultural market perceive advertising communications would provide regional marketers with valuable information regarding the standardization/ adaptation question. However, meaningful cross-cultural comparisons cannot be established if there is a lack of measurement invariance, and measurement instruments developed with ethnocentric (emic) approaches are bound to find little relevance in other cultures. The applicability of measurement instruments to other countries is also seriously questioned in an increasingly pluralistic marketplace. One cannot detract from the fact that the fundamental purpose of cross-cultural research should be to test whether concepts and consumer theories developed in one culture are equally valid and applicable in other cultures. In this respect, the ultimate goal of quantitative cross-cultural research lies in the equivalence of data. If data equivalence and evidence supporting a measure's invariance are lacking, then research should be confined to qualitative interpretation rather than conducting questionable quantitative comparisons because conclusions drawn are at best ambiguous and at worst erroneous. The primary objective of this study is to develop a Multi-cultural Advertising Response Scale (MARS) that is equivalent across cultures. Building on the popular method pioneered by Churchill (1979) for development of marketing constructs, we follow the Object-Attitude-Rater (O-A-R) procedure recently recommended by Rossiter (2000). Research is conducted among samples of Australian and Asian respondents for five standardized (global) television commercials. Results are reported, implications discussed, limitations are noted and future research directions suggested.

Advertising Response Scales Before the development of a new advertising response scale can be justified, it is necessary to review the various scales developed thus far. A review of the

Equivalent Multicultural Advertising Response Scale: MARS

27

literature highlights a wide array of measures that seek to capture viewers' responses to television advertising in particular. Among the most popular and frequently cited instruments are the Emotional Quotient (EQ) scale (Wells, 1964a, b; Wells, Leavitt & McConville, 1971), the Likeability scale (Leavitt, 1970) and the Viewer Response Profile (Schlinger, 1979). Work by Aaker and colleagues in this area, is also well regarded (see for example: Aaker & Norris, 1982; Aaker & Bruzzone, 1985; Aaker & Stayman, 1990). However, it was the pioneering work of Wells (1964) in particular that led to the popularization and use of rating scales in advertising. His Emotional Quotient scale originally consisted of 12 items that sought to measure the emotional appeal to print adverts. Subsequent work saw the development of 'Son of EQ' that consisted of twenty-six items that split into three factors: Attractiveness, Meaningfulness and Vitality that were shown to be significant predictors of recognition and recall. Wells, Leavitt and McConville (1971) developed an EQ scale version for TV commercials. The final analysis revealed six stable factors: Humour, Vigour, Sensuousness, Uniqueness, Personal relevance and Irritation. Leavitt (1970) focused on likeability of an ad as a means of predicting its attention getting ability. The seventy-one-item instrument that was developed resulted in seven 'attention getting' factors of a commercial, namely: Energetic, Personal relevance, Sensual, Familiar, Novel, Authoritative and Disliked. Schlinger's (1979) Viewer Response Profile (VRP) seeks to quantify consumers' subjective and affective reactions to advertisements. Unlike measurements of learning and comprehension, the VRP does not directly concern itself with the retention of claims, slogans, or other factual material but focuses instead on the emotional component of communication effects. The final thirty-two item scale consists of seven broad dimensions, namely: Entertainment, Confusion, Relevant news, Brand reinforcement, Empathy, Familiarity and Alienation. Puto and Wells (1984) proposed a theoretical structure of advertising effects based on the cognitive and affective elements of informational and transformational advertising that was used as the basis for their twenty-three-item scale. Ads can concurrently be informational and transformational in different degrees and the two dimensions are not mutually exclusive. For an advertisement to be deemed informational, it must present factual, relevant information about the brand, present information that is important to the potential consumer, and present data that the consumer accepts as being verifiable. An advertisement is judged to be transformational if it makes the experience of using the product richer, warmer, more exciting, and/or more enjoyable and it connects the experience of the advertisement to the extent that consumers cannot remember the brand without recalling the experience generated by the advertisement.

28

MICHAEL T. EWING, ALBERT CARUANA AND ANDY TEO

Aaker and Stayman, (1990) focus on television advertising and in looking at the factors that result from the instruments by Wells, Leavitt and McConville (1970), Schilinger (1979) and Aaker and Bruzzone (1981) recognise a remarkable degree of consistency across these studies. Four factors, namely: Informative, Dislikable, Warm, and Entertaining always seem to emerge, while factors labeled Familiar and Credible appear twice. Other factors such as Novel, Confusing, Stimulating, Brand reinforcing, and Clear are less consistent because items were not always included to tap them. Most advertising scales have been developed using exploratory procedures. Lastovicka (1983) first noted the paucity of published evidence of convergent and discriminant validity for any set of measures of viewer reaction to television commercials scales developed to date. He therefore set out to test this type of validity for three of the most popular copy-testing concepts, namely: Relevance, Confusion and Entertainment, using a 16-item scale. Lastovicka (1983) concludes that the underlying concepts, which the empirical measures were designed to tap, seem related to one another in a manner consistent with expectations. Zinkhan and Fornell, (1985) looked at the Wells EQ scale and Leavitt's commercial profile and examined the hypothesized factor structure and their predictive validity with respect to attitude and purchase intention. The EQ scale was found to be consistent with its postulated structure whereas the commercial profile was not (Zinkhan & Fomell, 1985). Nonetheless, the two profiles predicted attitude toward the brand equally well, but the reaction profile was superior with respect to purchase intention (Zinkhan & Fornell, 1985). In a later study, Zinkhan and Burton (1989) examined the dimensional structure and validity of the Wells' Reaction Profile, Leavitt's Multi-dimensional Profile and Schlinger's VRP. Findings indicated that the dimensions underlying Wells' Reaction Profile appeared stable, the Leavitt profile displayed acceptable stability, but the Schlinger profile did not appear sufficiently stable. EQUIVALENCE

IN CROSS-CULTURAL

RESEARCH

A review of the principal advertising instruments has indicated no universally agreed set of dimensions but a significant variety with some overlap at times. A further point is that all the research instruments have been developed within a U.S. context. While the importance of the U.S. market is not in doubt, the opening of previously sealed international borders, large migration streams, international tourism, increased cross-cultural communications, China's recent admittance into the WTO, and the general globalization of the economy all underline the increasing importance of multicultural marketing communications.


29

All too often, researchers have assumed that United States-based concepts and models were relevant to other countries without actual validation of model constructs or linkages and this assumption could lead to invalid cross-national inferences (Durvasula et al., 1993). Many studies have been exported from the West to non-Western countries but some of the issues examined in these studies can be of little relevance to non-Western cultures (Sin, Cheung & Lee, 1999; Van de Vijver & Leung, 1997). Measurement instruments developed with ethnocentric (eric) approaches are often bound to find little relevance in other cultures. There has often been a failure to realize that the very meaning of a scale might change and classical quality indicators such as reliability and validity might be strongly influenced by cultural factors (Salzberger, Sinkovics & Schelegelmilch, 1999). The same sentiment has been echoed by Steenkamp and Baumgartner (1998) who note that assessing the applicability of frameworks developed in one country to other countries is an important step in establishing the generalizability of consumer behavior theories. Indeed, it is common to find that all items of an instrument are assumed to be equivalent across cultures without any statistical checks to support this claim (van de Vijver & Leung, 1997). However, for such comparisons to be meaningful, the instruments used to measure the theoretical constructs of interest, must first exhibit adequate cross-national equivalence (Steenkamp & Baumgartner, 1998). For meaningful cross-cultural comparisons to take place it is necessary to undertake appropriate equivalence comparability and generalizability checks to establish measurement invariance. The resulting advertising instruments will be relevant across countries and will prove increasingly useful to both practitioners and academics. Scale Development Procedures Rossiter (2000) argues that Churchill's (1979) popular scale development method is but one 'cell' in a greater schema for measure development and has recently proposed an alternative Object-Attribute-Rater (O-A-R) procedure that focuses on first assessing content validity. In order to determine whether single items measures are sufficient, Rossiter (2000) specifies the need to follow a systematic content analysis procedure. This procedure also assists the researcher to decide whether unidimensionality and internal consistency are relevant. It is argued that the universally accepted practice of 'pruning' of items to improve a measure in the name of reliability could in fact detract from its validity. Perhaps the recommendation to compute reliability as a first assessment of an instrument is in part driven by the ease with which this can be computed. Of the three types of validity (content, predictive and construct validity),

30


Rossiter (2000) maintains that content validity is the most important as it has theoretically logical priority over construct validity. For Rossiter (2000), a 'concrete' object has a closely uniform meaning across respondents while a 'singular' object refers to either a single object or a set of reasonably homogenous objects. Attributes can also be either 'concrete' (such as age) or 'abstract' (such as Social Class or Personal Involvement). Constructs can result from abstract formed attributes (such as occupation, income, education and residence type in Social Class) or abstract eliciting attributes (such as: Important, Interesting, etc in Personal Involvement). The Response Profile type of measures used in advertising represents abstract formed attributes, that in the case of the VRP would include components such as Entertainment, Relevant news, Confusion, etc. Rossiter, (2000) argues that there is no need to employ factor analysis of many items to derive the components in the case of such constructs as these are 'concrete,' and can be represented by a single item. Rossiter (2000) further argues that the notion that single items must be unreliable is a misconception, and it is respondents, not items, that are unreliable. The use of multiple items to measure a construct results in over-measurement that can cause unwanted measurement interactivity with the true score, and thus reduce the validity of the measure (Rossiter, 2000). STAGES

IN DEVELOPMENT

OF MARS

Phase One: Obtaining Qualitative Data About Feelings Toward TV Commercials The first stage in the development of MARS consisted of obtaining qualitative data from a convenience sample of undergraduates from fourteen class sections of a university business school in Western Australia and Hong Kong (the Australian university runs programs in Hong Kong, Singapore and Malaysia). A total of 159 subjects were qualified and separated into two groups on the basis that they were: (a) international students from Asian countries; or (b) local Australian students of Western heritage. Local Australian students with an Asian heritage were excluded from this study as they could have assimilated into the local culture but still retained some of the traits from their country of origin. Their omission was primarily due to the consideration that the extent of assimilation was unknown and was beyond the scope of this study, and they could also possibly effect the survey results. The two final groups of students consisted of ninety-four and sixty-five respectively. In the first group, twentysix percent were male and seventy-four percent female, while the latter group consisted of forty-seven percent males and fifty-three percent female.


31

The initial ten commercials selected in this phase of the research covered a variety of goods and services such as computers (IBM and Apple), apparel (Levi Strauss and Nike), shaving needs (Gillette Sensor), air travel (Singapore Airlines), financial services (Visa), entertainment (Sony Playstation), automotive (Volvo) and soft drinks (Pepsi). These commercials were chosen as the subjects would have adequate experience with the various product categories. The specific commercials were chosen on the basis that they are global or standardised commercials which were developed to be shown all over the world (see Belch & Belch, 2001). According to de Mooij (1998), a worldwide brand name and uniform communication strategy is most appropriate and effective for culture-free products such as computers, wristwatches, pocket calculators, perfumes and credit cards. Current literature provides ample support that the ten commercials selected are indeed of a standardised and global nature (Belch & Belch, 2001; Campbell, 2000; Cardona, 1999; Cozens, 1998; Halliday & Petreeca, 1998; Jensen, 1998; Jensen & Cuneo, 1998; Lenihan & Kelly, 1998; Merrill & McCarthy, 1998). Each commercial used can be considered as a 'concrete' and 'singular' object (Rossiter, 2000). An open-ended questionnaire (see Appendix) was administered and the respondents were asked to describe how they felt with their own words after viewing each commercial. The words entertaining, confusing, relevant and likable were used as prompts as these words appeared consistently across the different scales by Schlinger (1979), Leavitt (1970), and Lastovicka (1983). Furthermore, subjects were told that they could use their native language or any language they felt comfortable with in expressing their feelings toward the commercials. It was beneficial to the international students from Asia as this facilitated the subjects' ease in expressing how they felt as English is generally not their first language. It also prevented the subjects from trying too hard to think of an appropriate word, and in the process lose their first spontaneous feeling toward the commercials. Phase Two: Selection of Scale Items A list of all the words used by the respondents without any attempt at changing or rephrasing them was compiled. A total of thirteen questionnaires were translated from Mandarin to English. This resulted in an initial yield of 300 words from both groups. To reduce the study to a more manageable size, duplications, slang, and synonyms were eliminated which resulted in a total of 142 words. A thesaurus was also used to assist in the refinement of words and phrases employed by the respondents. This procedure follows that employed by Wells (1964) in developing the EQ scale.

32

MICHAEL T. EWlNG, ALBERT CARUANA AND ANDY TEO

The list was then tabulated and words that appeared most frequently across the two groups of subjects were used as scale items. The resultant scale consisted of eight items with two of the scale statements worded negatively so as to reduce irresponsible or lazy response pattern tendencies (Churchill, 1979; Rossiter, 2000). A five-point graphical scale ranging from 'Strongly Disagree' (1) to 'Strongly Agree' (5) was added. The use of a five-point scale for MARS is in line with that employed in the pioneering work of Wells, Leavitt & McConville (1971) in the EQ for TV commercials. The fact that the questionnaire must be understood by the general public necessitates that the vocabulary must not be fancy, and the usual requirements of brevity and comprehensiveness urge that the questionnaire be as short as possible without omitting any important ideas (Wells, Leavitt & McConville, 1971; van de Vijver & Leung, 1997) The resultant draft scale comprising eight items has its origins in the Schlinger (1979) factors, but with some additions and modifications emanating from the series of fourteen focus groups conducted with students in Western Australia and Hong Kong.

Phase Three: Use of Experts The draft scale was then reviewed by a group of ten international advertising practitioners (two each from: Australia, Hong Kong, Singapore, South Africa and the United States). The experts were asked to examine the draft scale, thinking about how well it will translate into the different languages that they are familiar with, the potential problems they could foresee, and whether any changes should be made. The experts thought the draft scale was suitable, and in the context of testing a commercial against those criteria, it would travel well no matter the culture. There was also a general consensus that while the definition as to what is considered 'entertaining' or 'inspiring' may vary between cultures, the validity of those measures in the overall appeal of a commercial would generally transcend cultural barriers. The instrument appears in the Appendix.

Phase Four: Scale Administration Convenience sampling was used as MARS was administered to students enrolled in an undergraduate Commerce program. A total of 243 valid responses were collected that consisted of 124 Asian students and 119 Australian students. Sixty-six percent of respondents were female and thirty-four percent were male for the Asian students group, while the Australian students group consisted of 64% female and 36% male respondents. This procedure is similar to the earlier


33

work by Lastovicka (1983) and Puto & Wells (1984) where data was collected from undergraduates after a forced exposure to an advertisement in a theatre setting. It has been suggested that the use of student subjects in measurement development research threatens the external validity and generalizability of findings due to the non-representativeness and unique characteristics of the population (e.g. Well, 1993). However, in a cross-cultural study, well-matched (i.e. maximally homogenous within and between cultures) samples are more useful than representative samples because they allow more exact theoretical predictions and reduce the confounding effects of other factors (Hofstede, 1991). More importantly, students are accepted theory testing research subjects where the multivariate relationships among constructs, not the univariate differences between samples are being examined (Calder et al., 1981). Accordingly, students have been used as effective surrogates for non-students in a wide and diverse array of marketing and advertising studies (Yoo & Donthu, 2001). Respondents were qualified based on the fact that they were either Australians or ethnic Chinese of Southeast Asian origin. The subjects were shown a total of five commercials (Singapore Airlines, Visa, Nike, Sony Playstation and Pepsi), with each of the commercials played twice in accordance with industry copy-testing practices. The five commercials used for the administration of the scale were retained because they were thought to be most relevant to the target samples, and they would have adequate product usage experience. RESULTS Metric Properties of the MARS Instrument Confirmatory factor analysis was undertaken on the two covariance matrices that resulted from the Australian (pop 1) and Asian (pop 2) population. In line with Rossiter (2000) the CFA undertaken on the data assumed unidimensionality. Results (Table 1) show a good fit when a unidimensional construct is assessed but inspection of items indicated errors for the negatively worded items 3 and 11. If these negatively worded items are treated as composing a second separate factor, the fit obtained for the CFA for both groups improves. Although use is being made of a two-factor model, the 'second' factor made up of the negatively worded items is a separate factor only because of the negative wording of the items and the construct is still unidimensional (Christie & Geis, 1970; Robinson & Shaver, 1973; Saxe & Weitz, 1982; Welsh, 1956). A look at the Composite ot indicates that these are all greater than 0.5 providing support

34

Table 1.

M I C H A E L T. E W I N G , A L B E R T C A R U A N A A N D A N D Y T E O Results of CEA

for Five Ads Shown

One-Factor Population 1

and Two-Factor

to T w o P o p u l a t i o n s

Using a

Model.

Ad 1

Ad 2

Ad 3

Ad 4

Ad 5

O n e factor ×2 df p SRMR GFI AGFI CFI IFI Coefficient ot Composite ot VE Two factors ×2 df p SRMR GFI AGFI CFI IFI Coefficient etl Coefficient ct2 Composite etl Composite or2 VE 1 VE 2

45.11 20 0.0007 0.076 0.91 0.84 0.90 0.90 0.799 0.799 0.343

32.53 20 0.038 0.062 0.94 0.89 0.96 0.96 0.766 0.814 0.374

40.76 20 0.004 0.055 0.92 0.85 0.96 0.96 0.901 0.903 0.547

35.40 20 0.018 0.056 0.94 0.89 0.95 0.95 0.822 0.828 0.390

58.19 20 0.000 0.065 0.90 0.82 0.91 0.91 0.842 0.845 0.422

28.57 19 0.073 0.057 0.94 0.89 0.96 0.96 0.813 0.594 0.865 0.710 0.518 0.571

16.24 19 0.640 0.034 0.97 0.94 1.00 1.00 0.843 0.580 0.851 0.626 0.496 0.473

28.17 19 0.080 0.040 0.94 0.89 0.98 0.98 0.915 0.663 0.917 0.664 0.651 0.498

25.17 19 0.150 0.042 0.95 0.91 0.98 0.98 0.832 0.592 0.830 0.594 0.460 0.424

48.53 19 0.0002 0.052 0.92 0.84 0.93 0.93 0.884 0.4948 0.887 0.553 0.569 0.407

Population 2

Ad 1

Ad 2

Ad 3

Ad 4

Ad 5

One factor ×2 df p SRMR GFI AGFI CFI IFI

35.45 20 0.018 0.063 0.93 0.87 0.95 0.95

27.77 20 0.11 0.055 0.94 0.90 0.97 0.97

49.96 20 0.0002 0.066 0.90 0.83 0.92 0.92

41.77 20 0.003 0.062 0.92 0.85 0.93 0.93

35.92 20 0.016 0.043 0.94 0.89 0.96 0.96

Equivalent Multicultural Advertising Response Scale: MARS Table 1. Population 1 Coefficient ct Composite ct VE

Ad 1 0.813 0.818 0.373

35

Continued. Ad 2 0.811 0.817 0.372

Ad 3 0.867 0.867 0.453

Ad 4 0.824 0.831 0.393

Ad 5 0.878 0.880 0.482

Population 2

Ad 1

Ad 2

Ad 3

Ad 4

Ad 5

T w o factors ×2 df p SRMR GFI AGFI CFI IFI Coefficient txl Coefficient et2 Composite etl Composite ct2 VE 1 VE 2

20.06 19 0.39 0.05 0.96 0.92 1.IX) 1.00 0.790 0.663 0.788 0.733 0.394 0.602

19.57 19 0.42 0.043 0.96 0.93 1.00 1.00 0.827 0.496 0.831 0.693 0.453 0.591

28.62 19 0.072 0.046 0.94 0.89 0.97 0.98 0.867 0.717 0.867 0.731 0.522 0.581

30.62 19 0.044 0.054 0.94 0.89 0.96 0.96 0.779 0.718 0.768 0.587 0.370 0.422

32.56 19 0.027 0.040 0.94 0.89 0.97 0.97 0.864 0.661 0.864 0.662 0.519 0.496

Table 2. Population 1 (factor 1 with 2)

Ad 1

Test for Discriminant Validity. Ad2

Ad3

Ad4

Ad5

0.566 0.320 0.651 0.498

0.453 0.205 0.460 0.424

0.377 0.142 0.569 0.407

Ad4

Ad5

0.739 0.546 0.370 0.422

0.661 0.436 0.519 0.496

Correlation Correlation squared VE 1 VE 2

0.368 0.135 0.518 0.571

0.341 0.116 0.496 0.473

Population 2 (factor 1 with 2)

Ad 1

Ad2

Ad3

Correlation Correlation squared VE 1 VE 2

0.476 0.227 0.394 0.602

0.433 0.187 0.453 0.591

0.523 0.274 0.522 0.581

36

MICHAEL T. EWlNG, ALBERT CARUANA AND ANDY "lEO

for convergent validity. All loadings to dimensions are acceptable and t-values are high. The variance extracted (VE) is at times less than 0.5. However, this need not be of too much concern since: (1) we are not arguing for the independence of the two factors and although the factors are primafacie separate they are artifacts and the construct is unidimensional; (2) the VE test is sensitive to sample size and we have a sample size that is on the low side. Table 2 indicates that the squared correlations between the two factors was less than the average variance extracted for the scale therefore providing general support for discriminant validity (Yoo & Donthu, 2001).

Equivalence Multiple Group LISREL is appropriate for comparing measurement models from mutually exclusive groups of respondents that are clearly defined (J6reskog & S6rbom, 1989). We seek to investigate the equivalence of a measure of advertising across two different populations. The null hypothesis is that there are no differences between the perceptions across the different populations and that the resulting matrices are equivalent, i.e.:

HEadx: E(popl) = E(pop2) Acceptance of the null hypothesis can be interpreted as evidence of invariance across groups. Rejection of the null hypothesis necessitates the testing of increasingly restrictive hypotheses in order to identify the source of nonequivalence. Testing of the null hypotheses for the five different ads provides support for their acceptance - Table 3.

Table 3.

Results of Tests of Equivalence.

Hypotheses

HEadl

HEad2

I-l~ad3

HI~ad4

×z df p SRMR GFI CFI IF I

57.94 36 0.012 0.075 0.94 0.96 0.96

47.07 36 0.100 0.078 0.96 0.98 0.98

46.45 36 0.110 0.150 0.95 0.99 0.99

51.03 36 0.050 0.095 0.94 0.98 0.98

HEad5 93.01 36 0.00 0.097 0.91 0.93 0.93


37

DISCUSSION The purpose of this research was to develop a psychometrically robust and cross-culturally generalizable measure of viewers' responses to standardized television commercials. Our etic multicultural advertising response scale (MARS) comprises eight items loading onto a single dimension. This study has important practical and theory-building implications that benefit advertising evaluation in several ways. These are discussed shortly. MARS was applied to samples of Asian and Australian students using five global commercials to test for equivalence of the advertising response scale. Multiple-sample LISREL analysis was carried out on data from the two sample groups. Results indicate the presence of data equivalence, which contributes to the advancement and answers the call for meaningful cross-cultural comparisons. Evidence supporting the validity and reliability of MARS was also established, affirming Rossiter's (2000) claim that the applied quest in the real world was to find the shortest empirical measure, often a single-item measure that would predict well, in contrast to 'over-measurement' in the ill-advised pursuit of multi-item reliability. The MARS scale has considerable but qualified implications for multinational marketers and their advertising agencies. For the first time, a simple, culturally equivalent unidimensional advertising response scale has been developed and validated. Managers can therefore employ MARS to check for the 'standardizability' of television advertising campaigns. The scale can also be used by local managers who for example may have been given the option of choosing one of a number of campaigns developed by Head Office, i.e. those organizations practicing what Belch and Belch (2001) call 'pattern advertising.' However, it is important to point out that MARS is not detailed enough to be used in copytesting and advertising development (pre-testing) research. With only eight items, it simply does not offer the same diagnostic power of Schlinger's 32-item VRP for example. In fact, it has more in common with Schlinger's factors, than with her individual items. So, for example, one might establish, through the use of MARS, that a particular advertisement is not entertaining (item 1), and/or confusing (item 6). The researcher could then employ either the full 32-item VRP, or only the items associated with the 'Entertainment' and 'Confusion' dimensions. This way, the researcher is able to more precisely pinpoint the source of the confusion. MARS then, is a tool best suited for post-testing and tracking. In particular, multinational advertisers could employ MARS to establish normative post-test databases. Like most studies of a similar nature and scope, ours is not without limitations. Our sample size, while psychometrically acceptable, was still somewhat modest.

38


Moreover, while we feel that, like many researchers before us, we can justify student samples for this type of study, we acknowledge that a non-student sample would have various advantages. For example, older consumers' psychological attachments to brands (and hence advertising) may vary markedly from students, for several reasons (e.g. differences in length of exposure to the brand, nostalgic associations with the brand, information processing abilities). In addition, while our ads and product categories were quite diverse, they were certainly not exhaustive, and this could also be a limitation. Lastly, our choice of population groups (ethnic Chinese and Australians) is a further limitation. In a strict sense, the scale may be generalizable across only the two samples involved in this study. However, Ewing, Caruana and Zinkhan (2002) recently provided evidence of the cross-national equivalence of three of the Schlinger dimensions across five countries (Australia, England, Hong Kong, Singapore and Malaysia), but not of the U.S., so we would argue that MARS is generalizable in at least those same five countries. Despite these limitations, we hope our study will spur researchers to revise and validate our scale. In particular, we would be pleased to see further attention paid to M A R S ' external generalizability. Additional research should validate the measure using different ads for different product categories from different countries. In particular, we would like to see the instrument applied in the U.S. In addition, the measure needs to be validated among non-student samples. So, an agenda for future research might include: (i) translating the scale into different languages; (ii) collaborating with a global advertising research firm (such as Millward Brown); and (iii) applying the scale across more countries and product categories. There is also perhaps a need to validate the scale items through the use of focus groups, for example. Again, this would be expensive and logistically challenging, leading us to suggest it would only be possible in collaboration with an industry partner.

REFERENCES Aaker, D., & Stayman, D. (1990). Measuring Audience Perceptions of Commercials and Relating them to Ad Impact. Journal of Advertising Research, 30(4), 7-17. Aaker, D. A., & Bruzzone, D. E. (1985). Causes of Irritation in Advertising. Journal of Marketing, 49(2), 47-57. Aaker, D. A., & Norris, D. (1982). Characteristics of TV Commercials Perceived as Informative. Journal of Advertising Research, 22, 61-70. Belch, G. E., & Belch, M. A. (1998). Advertising and Promotion: An Integrated Marketing Communications Perspective (4th ed.). Irwin McGraw-Hill, USA. Campbell, L. (2000). Top footballers add flair to Nike global campaign. Campaign (April, 29).


39

Cardona, M. M. (1999). Visa still gets charge out of campaign. Advertising Age, 70(23), 6. Christie, R., & Geis, F. L. (1970). Studies in Machiavellianism. New York: Academic Press. Churchill, G. A., Jr. (1979). A Paradigm for Developing Better Measures of Marketing Constructs. Journal of Marketing Research, XVI (January), 64-73. Cozens, C. (1998). Saatchis beats JWT to retain pan-Euro Visa task. Campaign (November, 1). de Mooij, M. (1998). Advertising Worldwide (3rd ed.). Hertfordshire: Prentice Hall International. Durvasula et al. (1993). Assessing the Cross-National Applicability of Consumer Behavior Models: A Model of Attitude toward Advertising in General. Journal of Consumer Research, 19 (March) 626---636. Ewing, M. T., Caruana, A., & Zinkhan, G. M. (2002). On the Cross-National Generalizability and Equivalence of Advertising Response Scales Developed in the U.S. International Journal of Advertising, 21(3), 319-339. Hair, J. F. Jr., Anderson, R. E., Tatham, R. L., & Black, W. C. (1992). Multivariate Data Analysis (5th ed.). New Jersey: Prentice Hall. Halliday, J., & Petrecca, L. (1998). Volvo effort extends image positioning. Advertising Age, 69(13), 9. Jensen, J. (1998). Nike OK's $50 mil intro ad push for Alpha. Advertising Age, 69(40), 4. Jensen, J., & Cuneo, A. Z. (1998). Nike, W&K exec moves key to 'reinvent' plan. Advertising Age, 69(41), 24. J0reskog, K. G., & Strbom, D. (1989). LISREL 7: A guide to the program and applications (2nd ed.). SPSS Inc., Chicago. Lastovicka, J. L. (1983). Convergent and Discriminant Validity of Television Commercial Rating Scales. Journal of Advertising, 12(2), 14-23. Leavitt, C. (1970). A Multidimensional Set of Rating Scales for Television Commercials. Journal of Applied Psychology, 54, 427-429. Lenihan, R., & Kelly, J. I. (1998). Young & Rubicam gets major global assignment from Sony. Adweek, 39(4), 5. Merrill, C., & McCarthy, M. (1998). Final slice of Apple pie. Adweek, 39(26), 5. Puto, C., & Wells, W. (1984). Informational and Transformational Advertising: The differential effects of time." Advances in Consumer Research, 11,638~343. Robinson, J. P., & Shaver, P. R. (1973). Measures of Social Psychological Attitudes (Revised ed.). Ann Arbor, MI: Survey, Research Centre, Institute for Social Research. Rossiter, J. R. (2000). Construct Measurement in Marketing. Working Paper no. 0(O)1, Department of Marketing (April), 1-50. Salzberg, T., Sinkovics, R. R., & Schlegelmilch, B. B. (1999). Data Equivalence in Cross-Cultural Research: A Comparison of Classical Test Theory and Latent Trait Theory Based Approaches. Australasian Marketing Journal, 7(2), 23-38. Saxe, R., & Weitz, B. A. (1982). The SOCO scale: A measure of the consumer orientation of salespeople. Journal of Marketing Research, X1X (August), 343-351. Schlinger, M. J. (1979). A Profile of Responses to Commercials. Journal of Advertising Research, 19(2), 37-46. Sin, L. Y. M., Cheung, G. W. H., & Lee, R. (1999). Methodology in Cross-Cultural Consumer Research: A Review and Critical Assessment. Journal oflnternational Consumer Marketing, 11(4), 75-96. Steenkamp, J. E. M., & Baumgartner, H. (1998). Assessing Measurement Invariance in Crossnational Consumer Research. Journal of Consumer Research, 25(1), 78-90. Van de Vijver, F., & Leung, K. (1997). Methods and Data Analysis for Cross-Cultural Research. Sage Publications, U.S.A.

40

M I C H A E L T. EWING, A L B E R T C A R U A N A A N D ANDY TEO

Wells, W. (1964a). EQ, Son of EQ, and the Reaction Profile. Journal of Marketing 28 (October), 45-52. Wells, W. D. (1964b). Recognition, Recall, and Rating Scales. Journal of Advertising Research, 4, 2-8. Wells, W. D., Leavitt, C., & McConville, M. (1971). A Reaction Profile for TV Commercials. Journal of Advertising Research, 11 (December), 11-17. Welsh, G. S. (1956). Factor dimensions A and R. In: G. S. Welsh & W. G. Dahlstrom (Eds), Basic Readings on the MMP1 in Psychology and Medicine. Minneapolis: University of Minnesota Press. Yoo, B., & Donthu, N. (2001). Developing and validating a multi-dimensional consumer-based brand equity scale. Journal of Business Research, 52, 1-14. Zinkhan, G. M., & Burton, S. (1989). An Examination of Three Multidimensional Profiles for Assessing Consumer Reactions to Advertisements. Journal of Advertising, 18(4), 6-14. Zinkhan, G. M., & Fomell, C. (1985). A Test of Two Consumer Response Scales in Advertising. Journal of Marketing Research, 22(4), 447--452.


APPENDIX: MARS ITEMS 1. 2. 3. 4. 5. 6. 7. 8.

This commercial is entertaining. I feel inspired watching this commercial. This commercial is interesting. This commercial is annoying (R). I find this commercial memorable. I did not understand this commercial (R). I feel happy watching this commercial. This commercial is unique.

(R) indicates negatively worded item.

STANDARDIZED vs. SPECIALIZED INTERNATIONAL ADVERTISING CAMPAIGNS: WHAT WE HAVE LEARNED FROM ACADEMIC RESEARCH IN THE 1990s. Charles R. Taylor and Chad M. Johnson

ABSTRACT This paper provides a review of the academic literature regarding examining the standardization vs. specialization debate during the 1990s. While numerous studies have examined the validity of both strategies, it is clear that many companies are adopting an approach where broad strategy is standardized but actual advertising executions are localized as necessary. There is a need for future researchers to focus on how to effectively develop and implement standardized strategies as opposed to examining whether advertising should be standardized. Suggestions for further research are provided.

INTRODUCTION T h e issue of whether advertising should be standardized vs. adapted to local cultures has been a major focus in the advertising literature. While the debate

New Directions in International Advertising Research, Volume 12, pages 45-66. Copyright © 2002 by Elsevier Science Ltd. All rights of reproduction in any form reserved. ISBN: 0-7623-0950-4

45

46

CHARLES R. TAYLOR AND CHAD M. JOHNSON

regarding standardization of international advertising has been particularly heated since Theodore Levitt's 1983 article in Harvard Business Review, the argument has existed in international advertising circles for almost fifty years (Agrawal, 1995). Some believe that Levitt's assertion that the world is moving towards becoming a "global village" is questionable. Proponents of standardization cite that "worldwide transportation and communication are making consumer demands universal and cultural differences inconsequential." Specialization (or local adaptation) supporters believe that the true trend is towards micro-marketing and advertising campaigns that reach target customers on a more individual level (Moriarty & Duncan, 1990). Levitt's argument in favor of standardization is strikingly concise. Technological convergence is the driving force behind his argument in favor of standardization. The world is becoming increasingly cosmopolitan; therefore, people throughout the world are increasingly demanding the same things from their products, advertisements, etc. As the world becomes increasingly global, it is imperative that international marketers give customers what they want, not what they say. Cost savings, the globalization of markets and the need for global brands have been cited as key arguments for standardization (Levitt, 1983). In addition to Levitt's article, a great deal of early international advertising research provides support for the proponents of standardization (Elinder, 1965; Fatt, 1967). However, there is another side to the argument. Many international advertisers do not wholly concur with Levitt's assertion that the world is increasingly homogeneous. These marketers believe that differences and segments exist throughout the world. Different usage conditions, regulatory differences and varying consumer tastes all provide rationale for the specialization strategy. Many researchers have warned of the dangers regarding standardization and provided support for specialization strategies (Buzzell, 1968; Ricks et al., 1974; Reichal, 1989). Because of the differences of opinion described above, numerous researchers have turned to examining questions related to whether, and/or under what conditions it is appropriate, to standardize advertising programs. This research is generally deemed to be important for a few reasons. First is the clear need for multinational firms to engage in effective advertising in order to be successful. Regardless of cultural influences, advertising is effective when it elicits positive customer response, when it gains the customer's attention, creates some empathy and effectively communicates a message (Meyers, 1996). A second key issue related to advertising is that it is generally regarded to be the most difficult component of the marketing mix to standardize (Mueller, 1996). Among those holding a middle ground view on the standardization debate, there is an acknowledgement that standardized strategies must be carefully planned out.

Standardized vs. Specialized International Advertising Campaigns

47

For these reasons, the wide volume of research on standardization is warranted. The purpose of this paper is to analyze the academic research that has occurred during the 1990s regarding the debate of whether international advertisers should standardize or specialize their advertising campaigns. The goal of the paper is to provide a summary of knowledge accumulated in the academic research and to provide directions for future research. To perform this analysis an extensive review of the literature was conducted. Special emphasis was placed on studies focusing on variables that affect the decision to standardize or specialize international advertising campaigns. Among the areas that have attracted considerable attention are control issues, cultural factors (including individual cultural variables as well as general studies of the acculturation process and reaction to country of origin effects), and other factors including media effects and the impact of specific executional variables such as the use of humor and music on advertising effectiveness. The remainder of the paper will be composed of a review of the literature followed by a discussion of key findings from research conducted in the 1990s. Two issues that have received considerable attention - control issues and cultural issues - will be discussed separately. The paper will conclude with a discussion of directions for future research. REVIEW

OF LITERATURE

Control Issues Affecting International Advertising Strategy One area that has attracted considerable research in the 1990s is implementing organizational control when embarking on international advertising campaigns. While control is important to varying degrees throughout the organization, it has been found to be a particularly important consideration for advertising decisions (Samiee & Roth, 1992). Many managers of MNCs are finding the span of control associated with making advertising decisions affecting regions all around the world to be almost unbearable (Gould et al., 1999). The ability of companies and their ad agencies to maintain some semblance of control over information and implementation of strategies has been positively affected by trends such as new forms of information technology, enhanced usage of databases, media fragmentation, client desires for interaction/synergy and global and regional coordination (Kitchen & Schultz, 1999). The emergence of global integrated marketing communications has played a role in making control more complex. Technology and globalization trends continue at dramatic rates (Kitchen & Schultz, 1999), yet it has been asserted

48


that completely integrated communications may be nearly unattainable due to uneven media availability (Gould et al., 1999). While global integrated marketing communications is rapidly becoming a powerful tool at the disposal of international advertisers, it too creates control issues, as one must acknowledge that different cultures are at different stages of grasping its capabilities (Griffin et al., 1998). The client-agency relationship has received a lot of attention in recent literature on standardization as it relates to control. Certain cultures seem to readily accept the notion of delegating power to ad agencies. For instance, the Japanese utilize their ad agencies somewhat like a full-service marketing company to a greater extent than companies in Chile and the United States. Some of the other marketing services that the agencies can provide include sales promotion, public relations and marketing research. The main reason why many companies are unwilling to delegate more responsibilities to their agencies is that they fear a loss of control (Griffin et al., 1998). There has been a significant trend towards more integration between the client and the ad agency, but it has been observed that clients need to be careful of suffocating the creativity of the agencies (Kitchen & Schultz, 1999). Despite their experience, many ad agencies feel as if their ability to influence the standardization-specialization decision is "negligible" (Rosenthal, 1994). While many clients are reluctant to relinquish control to the agencies (Griffin et al., 1998), agencies tend to have a different perspective in regards to delegating to their affiliates. Due in large part to expenses and coordination issues, most ad agencies are regional. It has been argued that few agencies feel comfortable executing globally; therefore, most parent agencies have a "hands ofF' approach to supervising their foreign affiliates (Hill & Shao, 1994; Tai, 1997). The structures of the advertising agency and the corporate client are critical components that determine the effectiveness of international advertising. The unwillingness to centralize management is one of the biggest impediments to successful standardization strategies. The "not made here" phenomenon often plays a detrimental role in the coordination and planning of global campaigns. The outlook for the future of international advertising suggests that a few "mega-agencies" will hold a majority of international advertising accounts (Moriarty & Duncan, 1990). At least one study concluded that, ultimately, international advertisers need to consider how the client-agency relationship will impact the targeted audience, and these considerations need to influence relationship decisions (Gould et al., 1999). The power and influence of the advertising agency plays a crucial role in whether an international advertising campaign is ultimately standardized. A direct correlation has been found between the level of client pressure and the


49

overall use of standardized techniques. Ad agencies often strongly believe that creative impact should play the decisive role in the standardization-specialization decision rather than client pressures. However, client pressures (both cost and time-oriented) are increasing and appear to be leading to more standardized international campaigns than used to be the case (Duncan & Ramaprasad, 1995). In many ways, this makes perfect sense, as global strategic planning for large MNCs needs to be centralized at some level. Another control issue associated with standardized ad campaigns is measuring effectiveness. While overall ad agency performance is important, the most crucial measurement of interest to clients is the effectiveness of the advertisements that have been implemented (Davison & Grab, 1993; Meyers, 1996; Rice & Bennett, 1998). Evaluating international advertising effectiveness is a formidable task, and one that presents special challenges for standardized campaigns. International advertisers must deal with inherent country inconsistencies at the start of the advertising research process. Notably, it has been asserted that differences in what is affordable and what is possible in a given market must be established. It is simply not realistic, for example, to compare the same measures of effectiveness for a product that is new to a market versus one that is at the mature life cycle stage in another. Moreover, the country's advertising infrastructure in terms of level of development of a research industry as well as media penetration and availability must be taken into account. It has also been observed that it is imperative to anticipate the impact of the country's culture on the data (Meyers, 1996). Clearly, many global advertisers recognize the difficulty in measuring and comparing the success of global advertising campaigns. Some attempts are being made by the advertising industry to standardize measurement principles, as evidenced by five international advertising and television trade groups joining together in the most ambitious effort to date to standardize measurement principles for global television advertising (Ross, 1996). Nevertheless, the ability to measure the success of standardized campaigns in different countries remains a significant challenge. In summary, the 1990s saw research pay increased attention to control issues, which have become more complex as more firms have conducted standardized campaigns. The trend toward integrated marketing communications along with different levels of media and research infrastructures has made control an increasingly challenging issue. The research suggests that these challenges are even more difficult for those ad agencies that are located regionally and that the structure of the advertising industry often leads to clients driving the push toward standardization in spite of greater integration between clients and agencies. In fact, it has been found that the level of client pressure is closely linked to the use of standardized programs. Finally, it has been found that

50


measuring effectiveness also creates a control challenge in that it is often difficult to use common effectiveness measures as a result of differences in market characteristics and research infrastructure.

The Role of Culture and the Environment Culture is a critical issue in the standardization-specialization debate, yet it is difficult to define it and even more arduous to determine its impact (Roth, 1992; Harris, 1994). It has been suggested that classifying markets based on cultural assumptions is useful (Roth, 1992). Marketing is an extremely powerful force that can create significant change within cultures, and the issue of cultural imperialism arises as advertising originating in the dominant countries of the global business landscape infiltrate all corners of the globe. (Moriarty & Duncan, 1990).

General Perceptions of Advertising It has been noted that even countries that share geographic proximity can be quite different on a cultural level (Huff & Alden, 1998). International advertisers need to be sensitive to the far-reaching effects of their campaigns. In sharp contrast from the classical view of economic theory, some believe that dependency theory better describes the impact of international advertising on developing economies. Dependency theory suggests that developing countries are dominated by the messages of established countries, which results in the transfer of some of the dominant countries' cultural beliefs (Tansey & Hyman, 1994). Western advertising professionals have been accused of relying upon Western assumptions and erroneously believing that Western practices will automatically translate to foreign lands (Wells, 1994). It has been suggested that countries like China are "melting pots" for Eastern and Western cultural values (Cheng & Schweitzer, 1996). This phenomenon can also be seen in Japan because their advertising depicts both Eastern and Western values (Ford et al., 1998). Several studies have confirmed the notion that other cultures throughout the world do not necessarily share the same sentiments regarding advertising as Americans do. Therefore, global advertisers need to be keenly aware of these differing perceptions because they will affect how various cultures perceive advertisements. Thus, global advertisers must assess the cross-national applicability of advertising messages (Durvasula et al., 1993). Even a country like Great Britain that is quite similar to the United States in many ways has some distinctly different values from those that Americans embrace. To


51

illustrate this point, American ads deploy direct speech and emotional appeals, yet British ads more often utilize indirect language and humor (Caillat & MueUer, 1996). Rapidly changing countries such as Russia present particularly interesting challenges for international advertisers. In general, the Russian culture is more receptive to advertising than are Americans (Andrews et al., 1994). Great Britain also has a more favorable impression of advertising than its American counterparts (Nevett, 1992). On the other hand, a country like Saudi Arabia has many citizens who are not very receptive to certain types of Westem advertisements (AI-Makaty et al., 1996). The clear differences across countries that have been found in terms of general attitudes toward advertising point out the need to carefully finesse standardized approaches. In developing countries, advertisers may need to be sensitive to possible perceptions of excessive foreign influence on the culture. Even in more developed countries such as Japan and Taiwan, advertisers need to consider the impact of using "Western" approaches in societies characterized by the mingling of Eastern and Western values in advertising in recent years. Per capita ad expenditures also play a critical role in determining how different cultures view advertising. Countries, such as Greece and India, with low per capita ad expenditures are more concerned with the overall economic and social impacts of advertising than are people in the United States. People in cultures with high levels of advertising spending like the United States are more concerned with advertising practices. The net result is that individuals in countries with low levels of advertising expenditures view ads in a much more function-related manner (Andrews et al., 1991). Cultural Factors and Their Impact on the Ability to Use Standardized Advertising Numerous studies have hypothesized that cultural factors lead to differences in the types of advertising that are effective in different countries. Various cultural variables have been proposed to lead to differences in reactions to advertising executions: gender roles, Edward T. Hall's concept of context, and individualism vs. collectivism. Japan has had a long history of tradition-bound gender stereotypes. However, recent research suggests that these gender stereotypes are becoming less frequently utilized in advertisements. The net result is a more egalitarian view of the genders (Ford et al., 1998). Nevertheless, Maynard and Taylor (1999) found that portrayals of young Japanese females depicted in Seventeen magazine ads differed from portrayals of American females in the version of the magazine published in the United States. While the Japanese females were presented

52


as more traditionally "girlish," the American models were shown in more avantgarde, mysterious, or contrary poses. Cultures can be classified as either high or low context. High context cultures, such as South Korea and Japan, are intuitive and prefer indirect messages (Miracle, Chang & Taylor, 1992). Johanson's (1994) observation that the Japanese prefer a soft-sell approach is consistent with this cultural dimension. On the other hand low context cultures, such as the United States and some Western European countries, rely heavily on explicit communication (Taylor et al., 1997). The consequence for international advertisers is that they must consider these fundamental differences in how different types of cultures respond to the levels of information content. The United States is a country that is receptive to information-laden appeals. However, other countries such as France have been found to respond relatively more favorably to emotional appeals (Biswas et al., 1992; Taylor et al., 1996). Meanwhile, the Indian culture has been found to have a higher need for formality in persuasive communications than do citizens of the United States (Bandyopadhyay et al., 1994). The type of culture is very important in determining the persuasive effect of emotional appeals in international advertising. Numerous studies have focused on the impact of cultures being either collectivist or individualistic in nature (Wilcox et al., 1996). The individualism-collectivism variable depicts how individuals in a culture construe their concept of self. "Ego-focused" and "otherfocused" are terms that can be used in order to distinguish the type of emotions that a society's individuals rely upon. Ego-focused emotions, which tend to be focused on one's internal state, include feelings such as pride and anger. Otherfocused emotions, which tend to be associated with others in a social context, include emotions such as empathy and indebtedness. Ego-focused appeals lead to more favorable attitudes within individualistic cultures. Other-focused appeals lead to more favorable reactions within collectivist cultures (Aaker & Williams, 1998). It is important to note that in collectivist cultures, consumer behavior is determined in large part by social normative factors (Huff & Alden, 1998). The role of culture type is imperative for international advertisers to understand as they seek to determine which cultures are compatible enough to standardize advertising campaigns. It has been established that the relationship between culture and communication is important in determining the success of international advertising. Advertising creates a "transfer of value through communicative connections between what a culture conceives as desirable states of being and products" (Tansey et al., 1990). Cultures express their purposes and conduct their affairs via communication. The common symbols of communication are often not obvious to those outside of the culture (Moriarty & Duncan, 1990). Additionally,


53

the extent to which a culture is familiar with a brand should significantly impact how the message of an advertisement is communicated to the market. It has been suggested that a market that has previous experience with a brand will be more likely to receive it favorably than those who have not been previously exposed to the brand (Rice & Bennett, 1998). Therefore, it is obvious that international advertisers need to make concerted efforts to understand foreign cultures and properly tailor their communications to create the most effective advertisement.

The Acculturation Process and Standardization The acculturation process describes the process of two or more cultures being assimilated together. As the world becomes increasingly global and transient, it has been posited that the notion of acculturation will become more relevant to international advertisers as they seek to target more diverse markets. It has been suggested that the process of acculturation is somewhat slow, and the rate at which immigrants become accustomed to their new culture depends largely on the compatibility with their original culture. Language barriers and different processing styles are two possible causes of the slow rate of acculturation (Lee & Tse, 1994). International marketers seek to determine how consumer needs should impact their advertising strategies. A depth strategy is utilized when marketers base their images on a single set of consumer needs. On the other hand, a multiple set of needs is used in breadth strategies. The normative model of brand image management suggests that images should be based on a single set of consumer needs thus providing support for the depth strategy. However, generalizations regarding depth strategies should be limited because effectiveness does vary based upon the environment. There are some indications that breadth strategies may be more effective in low context cultures because explicit information plays a pivotal role in this type of culture (Roth, 1992).

Country of Origin and Standardization It has been well documented in the 1990s that the country of origin's image can have a significant impact on the how cultures throughout the world perceive the advertising message (Tse & Lee, 1993; Parameswaran & Pisharodi, 1994; Klein et al., 1998). It has been suggested that the level of animosity towards the producer country can predict consumers' likelihood of purchasing products. The consequence for international marketers is that they must understand the military, economic and diplomatic relationships between the producer's country and the target market's country. It is noted that while the United States is a largely admired country throughout the globe, there are some countries that do not hold

54


the United States i/a high esteem. Hiroshima and Hanoi are two markets in particular that have been questioned regarding the viability of American marketing opportunities (Klein et al., 1998). There are ways for marketers to override or at least mitigate the effects of negative country of origin effects by decomposing the "made in" statement into separate component and assembly information (Tse & Lee, 1993). On the other hand, there are potential opportunities that are created by the country of origin variable, yet these advantages have been underutilized due to "deficiencies in the definition and measurement of its facets." Enhanced knowledge of the country of origin effects can enable international advertisers to better understand the market, improve product positioning and develop more effective messages (Parameswaran & Pisharodi, 1994).

Other Pertinent Issues Affecting International Advertising Strategy Media Issues Even in similar magazines, print ads often vary across cultures (Tansey et al., 1990; Biswas et al., 1992). For example, French print ads typically deploy sex appeal and humor in the form of jokes. On the other hand, American print ads tend to rely on information and humor stemming from puns and satire (Biswas et al., 1992). However, research has shown that there are some areas in magazine advertising which support standardization. For instance, the magazine ads from the United States and Sweden both tend to portray their models as being very young and in a very positive manner. Additionally, American advertisements are quite Similar to magazine ads run in the Unites States more than twenty years ago (Wiles et al., 1996). Thus, significant opportunities for longstanding standardized campaigns seem to exist under the aforementioned circumstances. It would seem logical that many print ads would contain a significant influence from the home-country culture. However, research has shown that this homecountry effect is not very influential in Japanese advertisements. On the contrary, one study found that German advertisers are much more likely to allow their own culture to affect their print ads. The international advertising impact is that Japanese firms are much more likely to deploy specialization strategies, and German firms are more likely to use standardization techniques (Graham et al., 1993). Television advertisements are a dominant component of domestic and global marketing campaigns for American companies. Due to the fact that "few things are seen by more of a country's population than TV advertising," the significance of television advertising cannot be emphasized enough (Meyers, 1996). While television is virtually a global phenomenon, cultures have distinct ways of presenting programming and advertisements. Even in similar cultures such


55

as the United States and Great Britain differences exist. For instance, the British culture responds more favorably to ads that contain less information, deploy a soft sell approach and attempt to entertain the audience (Nevett, 1992). Also several attractive markets such as Japan have fewer network affiliates than the United States. Fewer network affiliates and little or no cable penetration serve to produce fewer media alternatives than are available in the United States (Johansson, 1994). Consistent with the aforementioned differences in attitudes toward advertising in general, there are some countries that are less receptive to television advertising. For example, some Saudi Arabians view television advertisements as threats to their social infrastructure and economic development, and there are other Saudis who feel that television advertising can help to accelerate modernization. It is clear to see that in a country such as Saudi Arabia where television advertising has only been permitted since 1986, global advertisers need to be sensitive to the local market's perception of television ads (A1-Makaty et al., 1996). In some cases, fundamental differences in how television ads are portrayed exist based upon the advertisement's country of origin. For instance, Chinese television commercials utilize symbolic cultural values. On the other hand, American commercials tend to rely on both symbolic and utilitarian values. However, there is some room for common ground between Chinese and American commercials. Western values like "modernity" and "youth" are utilized more frequently in Chinese commercials than Western values like "enjoyment" and "sex" (Cheng & Schweitzer, 1996). Executional Factors: Humor and Music Humor plays a significant role in advertising, and it has been shown to often be particularly effective in television advertising. Script-based semantic theory suggests that humor often creates incongruity for the audience on a global level. In other words, the basic cognitive structure approach may not be "culturebound" and, consequently, humor has the potential to be quite effective for international advertisers. It has been noted that international advertisers should consider grouping different countries based upon their value sets in order to determine what type of humor would be most effective (Alden et al., 1993). A possible way to group countries is based upon the level of informativeness and the level of emotion (Biswas et al., 1992). Music plays a pivotal role in adolescent consumer socialization, and music television in particular plays a critical role in consumption imagery. The genre of music video influences the type of imagery utilized. Dance videos use fashion-oriented imagery. Heavy metal videos contain the most direct references to brands. Rap videos contain the most frequent use of a blend of verbal and

56


visual references to consumption. Consumption imagery in music videos also varies across cultures. This is evidenced by the fact that there is more consumption imagery in American music television than in Sweden's (Englis et al., 1993). It is critical for international advertisers to recognize the significance of the interrelationship between music and culture. For instance, television ads in the Dominican Republic are "more likely to contain music, to have lyrics, to have extensive lyrics and to have self-related meaning" than American advertisements. Research indicates that these differences can be attributed to indigenous differences between the two cultures (Murray & Murray, 1996).

FINDINGS This section examines insight that has been gained on the general debates on whether, and when, to standardize.

Analyses of the Standardization Approach to International Advertising While the literature is beginning to converge on the notion that standardization of broad strategy or main selling proposition is often desirable, the issue of the level to which specific executional techniques can be standardized still gains attention. Research indicates that partial standardization is widespread, yet total standardization is quite rare (Harris, 1994). Harris' study is particularly important in that it effectively makes the case that most multinationals do use standardized advertising but that the extent to which they do it varies. International advertising managers who are "non-culturally" oriented use the standardized strategy quite frequently (Kanso, 1992). Markets that do not differ cross-nationally regarding a range of cultural variables lend themselves to standardization strategies (Roth, 1995). It has been recommended that practitioners should place a greater emphasis on how they standardize rather than whether they do. The "how" refers to decisions such as the type of creative format that should be used to convey the message (Harris, 1994). Economies of scale, message consistency and the ability to attract crossnational market segments have predominated the rationale for seeking to standardize over the years (Levitt, 1983). Other benefits include control over advertising content, stronger brand images and simplified strategic planning (Tansey et al., 1990). International advertisers of the 21st century must seek to ascertain when global campaigns using similar executions can be effective. It


57

has been asserted that standardized campaigns are effective when multicultural segments perceive the messages in "semiotically-equivalent ways" (Domzal & Kernan, 1993). Segments that share a common ethos should be the targets of standardized campaigns. The world's economically elite consumers and the postWorld War II generations in Western cultures are two segments in particular that are becoming increasingly homogeneous across the globe (Domzal & Kernan, 1993). However, only a limited number of studies have focused on reaching cross-national market segments. Clustering techniques have become quite popular methods to enhance standardization strategies by grouping countries (Sriram & Gopalakrishna, 1991; Zandpour & Harich, 1996). Countries that are not part of a common geographic region may still exhibit similar preferences in spite of cultural differences and a lack of regional proximity and, thus, are candidates for standardized campaigns derived from clustering (Zandpour & Harich, 1996). While these clustering techniques have proven to be of some use, it is notable that very little insight on how to develop and communicate to segments that cut across markets was developed in the 1990s. There are certain market conditions that favor the standardization approach for international advertising. Some researchers believe that global consumer segments are continually emerging and becoming more homogeneous. In fact, global consumer culture positioning is a construct that has received wide acceptance because it associates a brand with a distinct set of symbols that represent the emerging global culture (Alden et al., 1999). Industry factors such as high rates of technological change and short product life cycles also favor standardization (Samiee & Roth, 1992). Standardization opportunities have been found to be quite attractive in less affluent, developing markets as well (James & Hill, 1991). As has been known for some time, the product category is one of the most important factors in the standardization-specialization debate. Universal products, upscale products and state-of-the-art technology lend themselves to attractive standardization opportunities (Moriarty & Duncan, 1990). Product universality has been labeled a "necessary but not sufficient condition" for standardization (Tansey et al, 1990). Other product categories that are particularly conducive to standardized appeals are industrial products as well as new products (Moriarty & Duncan, 1990). Research has suggested that products that enable consumers to demonstrate a certain amount of self-expression are often viewed in similar ways across various cultures and are consequently attractive standardization candidates. Based upon their relationship to an individual's self-expression, food and fashion are two product categories in particular that seem to appeal to some consumer segments across cultures (Domzal & Kernan, 1993).

58


Global finns and those with a market extension philosophy are better suited to standardize their campaigns (Samiee & Roth, 1992). Standardized strategies also seem to be appropriate for low information content and image advertisements (Tansey et al., 1990). A message in which the principal message stresses a unique benefit of the product that is distinct from that offered by any competitor is a brand differentiating message. Based upon the fact that brand differentiating messages may not be a culture-bound phenomenon (Taylor et al., 1994), advertisements that utilize brand differentiating messages that can be perceived similarly across cultures are sound candidates for standardized campaigns.

Analyses of the Specialization Approach to International Advertising The question of the level to which specific executions need to be localized remains important. General proponents of specialization note significant differences in customers, cultural and socioeconomic conditions and market structures as evidence for the attractiveness of specialized campaigns (Roth, 1995). Other arguments for specialized techniques revolve around the notion that specialization is essential to the basic principle of the marketing concept. Supporters of specialization strategies insist that marketers must be sensitive to individual needs and advertising that portrays values of the indigenous culture tends to be more effective (Tansey et al., 1990). "Culturally" oriented international advertising managers are more likely to utilize the specialization strategy (Kanso, 1992). In fact, one study estimated that two out of every three international advertising campaigns are specialized for local market conditions (James & Hill, 1991). Some argue that the globalization debate is "empty" because individual managers must make individual decisions (Graham et ai., 1993). Enhanced technologies, such as integrated marketing communications, are enabling international marketers to gain a more complete understanding of the idiosyncrasies that differentiate markets. The result is that specialization strategies are more viable via these technologies (Kitchen & Schultz, 1999). Product categories that are most culturally bound, and subsequently difficult to standardize, include consumer products, established products and products with simple technologies (Moriarty & Duncan, 1990). Specialization has been deemed to be more necessary when the status and positioning of local brands are quite varied (Davison & Grab, 1993). The rationale for utilizing specialized advertising strategies with consumer goods is based upon the fact that consumers infer "psychological meaning" from goods to reflect their own values (Tansey et al., 1990).


59

Determining Which Strategy is Appropriate Cho et al. (1999) found that the content of Korean ads suggests that while a broad global theme could be effective for the Korean market, adjustments linked to culture are likely to be needed to be successful. In summary, they state that their findings support the arguments of James and Hill (1991), Duncan and Harris (1995) and Tai (1997) that the design of international advertising should be based on a common global theme, but the executions often have to be modified. This thinking represents the prevailing view among leading academics and practitioners. The literature of the 1990s suggests that companies have standardized their advertising campaigns to varying extents. Coke utilizes a truly global strategy. After deciding that a specialized advertising strategy resulted in a fragmented global image, Kodak recently embarked on a truly global standardized campaign for the first time in its history (Kim, 1998). By virtue of its western European campaign for example, Levi's deploys a less comprehensive standardized approach than does Coke (Hill & Shao, 1994). Levi's carefully considers the varying importance of social status and affiliation across cultures when advertising its Dockers (Roth, 1995). Colgate-Palmolive, Gillette and Philips are examples of companies that use regional campaigns (Hill & Shao, 1994). However, it is clear that many companies are shifting to using a general theme globally while making necessary adjustments (Harris, 1994). Studies have long analyzed the impact of culture when deciding whether to standardize or specialize advertising campaigns; however, it has been suggested that companies need to place greater emphasis on understanding the effects that their strategies have on their brand's image (Roth, 1995). One of the key issues regarding the debate of whether to standardize or specialize international advertising campaigns is definitional (Moriarty & Duncan, 1990; Samiee & Roth, 1992). Levitt's initial definition of standardization was "selling the same product the same way everywhere" (Levitt, 1983). On a strictly literal basis, standardization critics have compelling argument. Very few, if any, products are capable of being standardized based upon Levitt' s perhaps "idealistic" definition (Moriarty & Duncan, 1990; Tansey et al., 1990). Of course, standardization of any aspect of the marketing mix in its truest form is difficult. Most finns that embrace the idea of standardization seek a reasonable facsimile (Hill & Shao, 1994). McDonald's Corporation operates in fifty-nine countries on a relatively standardized platform. The company has standardized positioning strategies (Samiee & Roth, 1992). However, it still pays attention to local tastes. For example, the company offers beer on its menu in many markets throughout Europe, and it offers the new "bulgogi burger" in South Korea and pork-based sandwiches in the Czech Republic.

60


In terms of advertising, it has been noted that certain phases of the international advertising process are more apt to be standardized than are others. For instance, strategy is the component of the process in which standardization is most often deployed. Standardization is utilized less often in execution and least often in language (Duncan & Ramaprasad, 1995). On the other hand, researchers have pointed out that companies that ignore a market's fundamental cultural and religious beliefs will do so at their own peril (A1-Makaty, 1996). In most cases well-designed general strategies do not violate such beliefs. However, numerous studies do suggest that even when targeting similar segments across cultures variations in receptiveness to advertising styles and content exist and need to be taken into consideration. This is due to the fact that cultures vary among dimensions such as Hofstede's individualism, uncertainty, avoidance, power distance and masculinity (AlbersMiller & Gelb, 1996). Many advertisers believe that countries that share geographical proximity present opportunities for standardization; however, it has been found that this generality is often invalid (Keillor et al., 1996; Huff & Alden, 1998). Even when dealing with the youth segments in two neighboring countries such as the United States and Mexico, close attention must be paid to inherent differences in culture and attitudes. In fact, some research suggests that specialization is advisable when attempting to advertise to youth in both the United States and Mexico (Keillor et al., 1996). One study analyzing the relatedness of advertisements in the United States and Great Britain supports the view that specialization is more often the proper decision when attempting to advertise on an international basis (Calliat & Mueller, 1996). The consumers of these two cultures are exposed to distinctly different value sets and styles of advertisements. Therefore, despite the strikingly similar marketing factors of the United States and Great Britain, the study concluded that differences in culture alone are significant enough to force companies to utilize specialized campaigns (Caillat & Mueller, 1996). An analysis of the footwear industry provides evidence that standardized and specialized strategies can both be effective, even in the same industry. Nike maintains a standardized fitness and performance image in all of the markets that it operates. Nike leverages its singular functional image throughout the world. On the other hand, Reebok customizes its image on the basis of differences it perceives in customer preferences. Reebok's management believes that important differences exist not only across continents but also within countries and regions (Roth, 1995). It is clear that companies must carefully study each national market before making its decision regarding the level of adaptation, if any, that needs to be made (Tansey et al., 1990). It is pertinent to note that many international adver-


61

tisers lack specific guidelines regarding the instances when they should seek to standardize and when they should choose to specialize their campaigns (James & Hill, 1991). A clear need exists for some sort of framework to help guide decision makers through the difficult process of international advertising. To date, no widely accepted contingency framework to guide advertisers in making the decision on the extent of standardization to employ exists. The Standardization~Specialization Continuum It is fairly obvious that, at least in some respects, global consumers are converging, yet substantial differences between cultures still exist. In reality the standardization-specialization dilemma is a complex decision because it can be used in varying degrees across different components of the campaign and in different regions (Duncan & Ramaprasad, 1995). The result is that an increasing number of finns have adopted the notion of global strategies, but regional or local executions (if needed). Most finns seek to standardize as much as possible because they seek to gain economies of scale (Moriarty & Duncan, 1990). It has also been suggested that standardization is a "matter of degree" (Samiee & Roth, 1992). When international advertising practitioners cannot decide on either standardization or specialization entirely, they must choose from what has been called the "contingency perspective" (Agrawal, 1995). A perfect example of a company embracing a standardization strategy, but simultaneously fine-tuning its ads to appeal to local taste is IBM. The company has recently embraced the notion of becoming a global marketer by merging its business and marketing communications strategies into a singular worldwide strategic and marketing campaign. IBM utilizes the same footage in each country, but they add the local language to their advertisements via subtitles. IBM has discovered that customers throughout the world respond very similar to these campaigns that utilize a standardized appeal (McCullough, 1996). Often the standardization-specialization debate becomes a functional struggle. At the most basic level, the debate revolves around tradeoffs between cost savings of standardization and increased revenues generated by market segmentation. Managers with cost responsibilities may push for standardization while marketing managers might argue for specialization and more narrow segmentation (Graham et al., 1993). Many practitioners believe that there are many potential combinations of ways that executional elements can be standardized. In practice, even those who embrace standardization strategies, like IBM, do not adhere to Levitt's strict definition of standardization. Instead, standardization is often a "flexible and adaptive" policy (Harris, 1994).

62

CHARLES R. TAYLOR AND CHAD M. JOHNSON CONCLUSIONS

Major Findings Regarding the Debate After researching and assessing a representative collection of the international advertising academic literature of the 1990s, it is apparent that MNCs have moved toward combination approaches in which broad strategies of selling propositions are standardized but specific executions are adapted when it is needed and some aspects of advertisements such as language and nationality of the models utilized are almost always adapted. Historically, much of the academic research regarding this ongoing debate takes one of the two bipolar positions. Clearly, it is time to get past such thinking and focus more on the circumstances in which global strategies and global executions are prone to work. While some may continue the debate regarding whether standardization or specialization techniques are more effective, it is time to conceptualize this debate in terms of the Standardization-Specialization Continuum. More work on developing a widely accepted framework that delineates key situational factors relevant to the standardization decision and incorporates the various components of the process (strategy, execution, language) is needed.

Directions for Future Research Due to the volatile and dynamic nature of international marketing, it is important that research regarding the various components of the international advertising mix continue. A pitfall that many international advertising researchers seemingly fall prey to is the temptation to often look exclusively for the differences between cultures. However, knowledge of cultural similarities is quite significant as well. Continued research regarding the similarities and differences of cultures throughout the world is advisable. Research on inter-market segmentation (i.e. segmenting similar consumers across national borders) is badly needed. Empirical evidence on the extent to which practitioners have had success in developing marketing to cross-national segments is needed to provide additional insight. Frameworks that help explain the circumstances in which various intermarket bases of segmentation may be effective need to be put forward. It is also clear that more research on control issues as they influence standardization is required. In terms of control, research on the client-agency relationship in a multinational context may provide insight regarding whether client strategies are maximizing global brand equity. A comparative study of advertising effectiveness among clients that demand control over a standardized strategy versus clients that allow agencies more latitude could be enormously


63

insightful. However, such a study would need to address another key issue how to measure the effectiveness of standardized campaigns in different markets. An even loftier, though important, goal would be to develop a method of measuring an advertising campaign's contribution to cross-national brand equity, in addition to the impact in individual markets, Experimental research on cultural factors and their impact on advertising effectiveness in various markets is needed. As noted by Taylor, Miracle and Wilson (1997), content analyses have provided the basis for some hypotheses, but to really understand why individual cultural factors affect advertising effectiveness, experimental studies are required. Additionally, testing the importance of various executional factors in various markets could provide very valuable insights for advertisers. For example, studies following an approach similar to Stewart and Furse's (1986) study of U.S. television commercials in other markets could do much to develop our knowledge regarding the extent that standardization of executional variables is desirable. International advertising research regarding the standardization-specialization debate does not need to be akin to a political debate with both sides arguing at the other's expense. Standardization and specialization research in the next century should focus more on the applicability issues than justifications. Practical applications regarding how international marketers can and should make tough decisions, and ultimately implement them, is needed as opposed to the research which focuses on which strategy is always the best to utilize is needed to further knowledge.

REFERENCES Aaker, J. L., & Williams, P. (1998). Empathy vs. pride: the influence of emotional appeals across cultures. Journal of Consumer Research, 25(3), 241-261. Agrawal, M. (1995). Review of a 40-year debate in international advertising. International Marketing Review, 12(1), 26--48. Albers-Miller, N. D., & Gelb, B. D. (1996). Business advertising appeals as a mirror of cultural dimensions: A study of eleven countries. Journal of Advertising, 25(4), 57-70. Alden, D. L., Hoyer W. D., & Lee, C. (1993). Identifying global and culture-specific dimensions of humor. Journal of Marketing, 57(2), 64-79. Alden, D., Steenkamp, J-B. E. M., & Batra, R. (1999). Brand positioning through advertising in Asia, North America, and Europe:The role of global consumerculture. Journal of Marketing, 63(1), 75-87. A1-Makaty, S. S., van Tubergen, G. N., Whitlow, S. S., & Boyd, D. A. (1996). Attitudes toward advertising in Islam. Journal of Advertising Research, 36(3), 16-26. Andrews, J. C., Durvasula, S., & Netemeyer, R. G. (1994). Testing the cross-national applicability of U.S. and Russian advertising belief and attitude measures. Journal of Advertising, 23(1), 71-83.

64

CHARLES R. T A Y L O R A N D CHAD M. JOHNSON

Bandyopadhyay, S., Robicheaux, R. A., & Hill, J. S. (1994). Cross-cultural differences in intrachannel communications: The United States and India. Journal of International Marketing, 2(3), 83-100. Biswas, A., Olsen, J. E., & Carlet, V. (1992). A comparison of print advertisements from the United States and France. Journal of Advertising, 21(4), 73-83. Buzzell, R. D. (1968). Can you standardize international marketing? Harvard Business Review, 46, 103-113. Caillat, Z., & Mueller, B. (1996). Observations: The influence of culture on American and British advertising: An exploratory comparison of beer advertising. Journal of Advertising Research, 36(3), 79-90. Cheng, H., & Schweitzer, J. C. (1996). Cultural values reflected in Chinese and U.S. television commercials. Journal of Advertising Research, 36(3), 27--44. Davison, J. A., & Grab, E. (1993). The contributions of advertising testing to the development of effective international advertising: The Kit Kat case. Marketing and Research Today, 21(1), 15-24. Domzal, T. J., & Kernan, J. B. (1993). Mirror, mirror: Some postmodern reflections on global advertising. Journal of Advertising, 22(4), 1-20. Duncan, T., & Ramaprasad, J. (1995). Standardized multinational advertising: The influencing factors. Journal of Advertising, 24(3), 55-67. Dunn, W. S. (1994). Early days of international advertising education in the U.S. Journal of Advertising, 23(1), 111-114. Durvasula, S., Andrews, J. C., Lysonski, S., & Netemeyer, R. G. (1993). Assessing the crosscultural applicability of consumer behavior. Journal of Consumer Research, •9(4), 626--636. Elinder, E. (1965). How international can European advertising be? Journal of Marketing, 29, 7-11. Englis, B. G., Solomon, M. R., & Olofsson, A. (1993). Consumption imagery in music television: A bi-cultural perspective. Journal of Advertising, 22(4), 21-32. Fatt, A. C. (1967). Danger of local international advertising. Journal of Marketing, 31(1), 60-62. Ford, J. B., Kramer Voli, P., Honeycutt Jr., E. D., & Casey, S. L. (1998). Gender role portrayals in Japanese advertising: A magazine content analysis. Journal of Advertising, 27(1), 113-124. Gould, S. J., Lerman, D. B., & Grein, A. F. (1999). Agency perceptions and practices on Global IMC. Journal of Advertising Research, 39(1), 7-20. Graham, J. L., Kamins, M. A., & Oetomo, D. S. (1993). Content analysis of German and Japanese advertising in print media from Indonesia, Spain, and the United States. Journal of Advertising, 22(2), 5-16. Griffin, T., McArthur, D., Yamaki, T., & Hidalgo, P. (1998). Ad agencies' performance and role in providing communication services in Chile, Japan, and the United States. Journal of Advertising Research, 38(5), 65-75. Harris, G. (1994). International advertising standardization: What do the multinationals actually standardize? Journal of International Marketing, 2(4), 13-30. Hill, J. S., & Shao, A. T. (1994). Agency participants in multicountry advertising: A preliminary examination of affiliate characteristics and environments. Journal oflnternational Marketing, 2(2), 29--48. Huff, L. C., & Alden, D. L. (1998). An investigation of consumer response to sales promotion in developing markets: A three-country analysis. Journal of Advertising Research, 38(3), 47-58. James, W. L., & Hill, J. S. (1991). International advertising messages: To adapt or not to adapt (that is the question). Journal of Advertising Research, 31(3), 65-71. Johansson, J. K. (1994). The sense of 'nonsense': Japanese TV advertising. Journal of Advertising, 23(1), 17-24.


65

Kanso, A. (1992). International advertising strategies: Global commitment to local vision. Journal of Advertising Research, 32(1), 10-15. KeiUor, B. D., Parker, R. S., & Schaefer, A. (1996). Influence on adolescent brand preferences in the United States and Mexico. Journal of Advertising Research, 36(3), 47-56. Kim, H. (1998). Ogilvy helming global Kodak effort. Adweek, 39(47), 5-6. Kitchen, P. J., & Schultz, D. E. (1999). A multi-country comparison of the drive for IMC. Journal of Advertising Research, 39(1), 21-38. Klein, J. G., Ettenson, R., & Morris, M. D. (1998). The animosity model of foreign product purchase: An empirical test in the People's Republic of China. Journal of Marketing, 62(1), 89-100. Lee, W-N., & Tse, D. K. (1994). Changing media consumption in a new home: Acculturation patterns among Hong Kong immigrants to Canada. Journal of Advertising, 23(1), 57-70. Levitt, T. (1983). The globalization of markets. Harvard Business Review, 61, 92-102. Maynard, M., & Taylor, C. R. (1999). Girlish Images in Japanese vs. U.S. Seventeen Magazine Ads. Journal of Advertising, 28, 38-49. McCullough, W. R. (1996). Global advertising which acts locally: The IBM subtitles campaign. Journal of Advertising Research, 36(3), 11-16. Meyers, R. D. (1996). Consistent advertising evaluation from multi-country inconsistent data sets. Journal of Advertising Research, 36(3), RC2. Money, R. B., Gilly, M. C., & Graham, J. L. (1998). Explorations of national culture and wordof-mouth referral behavior in the purchase of industrial services in the United States and Japan. Journal of Marketing, 62(4), 76-87. Moriarty, S. E., & Duncan, T. R. (1990). Global Advertising: Issues and Practices. CIRA, 313-341. Mueller, B. (1996). International Advertising: Communicating Across Cultures. Belmont: Wadsworth Publishing Company. Murray, N. M., & Murray, S. B. (1996). Music and lyrics in commercials: A cross-cultural comparison between commercials run in the Dominican Republic and in the United States. Journal of Advertising, 25(2), 51-62. Nevitt, T. (1992). Differences between American and British television advertising: Explanations and implications. Journal of Advertising, 21(4), 61-73. Parameswaran, R., & Pisharodi, R. M. (1994). Facets of country of origin image: An empirical assessment. Journal of Advertising, 23(1), 43-55. Reichal, J. (1989). How can marketing be successfully standardized in European market? European Journal of Marketing, 23(7), 60-67. Rice, B., & Bennett, R. (1998). The relationship between brand usage and advertising tracking measurements: International findings. Journal of Advertising Research, 38(3), 58-66. Ricks, D. A., Alpan, J. S., & Fu, M. Y. (1974). Pitfalls in advertising overseas. Journal of Advertising Research, 14(6), 47-51. Rosenthal, W. (1994). Standardized international advertising: A view from the agency side. Journal of International Consumer Marketing, 7(1), 39-62. Ross, C. (1996). Global rules are proposed for measuring TV. Advertising Age, 67(33), 3--4. Roth, M. S. (1992). Depth vs. breadth strategies for global brand image management. Journal of Advertising, 2•(2), 25-36. Roth, M. S. (1995). Effects of global market conditions on brand image customization and brand performance. Journal of Advertising, 24(4), 55-73. Samiee, S., & Roth, K. (1992). The influence of global marketing standardization on performance. Journal of Marketing, 56(2), 1-20. Sayre, S. (1994). Images of freedom and equality: A values analysis of Hungarian political commercials. Journal of Advertising, 23(1), 97-107.

66

CHARLES R. T A Y L O R A N D CHAD M. JOHNSON

Sriram, V., & Gopalakrishna, P. (1991). Can advertising be standardized among similar countries? A cluster-based analysis. International Journal of Advertising, 10(2), 137-149. Stewart, D. W., & Furse, D. H. (1986). Effective Television Advertising: A Study of 1000 Commercials. Lexington, MA: Lexington Books. Tal, S. H. C. (1997). Advertising in Asia: Localize or regionalize? International Journal of Advertising, 16(1), 48-61. Tansey, R., Hyman, M. R., & Zinkhan, G. M. (1990). Cultural themes in Brazilian and U.S. auto ads: A cross-cultural comparison. Journal of Advertising, 19(2), 30-39. Tansey, R., & Hyman, M. R. (1994). Dependency theory and the effects of advertising by foreignbased multinational corporations in Latin America. Journal of Advertising, 23(1), 27-43. Taylor, C. R., Wilson, R. D., & Miracle, G. E. (1994). The impact of brand differentiating messages on the effectiveness of Korean advertising. Journal oflnternational Marketing, 2(4), 31-52. Taylor, C. R., Miracle, G. E., & Wilson, R. D. (1997). The impact of information level on the effectiveness of U.S. and Korean television commercials. Journal of Advertising, 26(1), 1-18. Taylor, R. E., Grubbs Hoy, M., & Haley, E. (1996). How French advertising professionals develop creative strategy. Journal of Advertising, 25(1), 1-14. Tse, D. K., & Lee, W. (1993). Removing negative country images: Effects of decomposition, branding, and product experience. Journal of International Marketing, 1(4), 25-48. Wells, L. G. (1994). Western concepts, Russian perspectives: Meanings of advertising in the former Soviet Union. Journal of Advertising, 23(1), 83-97. Wilcox, J. S., Gentry, J. W., Stricklin, M., & Jun, S. (1996). Advertising Presentations of the Independent vs. Interdependent Self to Korean and U.S. College Students. Advances in International Marketing, 7, 159-174. Wiles, C. R., Wiles, J. A., & Tjernlund, A. (1996). The ideology of advertising: The United States and Sweden. Journal of Advertising Research, 36(3), 57-66. Zandpour, F., & Harich, K. R. (1996). Think and feel country clusters: A new approach to international advertising standardization. International Journal of Advertising, •5(4), 325-344.

THE ASSOCIATION BETWEEN PROCESS AND PROGRAM ADVERTISING STANDARDIZATION: AN ILLUSTRATION OF U.S. MULTINATIONALS OPERATING IN INDIA Aruna Chandra, David A. Griffith and John K. Ryans, Jr.

ABSTRACT For well over forty years, academics have debated the effectiveness of the standardization~adaptation of international advertising, with practitioners beginning four decades earlier. As the debate has progressed, a critical distinction in the literature has developed related to this issue, that of process and program standardization. In this study, we examine the association of process and program standardization, inclusive of consumer market and environmental similarity, within a sample of U.S. multinationals operating in India. Results indicate a positive association between process and program advertising standardization of U.S. multinationals operating in India. Implications for academics and practitioners are presented.

New Directions in International Advertising Research, Volume 12, pages 67-83. Copyright © 2002 by Elsevier Science Ltd. All rights of reproduction in any form reserved. ISBN: 0-7623-0950-4 67

68

ARUNA CHANDRA, DAVID A. GRIFFITH AND JOHN K. RYANS, JR. INTRODUCTION

For nearly eight decades, international advertising standardization has been the central focus of academics and practitioners (Agarwal, 1995). Most notably, in the last forty years, a tremendous growth in academic conjecture and research has focused on this topic (e.g. Agarwal, 1995; Baalbaki & Malhotra, 1993, 1995; Laroche et al., 2001; Levitt, 1983; Solberg, 2000). As the communication revolution, inclusive of the Internet, shrinks the world and minimizes the cost of market access to both consumers and firms, the discussion over the standardization of advertising has intensified (e.g. Culter et al., 2000; Hu & Griffith, 1997; Onkvisit & Shaw, 1999; Sirisagul, 2000). The present day dominance of the world media coupled with a shift in focus of multinationals from saturated Western markets to the Big Emerging Markets (BEMs), such as India and China, necessitates a re-examination of the issue of the standardization/adaptation of advertising. For example, recent efforts by Culter et al. (2000) and Srivastava and Schoenbachler (1999) have suggested that advertising in India and the BEMs may require unique adaptations, thus suggesting that the simple extension of prior research may not be applicable. Clearly, the tremendous growth opportunity in India, and the other newly developing market areas constitute major opportunities in the world economic order. While the BEMs are generally characterized by relatively low per capita income, economic (and in many cases political) instability, and antiquated infrastructure, they often contain vast resources and largely untapped consumer potential. As such, BEMs have become a strategic area for market expansion, especially for multinational companies (MNCs). India, in particular, provides an excellent context for examining the use of advertising standardization within an emerging market for MNCs. While there are clear and obvious differences between India and the U.S., India offers some of the greatest opportunities of all BEMs as the consumer middle class grows in size and purchasing power. From a U.S. multinational perspective, India's liability of foreignness is mitigated by the country's numerous similarities to the U.S. market. For example, India has a welldeveloped legal system, a democratic political system, a mixed economy with well-developed capitalistic traditions (Chandrasekaran & Ryans, 1996), and a substantial middle class. It is in this context, i.e. the emerging market of India, that the advertising standardization issue for U.S. MNCs is considered. Underlying the advertising standardization debate is the issue of balancing the cost efficiencies of standardized approaches across markets with the benefits

The Association Between Process and Program Advertising Standardization

69

of locally responsive strategies tailored to particular markets (Baalbaki & Malhotra, 1993, 1995; Jain, 1989; Levitt, 1983; Ozomer & Prussia, 2000; Szymanski et al., 1993). Researchers contend that managers can standardize or adapt programs and/or processes (Griffith et al., 2000; Jain, 1989; Sorenson & Wiechmann, 1975). In general, program refers to the elements of the marketing mix, whereas process refers to the outline of activities that occur prior to, and during the development and implementation of the marketing program. While a significant amount of program-related research, i.e. specific tactics used, has been conducted (e.g. Baalbaki & Malhotra, 1993, 1995; Cavusgil & Zou, 1994; Jain, 1989; Kotabe & Omura, 1989; Samiee & Roth, 1992; Szymanski et al., 1993), little research has examined process-related issues, i.e. the tasks that aid in program development and implementation (Griffith et al., 2000; Shoham, 1995; Waiters, 1986). This is a critical omission within the literature as process standardization and its association with program standardization provides the foundation for achieving advertising standardization effectiveness. The purpose of this study is to empirically examine the association between process and program advertising standardization in India. In this study, the associations of process standardization, consumer market similarity and environmental similarity with program standardization of advertising are explored. First, a discussion of advertising standardization is presented followed by the development of hypotheses related to each key variable. A method section specifies the sample, the pre-testing, and the variable measurements. The hypotheses are tested using responses from a sample of U.S. multinational managers operating in India. A discussion of the results, their academic and managerial implications and directions for future research are then presented. STANDARDIZATION The issue of standardization has been the subject of intense academic debate for several decades (cf., Baalbaki & Malhotra, 1993; Onkvisit & Shaw, 1999; Solberg, 2000). The topic, first noted in the 1920s, garnered widespread disagreement, most notably in the divergent views of Goodyear's David Brown and Bausch and Lomb's Carl Propson in 1923. Whereas Brown (1923) viewed humanity as possessing common attributes, thus allowing for standardization, Propson (1923) argued that adaptation was necessary to appeal to divergent local markets. Subsequent examples in the early years established a pattern of disagreement on this fundamental issue, with some advertising agency and corporate advertising executives perceiving adaptation to be key to global success (Bari, 1979; Delaforce, 1964; Lindsey, 1964; Marcus, 1964; Sutton,

70

ARUNA CHANDRA, DAVID A. GRIFFITHAND JOHN K. RYANS, JR.

1974; Theophilopoulos, 1979; Vladimir, 1950), while others viewed standardization as appropriate (Barnes, 1968; Cornejo, 1958; Deschampsneufs, 1967; Elinder, 1964; Ettinger, 1969; Patterson, 1967; Peebles, 1967). Given the great importance of this debate to practitioners, academics began actively engaging in the study of the issue in the late 50s (Pratt, 1956, etc.) and 60s (Buzzell, 1969; Miracle, 1968; Ryans, 1969). From this time, research has continued to proliferate on the topic and has remained unabated (e.g. Agarwal, 1995; Donnelly & Ryans, 1969; Green et al., 1975; Harvey, 1993; James & Hill, 1991; Kanso, 1992; Kernan & Domzal, 1993; Onkvisit & Shaw, 1999; Shao et al., 1993; Solberg, 2000). In reality, a variety of external and internal factors impinge on the standardization decision, which involves an inherent trade-off between the economic benefits of leveraging a global brand identity via standardized strategies with the performance gains achieved when adapting to local market conditions and consumer preferences (Baalbaki & Malhotra, 1993, 1995; Bharadwaj & Varadarajan, 1993; Jain, 1989; Ozomer & Prussia, 2000; Szymanski et al., 1993). As such, the standardization/adaptation debate converges on the perception of, or movement toward consumer as well as environmental homogeneity/heterogeneity (Donnelly & Ryans, 1969; Hu & Griffith, 1997; Levitt, 1983; Shoham, 1995). Those researchers who view markets, as being homogeneous argue that the standardization of advertising is more "effective" as it allows the firm to capture cost efficiencies and thus increasing margins (Jain, 1989; Levitt, 1983; Peebles et al., 1978). Thus far, at least in theory, the standardization issue has been considered unidimensionally (exploring program standardization); in this study we contend that it is important to differentiate the types of standardization when exploring this issue. Standardization has two fundamental aspects, i.e. program standardization, such as the individual advertising tactics, employed within and across markets, and process standardization, inclusive of the development of a common method through which programs are implemented (Griffith et al., 2000; Jain, 1989; Sorenson & Wiechmann, 1975). While a significant amount of program-related research has been conducted (e.g. Baalbaki & Malhotra, 1993, 1995; Cavusgil & Zou, 1994; Jain, 1989; Kotabe & Omura, 1989; Samiee & Roth, 1992; Szymanski et al., 1993), little research has examined process-related issues (Griffith et al., 2000; Shoham, 1995). This is a critical limitation within the literature as Sorenson and Weichmann (1975) note that while the benefits of standardizing marketing programs may be situational or context-specific, MNCs can gain the greatest economies by standardizing the process through which they devise these programs.


71

Process Standardization

Process standardization, in the context of advertising, refers to the commonality in the method of advertising development and implementation, inclusive of the ad campaign development process, approach to media selection, pre-testing/ post-testing of an ad, etc. Through the standardization of a firm's advertising development and implementation processes, a firm can capitalize on economies of scale and scope. The theory underlying the drive toward process standardization in today's highly competitive business environment is best signified under the business axiom of "Think Global, Act Local." The axiom of "Think Global Act LocaL" suggests that firms can maximize the benefits of standardization and adaptation by developing consistent global processes across markets (i.e. think global), while simultaneously adapting programs to specific markets (i.e. act local). Through process standardization a firm is able to capitalize on the cost advantages of the development and implementation of worldwide best practices, while adapting its specific program elements to local markets, thus allowing it to increase its local responsiveness thereby allowing the firm to earn above normal rents. While a firm's processes may tend toward standardized best practices, the association between process and program standardization will vary by market context (under the noted assumption of 'Think Global, Act Local"). When firms operate in relatively similar markets a positive association exists between the two elements of standardization (i.e. a firm operating in the U.S. and Canada would tend to have both standardized processes and programs), while firms operating in dissimilar markets would tend to observe a negative association (i.e. a firm operating in the U.S. and Japan will tend to have a standardized process yet a localized program). Thus, an underlying assumption of this argument is that firms attempt to maximize standardization of processes regardless of market, and adapt programs when external conditions require it. As such, the relationship between process and program standardization is context specific. In this study, given the relative similarities between the U.S. and India markets stated earlier, it is theorized that: HI: The association between process standardization and program standardization of U.S. MNCs operating in India is positive. Consumer Market Similarity

The feasibility of advertising program standardization is reliant on the identification of homogeneous market segments across countries (BlackweU

72

ARUNA CHANDRA, DAVID A. GRIFFITH AND JOHN K. RYANS, JR.

et al., 1992). Market similarity exists when consumers in the finn's crossnational target market react similarly to marketing stimuli, thus sharing common behavioral response patterns and preference structures (Harvey, 1993; Hu & Griffith, 1997; Jain, 1989; Levitt, 1983; Ryans, 1969). Levitt (1983) identifies this phenomenon as "segment simultaneity," which refers to parallel segments across markets. This tenet would have validity, if a firm is able to identify and target a similar segment in India and the United States. The existence of simultaneous segments makes standardization a feasible strategy in this context. If, for example, a finn approaches two widely divergent target markets with the same product, differences in underlying consumer preference structures and response patterns would necessitate adaptation of the finn's advertising program to effectively match the needs of the market (Kanso, 1992; Kernan & Domzal, 1993). Yet, even if "a limited measure of homogeneity" (Onkvisit & Shaw, 1987) could be identified between the United States and India, the two countries could be targeted with similar, if not identical approaches. Hence, the greater the similarity of consumer markets targeted by a finn across countries, i.e. common behavioral response patterns and preference structures, the greater the economic and administrative benefits of advertising program standardization. As such, it is theorized that:

H2: The association between consumer market similarity and program standardization of U.S. MNCs operating in India is positive. Environmental Similarity Environmental similarity refers to the similarity of legal, political, economic, regulatory and marketing infrastructure a firm confronts in its international marketing efforts. Environmental elements often vary markedly from country to country and can dramatically influence a finn's ability to standardize its advertising programs (Donnelly & Ryans, 1969; Green et al., 1975; JaJn, 1989). Whereas similarities in environmental factors can facilitate standardization, differences across markets can substantially hinder standardization. For example, differences in laws governing advertising across countries, such as the use of mass communications, can significantly limit standardization, since adaptation is a precondition for advertising in the market (Boddewyn, 1982; Harvey, 1993). To illustrate, some countries tightly regulate advertising directed to children. Similarly, differences in political and economic systems may directly influence the firm's ability to standardize its message, since these inherent differences may result in differences in message interpretation. Further, marketing infrastructure differences, such as type of media available, will necessitate


73

adaptation of a finn's advertising program (Harvey, 1993; Jain, 1989). While no two countries are completely identical, the focal issue is whether the two countries are environmentally similar enough to support a standardized advertising program. Given the political and economic similarities between India and the U.S. mentioned earlier, it is hypothesized that: H3: The association between environmental similarity and program standardization of U.S. MNCs operating in India is positive.

METHODS Sample A systematic random selection method was used to draw 250 U.S. multinationals listed in the Hoover Directory of Fortune 750 finns and the Fortune 500 listing operating in India. Individuals identified in the source directories were contacted via mail and asked to have the person most involved with the advertising strategy in India complete the English language survey. Sixty-three surveys were returned yielding a 25.2% response rate. Partially completed surveys were discarded. Fiftyone usable surveys were retained for analysis (21.4% effective usable response rate). Respondents represented finns averaging $7.014 billion dollars in annual sales, employing on average 32,504 individuals and personally averaging 14.45 years of international experience. Early vs. late respondents were compared in relation to annual sales, number of employees, years of international experience, program standardization, process standardization, consumer market similarity and environmental similarity, using Armstrong and Overton's (1977) procedure. No significant differences (p < 0.05) were observed. Measures Measures were developed for this study based upon the conceptualizations brought forth in the literature in the context of advertising standardization. Program standardization was measured via a 7-item, 5-point Likert scale, assessing the degree of similarity of the finn's: (1) creative copy of an ad; (2) ad slogan; (3) brand; (4) setting of ad; (5) package appearance; (6) labeling; and (7) product image, in India and the U.S. (t~ = 0.82). Process standardization was assessed through a 4-item, 5-point, Likert scale, assessing the degree of similarity of the finn's: (1) ad campaign development process; (2) approach to media selection; (3) pre-testing of an ad; and (4) post-testing of an ad, in India and the U.S. (et = 0.81). Consumer market similarity was measured on an 8-item, 5-point, Likert scale regarding similarity in consumer segments in India and the U.S. (et = 0.82). Respondents were asked to rate India in comparison

74

ARUNA CHANDRA, DAVID A. GRIFFITH AND JOHN K. RYANS, JR.

to the U.S. on similarity of: (1) purchasing power and buying preferences; (2) business-to-business buying products; (3) idioms; (4) awareness of major brands; (5) consumer mobility; (6) response to basic appeals; (7) country-oforigin effects; and (8) market segments. Environmental similarity was operationalized via a 5-item, 5-point scale from "very dissimilar" to "very similar" (or = 0.75). Respondents were asked to rate India in comparison to the U.S. on similarity of: (1) political system; (2) free enterprise; (3) distribution system; (4) competitive environment; and (5) management style. ANALYSIS To initially test each hypothesis we examined the pair-wise correlations between the antecedent variables and advertising program standardization (see Table 1). As hypothesized, all antecedent variables are significantly related to program standardization. The correlations ranged from -0.316 to 0.493 and were each statistically significant at the 0.05 level. Both the statistical significance (p < 0.05) and the size (greater than 0.316) of the correlations, between the variables and advertising program standardization, suggests that each variable is both theoretically and managerially important to understanding advertising program standardization. Since the variables of process standardization and consumer market similarity are themselves interrelated we used multiple regression to partial out the individual association of each of the variables with advertising program standardization. This analysis (Table 2) provides a more rigorous Table 1.

Program Standardization Process Standardization Environmental

Similarity Consumer Market Similarity

Means, Standard Deviations and Correlations. Mean

Std Dev.

3.15

0.67

3.41

0.80

0.473*

2.59

0.63

-0.316"

3.36

0.61

0.493*

Note: *p Page and zero when Page -> Peli" Similarly, A D V g e is equal to Page minus Pc~i when Page > Peli and zero when Pdi > Page'

Possession and Effects of Power in Advertising Agency-Client Relationships

229

Sampling and Data Collection The sampling frame was the Korea Advertising Data Directory (1998). We first identified top 300 client firms in terms of their annual advertising expenditure. After deleting those client firms that are either non-for-profit organizations or refused to participate in the study, 100 client firms were selected through systematic random sampling. Through telephone calls to selected client firms, a key informant who is most knowledgeable about the advertising of the client firm's main product line was identified and contacted. The title of those key informants includes senior marketing director and senior marketing (or advertising) manager. Data collection began with a letter to the chosen key informants that introduced the research project. The letter contained a recommendation from Korea Broadcasting and Advertising Corporation (KOBACO). A few days later, we called the key informants and gained permission to deliver the questionnaire to the key informant's office in person. A follow-up call and another questionnaire was sent two weeks later to those that had not responded to the original questionnaire. For the purpose of delivering and collecting questionnaires from key informants, a group of college students were used. The final response rate after two mailings and multiple phone calls was 100% (100 out of 100 delivered). The questionnaire for client firms contained questions asking for the name of the advertising agency and major contact person in that agency. Those who were identified as the key contact persons by the client firms' key informants were asked to participate in the survey with the explanation on the purpose of the survey and recommendation by KOBACO. All advertising agencies agreed to participate in the survey. We went through the same process for data collection for advertising agencies except that the key informants in advertising agencies were asked to talk about the focal client firm throughout the questionnaire. In effect, we collected data from both sides of agency--client dyads: 100 client firms and corresponding 100 advertising agencies. Measure Validation The measurement items first were examined in terms of their item-to-total correlations. Since power and commitment are operationalized as formative scales, coefficient alpha and exploratory factor analysis were conducted only for the commitment scales. Coefficient alphas were 0.80 for agency commitment and 0.82 for client commitment. Exploratory factor analysis result also shows that agency commitment and client commitment are internally consistent, yet different from each other. Therefore, convergent and discriminant validities of

230

CHANGHO OH AND STEPHEN KEYSUK KIM

the commitment scales are established. Pearson correlation and exploratory factor analysis results are reported on Table 1 Since we could not conduct factor analysis for formative scales of power and commitment for measure validation purposes, the convergent validity of the power and communication scales was tested by calculating correlation between each scale and a convergence item. For the agency power scale, its correlation with a convergent item (i.e. the extent to which the agency contributed to marketing and advertising goal attainment of the client firm) was calculated. Correlation between agency power and this item was 0.79 ( p < 0.001). For the agency communication scale, its correlation with a convergent item (i.e. the agency's responsiveness to client requests) was calculated. The correlation was 0.39 (p < 0.001). These results support the convergent validity of the power and communication scales. The discriminant validity of the power and communication scales was checked by correlation matrix on Table 1. Correlations between power

Table 1.

Pearson Correlation and Exploratory Factor Analysis Results.

A. Descriptive Statistics and Pearson Correlation 1

2

1. Clientpower 2. Agency's commitment 3. Agency's communication 4. Agency power 5. Client's commitment 6. Client's communication

1.00 0.56 0.51 0.12 0.12 0.18"

1.00 0.43 0.11 0.17 0.11

Mean Standard deviation

4.51 1.13

5.16 1.00

3

4

5

6

1.00 -0.04 0.00 0.15

1.00 0.67 0.54

1.00 0.52

1.00

25.52 4.71

4.98 0.89

5.02 0.95

25.86 5.41

* Coefficients larger than 0.18 are significantat the 0.05 level. B. Exploratory Factor Analysis of Agency Commitment and Client Commitment

Items Client commitment 1 Client commitment2 Client commitment 3 Client commitment4 Agency commitment I Agency commitment2 Agency commitment 3 Agency commitment4

Factor 1

Factor 2

0.83 0.77 0.87 0.74 0.06 0.08 -0.01 0.13

0.15 0.12 0.08 0.04 0.83 0.84 0.78


231

and communication are 0.56 for agency data and 0.54 for client data, which are significantly different from unity. These results provide an evidence of discriminant validity. The results of key informant bias check are available upon request. ANALYSIS

AND RESULTS

A series of regressions was conducted with power as the independent variables and commitment and communication as dependent variables. We used the Generalized Least Square option of SPSS to take the correlation between commitment and communication into account in model estimation. The regression results are reported on Table 2.

Possession of Power and Commitment Analysis results in Table 2.A indicate that an agency's commitment to the relationship increases as power of that client firm increases (b = 0.50, p < 0.001) and a client firm's commitment to the relationship increases as power of that agency increases (b = 0.72, p < 0.001), in full support of H l a and Hlb. Analysis results on Table 2.B indicate that an advertising agency's commitment to the relationship increases as total power increases (b = 0.27, p < 0.001). Similarly, a client firm's commitment to the relationship increases as total power increases (b = 0.37, p < 0.001). Thus, H2a and H2b are fully supported. As was hypothesized, we found that an agency's commitment to the relationship increases as relative power of that client firm increases (b = 0.29, p < 0.05). Similarly, a client firm's commitment to the relationship increases as relative power of that agency increases (b = 0.31, p < 0.001). Therefore, H2c and H2d are fully supported. Analysis results on Table 2.C indicate that total commitment between agency and client increases as total power increases (b = 0.64, p < 0.001). Conversely, the asymmetry of commitment between agency and client decreased as total power increased (b = -0.04, n.s.), but it was not statistically significant. Therefore, H3a is supported and H3b is not supported by data. As for the effect of relative power of a firm on commitment, we found that total commitment does not change significantly as relative power of either client firm (b = -0.12, n.s.) or agency firm (b = 0.11, n.s.) increases, in full support of H3c and H3d. Although we hypothesized that the asymmetry of commitment between agency and client does not change as relative power of client (or agency) increases (H3e and H3f), the analysis results indicate that asymmetry of commitment increases as relative power of a client firm increases (b = 0.62, p < 0.001). Relative power of an agency also has a similar, albeit weaker, effect on

232


-

,y

,-?

o

"g.i

~

I e~

~'~

~'~

r-'-

[.-,

'~

~ .]

.iS

0

¢5 V

45 ¢5 V

L; ,.:.; o.

~

V

.i5

V

?z

Possession and Effects o f P o w e r in Advertising A g e n c y - C l i e n t Relationships

0

25

~

~5

zz

~z

e.

0

e-

e~ e~

t--

~5

V

~5

o e~

eq

V ~.

~5

L;

~Z

~5 V

25

kl

233

234


asymmetry of commitment (b = 0.17, p < 0.10). Therefore, H3e and H3f are not supported by data.

Possession of Power and Communication The analysis results on Table 2.A indicate that an advertising agency's communication increases as power of that client firm increases (b = 2.12, p < 0.001) and a client finn's communication increases as power of that agency increases (b = 3.20, p < 0.001). Therefore, H4a and H4b are fully supported. We found that both an agency's communication (b = 0.83, p < 0.01) and a client firm's communication ( b = 1.71, p < 0 . 0 0 1 ) increase as total power in the dyad increases. Therefore, H5a and H5b are fully supported. We also found that an advertising agency's communication increases as relative power of its client firm increases (b = 1.99, p < 0.01), in support of H5c. However, a client firm's communication with its advertising agency does not change significantly as relative power of its agency increases (b = -0.07, n.s.). Therefore, H5d is not supported. As hypothesized, the total communication increases as total power increases (b = 2.54, p < 0.001). However, the asymmetry of communication does not decrease as total power increases (b = 0.16, n.s.). Therefore, H6a is supported and H6b is not supported by data. In support of H6c and H6d, the total communication does not change significantly as relative power of the client finn (b =-1.84, n.s) or the advertising agency (b =-1.03, n.s.) increases. Although we hypothesized an insignificant change, the analysis results suggest that the asymmetry of communication increases significantly (b = 2.45, p < 0.001) as relative power of a client finn increases. In contrast, the asymmetry of communication does not change significantly (b = -0.15, n.s.) as relative power of the agency firm increases. Therefore, H6e is not supported and H6f is supported by data.

DISCUSSION Speculation on Unexpected Results The effect of asymmetry of power on asymmetry of commitment. Although we hypothesized that asymmetry of commitment does not change as relative power of either an advertising agency or a client firm increases (H3e and H3f), the analysis results indicate that the gap in the commitment increases as relative power of the client firm (b = 0.62, p < 0.001) or the agency (b = 0.17, p < 0.10)


235

increases. These results suggest that one firm's change in commitment due to power asymmetry is not fully offset by the other firm's change in commitment to the opposite direction. Specifically, a comparison of two regression coefficients indicates that the relative power of a client firm has stronger effect on the asymmetry of commitment than relative power of an agency firm does (t = 2.78, p < 0.01). Thus, it is the advertising agencies that need to be more attentive to the power gap issue.

The effect of asymmetry of power on asymmetry of communication. As hypothesized, we found that the asymmetry of communication does not change as relative power of the agency increases (b =-0.15, n.s.). Yet, the asymmetry of communication increased significantly as relative power of a client finn increases (b = 2.45, p < 0.001). It is quite clear that relative power of a client firm has a stronger effect on the asymmetry of communication than relative power of agency does (t = 3.08, p < 0.01). The analysis results on Table 2.B indicate that the major cause of this difference is a client finn's steep decrease in communication. That is, a client finn lowers communication with its agency significantly as its relative power increases (b =-3.83, p < 0.001), whereas a client finn's communication does not change much when relative power of its agency firm increases (b = -0.07, n.s.). These different responses to power gap suggest that agencies should pay more attention to the power asymmetry and also highlight the need to study both sides of a dyad. Implications of the Study Results Effects of power on commitment and communication at the individual firm level. The analysis results on Table 2.A clarify two major issues in research on power: theoretical and methodological. From a theoretical standpoint, the results suggest clearly that power of a finn is conducive to positive attitude (i.e. commitment) and behavior (i.e. communication) of both advertising agencies and client firms. Therefore, the allegation that power is a symptom of dysfunctional relationships (Young & Wilkinson, 1989) needs to be dropped. Instead, the construct of power should be treated as a strategic variable that drives long-term, close ties between firms. From a methodological standpoint, the results suggest that operationalizing and measuring power with a "role performance" approach (Frazier, 1983) explains a significant amount of variation in commitment and communication. Note that we did not include any covariate in the regression models, yet R-squares on Table 2.A ranges from 0.26 (agency total communication) to 0.46 (client commitment). These results contrast with the prior analysis results based

236


on "replaceability" measure of power that had R-squares ranging from 0.04 (conflict) to 0.08 (distributor commitment) (Kumar, Scheer & Steenkamp, 1995). These differences in explanatory power suggest strongly that the "role performance" approach of power does a better job and should be used. From a managerial standpoint, the results on Table 2.A suggest a clear divergence on the effect of one's own power and exchange partner finn's power on commitment and communication. For example, it is really the power of the client finn that motivates an agency to commit more to the relationship and engage in more communication with the client finn. In contrast, the power of an agency as is perceived by the client firm does not have any significant effect on either the agency's commitment (b = 0.04, n.s.) or communication (b = -0.54, n.s.). This divergence suggests the possibility of a perceptual gap between a client finn and the agency itself on the power of that agency, thereby highlighting the need to know perceptions and opinions of both parties. Effects o f p o w e r on commitment and communication at the cross level The

results in Table 2.B show a very consistent pattern of the differential effects of relative power on an individual firm's commitment and communication. That is, a finn's own relative power advantage lowers its commitment to the relationship and communication with the exchange partner finn, whereas a finn increases commitment and communication as the relative power of the partner firm increases. To the best of our knowledge, this is the first clear empirical evidence that firms adjust their commitment and communication in a discriminating fashion as the power condition in a dyad changes, although this effect has been alluded to in the literature (Kumar, Scheer & Steenkamp, 1998). From a managerial standpoint, we observe slightly different effects of the power gap on commitment and communication on Table 2.B. That is, one finn's commitment to the relationship is clearly driven by the direction of power advantage within a dyad. Thus, a more powerful film decreases its commitment, whereas a less powerful finn increases its commitment, to the relationship. In contrast, power gaps appear to have divergent effects on communication. That is, an agency increases communication with its client firm significantly (b = 1.99, p < 0.01), but a client finn decreases communication with its agency significantly ( b = - 3 . 8 3 , p < 0 . 0 0 1 ) , as relative power of that client firm increases. In contrast, neither an agency (b=--0.96, p < 0 . 1 0 ) nor a client finn (b =--0.07, n.s.) changes communication level significantly as relative power of an agency increases. This asymmetry between a client and an agency suggests that, at least in the context of South Korean advertising industry, advertising agencies are expected to initiate and maintain a certain level of communication regardless of its power advantage over its client finn.


237

Interestingly, both agencies and clients appear to be aware of this "unwritten rule" of the advertising industry. Effects of power on commitment and communication at the dyad level The results on Table 2.C indicate that our hypotheses on the effects of total power on total commitment and total communication are fully supported. That is, total commitment and total communication increase as total power increases, while asymmetry of commitment and communication do not change much as total power increases. Thus, the constructive nature of power for exchange parties' attitude and behavior is supported at the dyad level as well. We also found that total commitment and total communication do not change much as either relative power of a client firm or an advertising agency increases. This result clearly shows the "compensatory" nature of power advantages for commitment and communication at the dyad level. That is, the effect of a firm's power advantage is offset by the effect of the other firm's power disadvantage, thereby leaving the net effect of power gap at the dyad level stable. Relatedly, this result also corroborates a dyadic nature of interfirm relationships, reminding us the danger of examining only one side of a dyad in the study of interfirm relationships. From a managerial standpoint, the above results indicate that what matters more for commitment and communication at the dyad level is the magnitude of total power rather than the extent of power gap between an agency and its client firm. Therefore, "expanding the size of the pie" through collaboration (Jap, 1999) should be the focus of the client-agency relationship instead of worrying about the imbalance of power between the two firms (cf. Heide & John, 1988). Thus, from an agency's perspective, the best strategy is to reduce the power gap and to expand the total power at the same time. Another insight from Table 2.C is the divergent effects of power advantages on asymmetry of commitment and communication at the dyad level. That is, asymmetry in commitment and communication does not change much as relative power of agency increases (b=0.17, p < 0 . 1 0 and b = - 0 . 1 5 , n.s. for commitment and communication, respectively). In contrast, asymmetry in commitment and communication increase significantly (b = 0.62, p < 0.001 for commitment b = 2.45 p < 0.001 for communications) as relative power of a client firm increases. Therefore, it is the condition in which a client firm has more power that changes dyad members' attitude and behavior. This result suggests two implications. First, as a seller of advertising service, an agency may be expected to remain committed to the relationship and maintain communication with its client firm even if its power over a client firm increases. On the other hand, as a buyer of advertising service, a client firm has more leeway in adjusting commitment to its agency and communication with that

238


agency as power condition changes. Second, this divergence also suggests that it is the advertising agencies that should be more vigilant about imbalance in power conditions and make efforts to expand the magnitude of total power through client support programs and superior role performance.

Limitations of the Study and Further Research Directions The paper has two potential limitations. From a conceptual standpoint, our measurement of power did not consider the "replaceability" dimension of dependence (Emerson, 1962). Although the analysis results indicate that "role performance" measure of power does an excellent job of explaining the variations in commitment and communication, the study is still vulnerable to a criticism that a full domain of power has not been tapped and measured. From a methodological standpoint, using an absolute difference scores of asymmetry in commitment and communication and spline measure of power asymmetry as was suggested by Kumar, Scheer and Steenkamp (1995, 1998) carries the potential problem of missing the feasible nonlinear functional relationships between power and outcome variables (Edwards, 1994). Further research on client-agency relationships can be expanded in two ways. First, the link between power and other outcome variables such as control and economic performance should be examined. The relationship between power and control in a dyad remains elusive despite its importance for managerial purposes. Similarly, the link between power and economic outcome variables should be clarified (Buchanan, 1992). If indeed total power is conducive to positive attitude and behavior of dyad members and the gap in power does not change asymmetry of attitude and behavior of dyad members, total power should also lead to a better economic performance of a dyad. Yet, this speculation has not been tested empirically. Second, the relationship between power at the interpersonal level and power at the interfirm level deserves more research attention. The importance of interpersonal issues appears quite high in advertising agency-client firm relationships in which power of an account executive can make a significant difference. Power at the interpersonal level should be examined further and related to commitment and communication at the interpersonal level. ACKNOWLEDGMENTS The authors thank anonymous reviewers for their helpful comments and C. Ray Taylor for his encouragement and support. Financial support by Korea Broadcasting and Advertising Corporation for this study is gratefully acknowledged.


239

REFERENCES Anderson, E., & Weitz, B. A. (1992). The Use of Hedges to Build and Sustain Commitment in Distribution Channels. Journal of Marketing Research, 29, 18-34. Anderson, J. C,, & Gerbing, D. W. (1991). Predicting the Performance of Measures in a Confirmatory Factor Analysis With a Pretest Assessment of Their Substantive Validities. Journal of Applied Psychology, 76, 732-740. Anderson, J. C., & Narus, J. A. (1990). A Model of Distributor Firm and Manufacturer Firm Working Partnerships. Journal of Marketing, 48, 62-74. Boyle, B., Dwyer, F. R., Robicheaux, R. A., & Simpson, J. T. (1992). Influence Strategies in Marketing Channels: Measures and Use in Different Relationship Structures. Journal of Marketing Research, 29, 462~,73. Buchanan, L. (1992). Vertical Trade Relationships: The Role of Dependence and Symmetry in Attaining Organizational Goals. Journal of Marketing Research, 29, 65-75. Edwards, J. R. (1994). The Study of Congruence in Organizational Behavior Research: Critique and a Proposed Alternative. Organizational Behavior and Human Decision Processes, 58, 51-100. Emerson, R. M. (1962). Power-Dependence Relations. American Sociological Review, 27, 31-41. Frazier, G. L. (1999). Organizing and Managing Channels of Distribution. Journal of the Academy of Marketing Science, 27, 226-240. Frazier, G. L. (1983). On the Measurement of Interfirm Power in Channels of Distribution. Journal of Marketing Research, 20, 158-166. Frazier, G. L., Spekman, R. E., & O'Neal, C. R. (1988). Just-In-Time Exchange Relationships in Industrial Markets. Journal of Marketing, 52, 52-67. Geyskens, I., Steenkamp J-B. E. M., & Kumar, N. (1999). A Meta-Analysis of Satisfaction in Marketing Channel Relationships. Journal of Marketing Research, 36, 223-238. Gundlach, G. T., & Cadotte, E. R. (1994). Exchange Interdependence and Interfirm Interaction: Research in a Simulated Channel Setting. Journal of Marketing Research, 29, 516-532. Heide, J. B., & John, G. (1988). The Role of Dependence Balancing in Safeguarding TransactionSpecific Assets in Conventional Channels. Journal of Marketing, 52, 20-35. Jancic, Z., & Zabkar, V. (1998). Establishing Marketing Relationships in the Advertising Agency Business: A Transitional Economy Case. Journal of Advertising Research, 36, 27-36. Jap, S. (1999). Pie-Expansion Efforts: Collaborative Process in Buyer-Seller Relationships. Journal of Marketing Research, 36, 461-475. Kim, K., & Frazier, G. L. (1997). On Distributor Commitment in Industrial Channels of Distribution: A Multicomponent Approach. Psychology and Marketing, 14, 847-877. Kumar, N., Scheer, L. K., & Steenkamp, J-B. E. M. (1995). The Effects of Perceived Interdependence on Dealer Attitudes. Journal of Marketing Research, 32, 348-356. Kumar, N., Scheer, L. K., & Steenkamp, J-B. E. M. (1998). Interdependence, Punitive Capability, and the Reciprocation of Punitive Actions in Channel Relationships. Journal of Marketing Research, 35, 225-235. Lusch, R. F., & Brown, J. R. (1996). Interdependency, Contracting, and Relational Behavior in Marketing Channels. Journal of Marketing, 60, 19-38. Michell, P. C. N., & Sanders, N. H. (1995). Loyalty in Agency~Client Relationships: The Impact of the Organizational Context. Journal of Advertising Research, 35, 9-23. Mohr, J., Fisher, R. J., & Nevin, J. R. (1996). Collaborative Communication in Interfirm Relationships: Moderating Effects of Integration and Control. Journal of Marketing, 60, 103-115.

240

C H A N G H O OH AND STEPHEN KEYSUK KIM

Mohr, J., & Nevin, J. R. (1990). Communication Strategies in Marketing Channels: A Theoretical Perspective. Journal of Marketing, 54, 36-51. Morgan, R. M., & Hunt, S. D. (1994). The Commitment-Trust Theory of Relationship Marketing. Journal of Marketing, 58, 20-38. Oh, C. (1998). Patterns of advertising agency-client Relationships in South Korea. Seoul, South Korea: Korea Advertising Association. Prendergast, G., Shi, Y., & West, D. (2001). Organizational Buying and advertising agency-client Relationships in China. Journal of Advertising, 30, 61-71. Rosenbloom, B. (1999), Marketing Channels: A Management View (6th ed.). Fort Worth, TX: The Dryden Press. Rust, R., Lemon, K. N., & Zeithaml, V. A. (2001). Driving Customer Equity: Linking Customer Lifetime Value to Strategic Marketing Decisions. MS1 Working Paper No. 01-108. Segal, M. N. (1989). Implications of Single vs. Multiple Buying Sources. Industrial Marketing Management, 18, 163-178. Wilson, D. T., & Mummalaneni, V. (1988). Modeling and Measuring Buyer-Seller Relationships. ISBM Working Paper No. 3-1988. Young, L. C., & Wilkinson, I. F. (1989). The Role of Trust and Cooperation in Marketing Channels. European Journal of Marketing, 23, 109-122.


241

APPENDIX (Questionnaire for Client Firms) I. Power How well does your advertising agency perform on the following tasks? (1: very dissatisfied 7: very satisfied) 1. Role content (a) understanding on market and consumers (c) suggestion of advertising concept (e) media planning (g) event and special promotion

(b) suggestion of good marketing ideas or tactics (d) creativity (f) media management and execution (h) market research and ad effectiveness research

2. Role process (a) understanding for our decision-making process (c) easy coordination in execution

(b) easiness to reach agreement in decision making (d) responsiveness to client requests

3. Relationship (a) competence of the agency staff

(b) personal relationship between AE and client advertising staff

(c) stability of the agency personnel II. Client Commitment How much do you agree with the following statements (1: Strongly disagree - 7: Strongly agree) 1. 2. 3. 4.

We treat the relationship with this agency as a long-term, cooperative relationship. We have been investing much time and effort to the relationship with this agency. This agency and our firm have strong business ties. We are willing to switch to a new agency if that agency offers better terms ®.

III. Communication How often does your staff meet with the people from the advertising agency to discuss advertising issues? (1: no interaction at all - 7: very frequent interaction) 1. Between top management (a) office meetings and visits (c) exchange of documents

(b) telephone calls (d) informal meetings (dinner, sports, etc.)

2. Between senior managers (a) office meetings and visits (c) exchange of documents


3. Between advertising staff (a) office meetings and visits (c) exchange of documents


SELF-ESTEEM, LIFE-SATISFACTION AND MATERIALISM: EFFECTS OF ADVERTISING IMAGES ON CHINESE COLLEGE STUDENTS Shuhua Zhou, Fei Xue and Peiqin Zhou

ABSTRACT This paper investigated the effects of exposure to advertising images on quality of life issues. Affordable products and unaffordable products, as well as control images, were presented to participants of the experiment. Subjects' self-esteem, life-satisfaction and materialism were measured using a battery of scales. Results indicated affordable product images did not affect any dependent variables. However, exposure to unaffordable products enhanced, rather than decreased, Chinese students' self-esteem and life-satisfaction, contrary to results found in previous social comparison research. This suggested that Chinese students were rather optimistic about their future and they used future orientation in their prediction of success. Implications of the study were discussed.

New Directions in International Advertising Research, Volume 12, pages 243-261. Copyright © 2002 by Elsevier Science Ltd. All rights of reproduction in any form reserved. ISBN: 0-7623-0950-4 243

244

SHUHUA ZHOU, FEI XUE AND PEIQIN ZHOU

INTRODUCTION Advertising conveys cultural symbols by means of well-crafted images with inherent values. People are exposed to countless advertising images every day in the form of billboards, magazine images, and television commercials. Critics are concerned that advertising is directed towards those with money, marginalizing those who are poor and low in social and economic status (Englis, 1992). Because the purpose of an advertisement is to arouse interest and motivate desire, many wonder about the effects of advertising on viewers' perception of the self and life in general. As advertising is often about the glamorous, the better and the more, the images often present an idealized version of life. These idealized images work through one's psyche in two mechanisms. First, it leads to self-comparison with those images, in which the consumer may feel less satisfied with his or her current circumstance and strive to achieve the idealized state. Second, the self-comparison can result in an upward shift of consumers' expectations for the standard of living that they believe they should achieve (Richins, 1995). As a consequence of such a comparison, people may feel frustrated if they do not obtain the "better life" presented in advertising. Almost without exception, the idealized images presented in advertising focus on the material aspect of life. Many researchers argue that, because the emphasis on material possessions has both a functional value and a symbolic one, longterm exposure to such images will lead to higher materialism. Ultimately, this will drive people to evaluate the quality of general life by the quality of material life, which typically results in a negative evaluation of life satisfaction (e.g. Sirgy et al., 1998; Zinkhan, 1994; Sirgy, 1998). However, most research on quality of life issues treats exposure to advertisement as a global construct. Little empirical research has been done focusing on the nature of advertised products on viewers' perception of their self worth. It is important to understand what enhances and what interferes with a person's evaluation of the self and his or her life in relation to the features and characteristics of the advertised products. This study used social comparison theory to test the effects of exposure to two types of advertised products, either affordable or unaffordable for the participants at the time of the study. Three dependent variables were examined regarding exposure to such images: perception of life satisfaction, self-esteem and sense of materialism. This study therefore addressed an important area that needed more investigation: the impact of advertising on issues dear to the evaluation of a happy life in a society saturated with media images. By examining the effects of different images, this study further explored how

Self-Esteem, Life-Satisfaction and Materialism

245

advertising provided or activated constructs in viewers' minds and how people used these as criteria reference in their subsequent judgment and interpretation of concerns about quality of life. THE

CASE

COUNTRY

We chose a fast developing country, the People's Republic of China, as our target of investigation. The variables involved in this study, self-esteem, life satisfaction and sense of materialism, are particularly important in the case of an evolving society where materialistic consumption is an emerging concept. Arguably, effects of advertising images would manifest themselves more readily with the benefit of a sharp contrast between traditional Confucian values and the newly found consumer freedom. Thanks to the policy of openness and reform enacted in 1978, China has become one of the epitomes of economic success. In 1992, the country furthered a vigorous market-oriented reform, which became a major impetus responsible for the emergence of a consumer market and consumer culture. The mass media, especially advertising, reinforced this process (Wei & Pan, 1999). Under these circumstances, materialistic possession began to gain momentum. Most families in China now have electronics, such as television sets, washing machines and refrigerators. These items were considered unaffordable luxuries twenty years ago. A survey conducted in Shanghai indicated that in 2000 there were 147 television sets, 102 refrigerators, and ninety-three washing machines per hundred families. The counterpart numbers in 1985 were twenty-two, twenty, and twenty-six. 1 On the other hand, advertising, which was accused of reinforcing capitalism and hedonism and was absent from the communist media landscape since the Cultural Revolution in the 1960s, came back to China in 1978 and helped spur the growth of a market-oriented economy. Tremendous strides have been made since then. It was reported that from 1987 to 1996 advertising expenditures in China increased by almost 900%. 2 Wang (2000) surveyed the advertising environment in Beijing on a day in 1996 and reported a total of 386 ads encountered. Understandably, most of these commercials emphasized materialistic consumption, which was incompatible with traditional Confucian and Communist values. While life satisfaction used to depend on how favorable one was viewed by the Party and little on material possession, it would be interesting to see the impact of these newly gleaned concepts on viewers' sense of materialism, self-esteem and ultimately, life satisfaction.

248


is a function of the social environment (Mruk, 1999). Self-esteem is the result of comparison between the ideal self and the current, actual, or "real" self. The smaller the gap is between the ideal and the actual, the higher the self-esteem. Conversely, the greater the gap, the lower the self-esteem. Advertisements can very well serve as the stimuli that remind and highlight the gap between the ideal and the reality. The study of Gulas and McKeage (2000) has testified that advertisements can change viewers' self-esteem. Generally speaking, after comparison with the idealized model in advertisements, viewers have lower scores on self-esteem. It is reasonable to predict that lacking the financial ability to buy the advertised products will make people feel worse about themselves. This discussion suggests the following hypotheses: H2a: Exposure to ads of currently affordable products and services will not influence viewers' self-esteem. H2b: Exposure to ads of currently unaffordable products and services will lower viewers' self-esteem.

It has been consistently reported that self-esteem is positively related to life satisfaction. Lewinsolm et al. (1991) stressed that self-esteem is one of the determinants of life satisfaction in their conceptual model of life-job satisfaction. Sekaran (1986) suggested self-esteem is the best predictor of life satisfaction because self-esteem helps surmount negative experiences, which in turn helps the individual to experience the good things in life. Moreover, those who view themselves in a positive way tend to observe life events more positively and may therefore show higher life satisfaction. Thus we expected a positive relationship between self-esteem and life satisfaction to emerge in the present study. H3: Self-esteem is positively related to life satisfaction. Materialism

Materialism denotes the importance one attaches to their worldly possessions (Belk, 1985). According to Belk (1984), the highest levels of materialism are assumed when such possessions take up a central place in one's life and become the greatest source of satisfaction and dissatisfaction in life. For a materialist, possessions are central to his or her life in that he or she feels that increased consumption increases his or her satisfaction with life. Materialism may become a problem in situations where the physical goal of consumption overshadows all other goals of self and interactive development (Belk, 1985; Richins & Dawson, 1992).


249

Materialism encourages customers to buy and consume. It leads to the idea that more is better. A desire for more means increased sales, so advertisers often incorporate materialistic appeals in their messages. Such appeals facilitate the attainment of corporate objectives (in terms of market share or profitability) because if consumers lose focus on material acquisitions and place an emphasis on non-material concerns, then they will consume less from the economic system, leaving marketers with fewer consumers and less demand for their goods (Zinkhan, 1994). Materialism is observed to influence people's perception of life satisfaction. Richins and Dawson (1992) conceptualized materialism as a consumer value with three components: success (possession-defined success), centrality (acquisition centrality), and happiness (acquisition as the pursuit of happiness). Possession-defined success suggests that materialists tend to judge their own and others' success by the number and quality of possessions accumulated; acquisition centrality suggests that materialists make possessions the focus of their lives; acquisition as the pursuit of happiness factor suggests that materialists regard possessions as essential to their satisfaction and well-being in life. Thus the value of possessions for materialists lies in their ability to both confer status and to project a desired self image. Materialism is thus a dispositional trait or personality that negatively influences life satisfaction. The reason is that materialism involves a tendency to experience negative emotions. Belk (1985) suggested that materialistic people are usually possessive, nongenerous, and envious. Several studies have examined the relationship between materialism and happiness or life satisfaction (e.g. Belk, 1984; Richins & Dawson, 1992). This literature review leads to the following hypothesis: H4: Materialism is negatively related to life satisfaction.

EXPERIMENTAL STIMULI We selected twenty-four print advertisements from mainstream Chinese magazines and divided them into three groups. Eight of them were used for the control group, with the ads featuring PSA and health issues. Eight of them featured affordable products such as orange juice and shampoo. Another eight were ads of products and services incompatible with subjects' current financial status, such as automobiles and overseas trips. Experimenters selected the ads if they were of a full page and the product was the central visual element. Manipulation checks indicated that categorization were adequate, as data

250


indicated that the majority (90.6%) of students' monthly allowance fell below 1000 yuan (approximately $125) and 97.3% of their annual family income was lower than 100, 000 yuan (approximately $12,500). An import family sedan, such as the Honda Accord, sold for more than 300,000 yuan (approximately $37,500) in the Chinese market, hardly within reach for most families. PARTICIPANTS

AND PROCEDURE

The participants, seventy-five junior college students at a large Southeastern Chinese university, were randomly divided into three groups and exposed to different print advertisements we selected: advertisements for public good, advertisements for currently affordable products (such as shampoo, orange juice, etc.), and advertisements for currently unaffordable products and services (automobiles and overseas trips). There were thirty-five males and forty females. Each group contained relatively equal numbers of male students and female students to minimize the impact of gender differences. Participants were asked to view eight advertisements in each group and answer two questions unrelated to our experiment to make sure they paid attention to the advertisements. College students were chosen for this experiment for several considerations. In one of the manipulated conditions, subjects were exposed to luxury product images. If anyone would be able to afford them in China, college students would be the most likely consumers of luxury products in the future. In an effort to sustain its economic boom, the Chinese government has been putting technology initiatives among their highest priorities, and they have rewarded thousands of scientists with generous compensation for their innovations. The intention is to instill the idea that knowledge equates to power and money. Therefore, if advertisers of luxury products want to secure the Chinese market in the long run, college students should reasonably be the target audience. On the other hand, one could reasonably argue that if effects on quality of life issues were detected in college students after exposure to unaffordable products, there would be good reason to suspect that a more dramatic effect existed for the general population. College students were assumingly more rational human beings with promises of future affordability. With dimmer indications in either affordability or rationality, the adverse effect of exposure to luxury items presumably would be greater for the general public as viewers might have difficulty discerning advertised reality and the real world. However, student participants also limited our generalization scope. College students, after all, were special in many ways, especially in the Chinese context. Therefore, any findings should be interpreted with caution.


251

MEASURES After viewing the advertisements in the given condition, participants were required to answer fifty questions including the test scales and several filter questions. A combination of measures was adopted in this study. The three-item life satisfaction measurement scale was developed from the congruity lifesatisfaction scales by Sirgy et al. (1998), asking how satisfied participants were compared to either what they had ideally hoped to become, or to what their parents wanted them to be, or to the accomplishments of their friends. The reported alpha for these measures ranged from 0.744 to 0.934, depending on the country in which they were used. China registered the lowest coefficient alpha, a fact substantiated again in this study as we recorded a coefficient alpha of 0.73. This might have to do with the adoptability of the scales in China. These scales all seemed to measure individual achievements as an indicator of life-satisfaction. Individual achievements were downplayed by the communist regime which advocated collective achievements for decades. It could also be that some nuances were lost when the original scales were rendered into the Chinese language. Richins and Dawson's (1992) eighteen-item scale was used to measure materialism (see Appendix 1), with a coefficient alpha of 0.68, while the statistics reported in the original article stood at 0.80. Again, this was likely caused both by a longtime sneering campaign by communist ideologists and translation problems. On the other hand, Rosenberg's (1965) ten-item scale was used to measure self-esteem (see Appendix 2), with the coefficient alpha tested to be 0.74, compared to the original statistics reporting at 0.72. All responses were recorded on five-point Likert scales ranging from "strongly disagree" (1) to "strongly agree" (5). The reversed scales were recoded after data were collected. RESULTS

Life satisfaction The first hypothesis stated that exposure to ads of currently affordable product would not influence viewers' life-satisfaction. Results supported this hypothesis. There was no significant difference between the life satisfaction of the control group (M = 2.56, SD = 0.70) and that of the affordable group (M = 2.72, SD = 0.53), t ( 4 8 ) = - 0 . 9 1 , p > 0 . 1 0 . However, the hypothesis that exposure to currently unaffordable product would lower viewers' life-satisfaction was not

252


supported. On the contrary, the mean satisfaction index (M = 2.93, SD = 0.70) was significantly higher than that of the control group (M = 2.56, SD = 0.70), t (48) = - 1.89, p < 0.05 (one-tailed test). This result represented a stark contrast to our hypothesis. That is, exposure to currently unaffordable product advertisements significantly improved Chinese college students' life-satisfaction.

Self-Esteem The hypothesis that exposure to currently affordable product would not influence viewers' self-esteem was supported. There is no significant difference between the self-esteem of the control group (M = 3.20, SD = 0.45) and that of the affordable group (M = 3.10, SD = 0.39), t (48) = 0.91, p > 0.10 (two-tailed test). However, results did not support the hypothesis that exposure to ads of currently unaffordable products would lower viewers' self-esteem. On the contrary, and similar to our findings on life-satisfaction, exposure to ads of currently unaffordable product significantly improved Chinese college students' self-esteem (M = 3.52, SD = 0.48), t (48) = -2.39, p < 0.05 (one-tailed test).

Self-Esteem and Life-Satisfaction In conformity with results of previous research, a positive correlation between self-esteem and life-satisfaction was observed (r = 0.57, p < 0.00), supporting Hypothesis 3 which stated that higher self-esteem corresponded to higher lifesatisfaction and vice versa.

Materialism and Life-Satisfaction Results indicated that there was no correlation between materialism and lifesatisfaction (r = 0.02, p > 0.10) in the case of Chinese students, failing to support Hypothesis 4 which predicted that materialism had a negative correlation with life satisfaction. DISCUSSION Research in social comparison suggests that media viewers are prone to use images in advertisements to make upward comparisons, resulting in inferior evaluation of their self worth and the social environment because the images in advertising are often superior to those in real life (e.g. Richins, 1991; Gulas & McKeage, 2000). In this study, advertising images were not treated as a global construct. Rather, we manipulated them into two disparate categories:


253

affordable and unaffordable items. It was expected that affordable products, such as shampoo, orange juice and candy, were too trivial to have an impact on people's life-satisfaction, self-esteem and sense of materialism. However, viewing unaffordable products, or automobiles and pleasure trips overseas, was expected to activate upward comparison in Chinese college students. Consequentially, participants' evaluation of their current situation would be lowered. As predicted, exposure to affordable products did not have any significant impact on either life-satisfaction or self-esteem, supporting Hypotheses la and 2a. These lent credence to our contention that treating all images in advertisements as a global construct might not be a precise representation and that careful categorization might offer us more insight into how different images might bear different impact. On the other hand, self-esteem and life satisfaction were found to be positively correlated, supporting Hypothesis 3 and confirming previous findings in the literature. However, statistical tests failed to support Hypothesis 4, which stated that materialism was negatively correlated with life-satisfaction. Results indicated that neither affordable products nor unaffordable products had any effect on viewers' sense of materialism. Several reasons could be responsible for this. It could be that after years of communist indoctrination for the lofty goals of utopian equality, worldly possession was still shunned if not despised. On the other hand, non-material quality of life concerns have been a central part of Confucian teaching over the years, in that happiness was defined more in terms of one's knowledge acquisition than materialistic consumption. Of course, the possibility also existed that our participants were not representative of the general population. Barring such a scenario, the fact that there were no effects on Chinese students' desire to buy and consume after exposure to either affordable or luxury items does not bode well particularly for a developing market economy in China. As Wei and Pan (1999) argued, Chinese are adopting an official discourse on establishing a market economy to legitimize their fever for consumption. Results of this study were not especially encouraging in this regard. The more interesting findings, however, came from exposure to unaffordable products, in which results ran opposite to the predicted direction. Exposure to ads of currently unaffordable products enhanced Chinese college students' perception of self-esteem and life-satisfaction. Most studies on social effects of advertisement focus on populations in developed countries, especially in the U.S. where the general population tends to pessimistically consider their failure of having the good life presented in ads instead of being skeptical at the believability of the life presented in advertising images (Williams, 1981). In this study, we might be dealing with a student

254


population who is very confident and optimistic about their future. The evidence of which came from a question in which we asked participants to predict when they would be able to afford a car and a pleasure trip overseas. Table 1 revealed that they were hopeful of their financial future. To these students, although the advertised luxury items were temporarily beyond their reach, their predicted affordability was not far from the horizon. 82.7% students agreed that they would be able to afford an overseas trip in ten years, while 54.7% predicted car ownership within 10 years. It is possible that these luxury items were viewed as potentially affordable to them, and they represented a higher quality of life in the future, thus enhancing participants' evaluations. There was also a negative relationship between predicted affordability of an automobile and self-esteem (r = -0.37, p < 0.001), indicating that the sooner the students predicted they would be able to afford a car, the higher self-esteem they had. The negative relationship also existed between the predicted time to afford an overseas trip and self-esteem (r = -0.27, p < 0.05). Several factors are accountable for this optimism in Chinese college students. First, China's continuous economic development over the last two decades has caused dramatic improvement in life quality. From 1978 to 1997, China's Gross Domestic Product (GDP) growth rate surpassed 10% every year, higher than any other country's for the same period. The poor population declined from 33% in 1978 to 4% in 1997 (Li, 1998). A survey done in twenty-two major cities in China in 2000 revealed that 41.6% of subjects agreed that their quality of life had been improved in 2000, and 40.4% predicted that the quality of life would be further improved in 2001. In other words, optimism continues to run high in regards to life improvement. It is also no secret that Chinese college students can reasonably dream of a promising future. In 1997, only four out of every one hundred college-age youths were able to enroll in either a college or a university (Cai & Tian, 2000). Table 1.

Prediction of Affordability of Overseas Trip and Car.

Prediction of affordability Within this year Within 1-2 years Within 3-5 years Within 5-10 years After 10 years Never have the chance

Overseas trip Number Percentage 3 8 22 29 12 1

4.0 10.7 29.3 38.7 16.0 1.3

Buying a car Number Percentage 0 0 6 35 33 1

0 0 8.0 46.7 44.0 1.3

255


Completion of a higher education inevitably improves the graduates' chances of becoming China's elites in the social and economic hierarchy. From this perspective, they have every reason to be buoyant about higher self-esteem and life satisfaction. If we were to understand self-esteem from a formula proposed by James (1983) more than a century ago, we would understand how optimism in Chinese students would readily translate into self-esteem. The formula states: Self-Esteem -

Su ccess Pretentions

In this experiment, the denominator, or pretensions, was equally distributed in the control group and in the experimental groups. Therefore, the only variable that could account for an increase in self-esteem would be an increase of the numerator of "success." Chinese student participants might be experiencing "future orientation" (Cameron, 1977), which stated that if one's best life accomplishments seemed to lay ahead, attention would be future directed. Gulas and McKeage (2000) argued that college students were more likely to do social comparison because they had not been fully developed and their futures bore far more potential than most others in the population. To understand Chinese college students' processing of advertising, it seems that James' formula can be modified as follows: Self-Esteem =

Future Success Pretentions

Keeping the pretensions constant, an increase of future success will account for an increase in self-esteem. The result is consistent with Shobe et al.'s (2001) hypothesis that an enhanced orientation toward the future may increase an individual's social and economic well being. It would be interesting to see whether the enhancing effects of unaffordable ads will remain over time in Chinese college students. As pointed out earlier, Chinese college students use future success to evaluate their self-esteem and life satisfaction. If their expectations about the future are met, we may predict that they will maintain high self-esteem and life-satisfaction. If their expectations exceed the reality and they do not realize the successes they anticipate, it is reasonable to predict that would have a negative effect. To date, most Chinese cannot afford to have an overseas trip or a car. It is reported that in 1999, only 4.26 million Chinese traveled outside Mainland China for private activities) and this represented less than 0.04% of the overall population in China. Statistics

256


reported that automobile ownership was 1.14 for every one hundred people in China 4. Compared with the current living standard in China, Chinese college students' expectation about the future is high. Whether these expectations will be met in the future is a key to understand whether "future success" can be used as a constant evaluator for an optimistic audience.

LIMITATIONS Care should be taken in generalizing these results to college students and the general population due to the following reasons. First, the exposure to advertisements in the experiment was different to how people view advertisements in real life. In this experiment, the participants viewed eight print advertisements and answered two questions about each ad. To answer the questions in the experiment, the participants had to pay more attention to advertisements that they may skip in real-life viewing. Second, Chinese college students have their own characteristics that make them different from college students in other countries. They are also different from other segments of the Chinese population. Third, the implications of these results are open to debate. Although in this study, Chinese college students are influenced by luxury advertisements positively, it is hard to argue that over time this effect may sustain. As discussed above, Chinese college students take these luxury products for granted as part of their future, and it is uncertain whether this is a realistic projection. Thus, the currently "positive" effects may be the cause of detrimental long-term effects of luxury ads.

IMPLICATIONS AND DIRECTIONS FOR FUTURE RESEARCH Results in this study showed that ads of unaffordable products and services exerted different influence on Chinese college students' self-esteem, as compared to previous researched populations. Future research should examine whether these findings are applicable to other comparable populations, such as college students in western countries, to see if different cultures cultivate a different mindset toward advertised products. It would also be interesting to utilize students from developed Asian countries such as Singapore, South Korea and Japan to see if similar cultures nurture a similar mentality in interpreting advertising images. However, it would be advisable to insert a couple of control variables in the measurement instrument to further tease out the effects of exposure on different segments of a student


257

population. One of those should be the inclination to social comparison - or how likely a person is to comparing status with others. Another potential control is to use sense of materialism as a possible confound rather than a dependent variable. Effects of exposure to luxury items may be more fine-tuned if personality and materialism are controlled as covariates. F r o m another perspective, future studies may also pursue other segments of the Chinese population to see if these ads have the same impact. It is reasonable to assume that there are some other segments that are not as optimistic as these college students and may be more vulnerable to some unintended effects of luxury advertisements. This experiment can serve as a starting point in understanding the effects of advertisements in an emerging, non-negligible market such as China. The results indicate that China's population has its own characteristics and should be taken into consideration in future research. These unique characteristics may call for alternative approaches to understand social comparison and the mechanisms involved regarding different populations.

NOTES 1. Statistics reported by the Shanghai Statistics Bureau, see: http://www.statssh.gov.cn/shtj/tjnj/2OO l/tables/3_13.htm

2. Figure reported in Asian Adspend: A Review of its Development and Future Prospects. International Journal of Advertising, •7(2), 255-262. 3. Statistics reported by the Chinese Statistics Bureau, see: http://www.stats. g o v. c n/sjjw/nd sj/z g nj/2 OOO/R O1c. h tm

4. See Xinhua News Agency (September 27, 2001), www.xinhua.gov.cn

ACKNOWLEDGMENTS The authors would like to thank Dr. Joseph E. Phelps and Dr. D a v i d RoskosEwoldsen at the University o f Alabama, as well as our editor Dr. Charles R. T a y l o r and other anonymous reviewers for their invaluable input in this project.

REFERENCES Belk, R. (1984). Three Scales to Measure Constructs Related to Materialism: Reliability, Validity, and Relationships to Measure of Happiness. In: T. Kinnear (Ed.), Advances in Consumer Research (Vol. 11, pp. 291-297). Provo, UT: Association of Consumer Research. Belk, R. (1985). Materialism: Trait Aspects of Living in the Material Word. Journal of Consumer Research, 12 (December), 265-280.

258

S H U H U A ZHOU, FEI X U E AND PEIQIN Z H O U

Cai, B., & Tian, Y. (2000). Chinese Education: The Bright Spot in China's Economic Growth. Chinese Education and Society, 33(1), 53-59. Cameron, P. (1977). The Life Cycle: Perspective and Commentary. Oceanside, NY: Dabor Science Publications. Dollard, J. (1960). Fear of Advertising. In: C. H. Fryburger (Ed.), The Role of Advertising (pp. 307-317). Homewood, IL: Irwin. Durgee, J. (1986). Self-Esteem Advertising. Journal of Advertising, 15(4), 21-28. Englis, B. G. (1992). The Willing Suspension of Disbelief and Its Importance in Understanding Advertising Effects. In: C. T. Allen et al. (Eds), Marketing Theory and Applications (pp. 203-208). Chicago: American Marketing Association. Fcstinger, L. (1954). A Theory of Social Comparison Processes. Human Relations, 7 (May), 117-140. Gulas, C. S., & McKeage, K. (2000). Extending Social Comparison: An Examination of the Unintended Consequences of Idealized Advertising Imagery. Journal of Advertising, 29(2), 17-28. Hong, S. M., & Giannakopoulos, E. (1994). The Relationship of Satisfaction with Life to Personality Characteristics, 128(5), 547-559. James, W. (1983). The Principles of Psychology. Cambridge, MA: Harvard University Press. (Original work published 1890.) Leelakulthanit, O., Day R., & Waiters R. (1991), Investigating the Relationship between Marketing and Overall Satisfaction with Life in a Developing Country. Journal of Macromarketing (Spring), 3-23. Lewinsohn, P. M., Redner, J. E., & Seeley, J. R. (1991). The Relationship between Life Satisfaction and Psychosocial Variables: New Perspectives. In: F. Strack, M. Argyle & N. Schwarz (Eds), Subjective Well-Being: An Interdisciplinary Perspective (pp. 141-169). Oxford, England: Pergamon Press. Li, P. (1998, March 21st). Report on the Work of the Government (address at the 15th People's Congress). People's Daily, 1. Lyubomirsky, S., & Ross, L. (1997). Hedonic Consequences of Social Comparison: A Contrast of Happy and Unhappy People. Journal of Personality and Social Psychology, 73(6), 1141-1157. Martin, M. C., & Gentry, J. W. (1997). Stuck in the Model Trap: The Effects of Beautiful Models in Ads on Female Pre-Adolescents and Adolescents. Journal of Advertising, 26(2), 19-33. Mruk, C. J. (1999). Self-Esteem: Research, Theory, and Practice (2nd ed.). New York: Springer Publishing Company. Richins, M. L. (1991). Social Comparison and the Idealized Images of Advertising. Journal of Consumer Research, 18(1), 71-83. Richins, M. L., & Dawson, S. (1992). A Consumer Values Orientation for Materialism and Its Measurement: Scale Development and Validation. Journal of Consumer Research, 19(3), 303-316. Richins, M. L. (1995). Social Comparison, Advertising, and Consumer Discontent. American Behavioral Scientist, 38(4), 593--607. Rosenberg, M. (1965). Society and the Adolescent Self-Image. Princeton, NJ: Princeton University Press. Sekaran, U. (1986). Significant Differences in Quality-of-Life Factors and Their Correlates: A Function of Differences in Career Orientations or Gender? Sex Role, 5/6, 261-279. Shobe, M., & Page-Adams, D. (2001). Assets, Future Orientation, and Well-Being: Exploring and Extending Sherraden's Framework. Journal of Sociology and Social Welfare, 28(3), 109-127.


259

Sirgy, M. J. (1998). Materialism and Quality of Life. Social Indicators Research, 43(3), 227-260. Sirgy, M. J., & Lee, D. J. (1998). Does Television Viewership Play a Role in the Perception of Quality of Life? Journal of Advertising, 27(1), 125-143. Wang, J. (2000). Foreign Advertising in China: Becoming Global, Becoming Local. Iowa: Iowa State University Press. Wei, R., & Pan, Z. (1999). Mass Media and Consumerist Values in the People's Republic of China. International Journal of Public Opinion (Spring), 75-96. Williams, R. (1981, March 20th). Corrupting the Public Imagination. Christian Science Monitor, 22. Zinkhan, G. M. (1994). Advertising, Materialism, and Quality of Life. Journal of Advertising, 23 (June), 1-4.

260


APPENDIX 1: M A T E R I A L I S M

SCALE

Success (1) I admire people who own expensive homes, cars, and clothes. (2) Some of the most important achievements in life include acquiring material possessions. (3) I don't place much emphasis on the amount of material objects people own as a sign of success. (4) The things I own say a lot about how well I'm doing in life. (5) I like to own things that impress people. (6) I don't pay much attention to the material objects other people own.

Centrality (1) (2) (3) (4) (5) (6) (7)

I usually buy only the things I need. I try to keep my life simple, as far as possessions are concerned. The things I own aren't all that important to me. I enjoy spending money on things that aren't practical. Buying things gives me a lot of pleasure. I like a lot of luxury in my life. I put less emphasis on material things than most people I know.

Happiness (1) (2) (3) (4) (5)

I have all the things I really need to enjoy life. My life would be better if I owned certain things I don't have. I wouldn't be any happier if I owned nicer things. I'd be happier if I could afford to buy more things. It sometimes bothers me quite a bit that I can't afford to buy all the things I like.


261

APPENDIX 2: SELF-ESTEEM SCALE

Self-esteem (1) (2) (3) (4) (5) (6) (7) (8) (9) (10)

On the whole, I am satisfied with myself. At times, I think I am no good at all. I feel that I have a number of good qualities. I am able to do things as well as most other people. I feel I don't have much to be proud of. I certainly feel useless at times. I feel that I am a person of worth, at least on an equal plane with others. I wish I could have more respect for myself. All in all, I am inclined to feel that I am a failure. I take a positive attitude toward myself.

ADVERTISING COMMUNICATION IN AUSTRALIA: A COMPARISON OF INFORMATION USED BY AUSTRALIAN, JAPANESE AND U.S. FIRMS Cameron James Hughes and Michael Jay Polonsky

ABSTRACT This paper examines the information content of print advertisements for high involvement products in Australia, by Australian, Japanese and U.S. firms. Paired comparisons between firms found that the information in Japanese firms, Australian advertising is relatively more similar to information in Australian firms' advertising than is the information in U.S. firms'Australian advertising. Comparisons between the results of this study and previous works found that the information used in Australian advertising by both Japanese and U.S. firms differed from advertising within the two respective home markets. This may support the view that there is some localization of information taking place.

New Directions in International Advertising Research, Volume 12, pages 263-280. Copyright © 2002 by Elsevier Science Ltd. All rights of reproduction in any form reserved. ISBN: 0-7623-0950.4

263

264

CAMERON JAMES HUGHES AND MICHAEL JAY POLONSKY INTRODUCTION

The debate as to whether finns should standardize or localize their advertising in formation has existed for well over 30 years, with various authors arguing both cases (Onkvist & Shaw, 1999; Ramarapu et al., 1999). For example Elinder (1961), Kreutzer (1988), Levitt (1983) and Schnaar (1991) are just a few who have argued that advertising should be standardized globally to achieve economies of scale, whereas authors such as Kotler (1986), Mueller (1992) and Wind (1986) are only a few of those who have suggested that advertising should be customized to accommodate the cultural idiosyncrasies of consumers in foreign markets. The proponents of standardisation contend that consumers are converging on a global scale due to the forces of technology, the accessibility of information, migration, tourism and modern transport (Elinder, 1961; Dichter, 1962; Levitt, 1983). Essentially, if consumers are converging toward a homogenous unit, who respond with equal enthusiasm to the same marketing stimuli, then standardised advertising campaigns will be the most effective means of competing on world markets. Alternatively, if national tastes and cultural idiosyncrasies are unyielding, or if people in different countries simply interpret information differently (Phillips, 1997), then competitive advantages will only be obtained through the conceptualization and implementation of advertising strategies that are reflective of the cultural empathy of the marketer (Leach & Liu, 1998). Unfortunately there have been few attempts to model the factors that influence whether standardization is indeed an appropriate approach (Laroche et al., 2001). According to the proponents of customization, convergence is a case of 'marketing myopia' (Hill & Still, 1986; Kacker, 1972; Kotler, 1986, Toyne & Waiters, 1989; Whitelock, 1987; Wind et al., 1973). Cultures may appear to be converging, however, this is only occurring on a cosmetic level. Hofstede (1991) argues that, although national cultures in the contemporary world give the appearance of having converged, in reality this has occurred only in superficial manifestations, such as similarities in dress, consumer products, television, movies, and sport. According to Hofstede, the deep underlying values, which determine the meaning people give to these activities and practises, have not changed, thus highlighting the need to use culturally relevant stimuli within host markets (Leach & Liu, 1998; Phillips, 1997). Essentially, the ability to appeal to the motivational pattern of consumers across international markets with a standardised advertising approach depends upon the relative influence and rigidity of certain environmental contingencies, one of the most significant of which is the cultural environment. For example, in communications theory, it is well established that individuals' cultural

Advertising Communication in Australia

265

environment significantly affects the way in which they perceive information (Phillips, 1997). Accordingly, if senders (i.e. marketers) of a message live in a cultural environment that is different from their intended audience, they will need to understand the cultural background of the receiver (Kanso, 1992; Mueller, 1987; Phillips, 1997; Roth, 1995). Kanso (1992) posits that international advertisers should employ frames of reference that are equally understood by their target audience. The intemational advertiser must also match the appeal used with the cultural characteristics of the target market (Caillat & Mueller, 1987; Hong, et al., 1987; Mueller, 1992; Rice & Lu, 1988). If an appeal is inconsistent with the local culture's motivational pattem then the advertisement will fail to achieve its objectives (Leach & Liu, 1998). For example, Ricks et al. (1974) attributed most international advertising blunders to a lack of cultural awareness. Consequently, advertising strategies must be conceived with an element of cultural sensitivity and recognition of consumer and market heterogeneity. While advertising strategies within foreign markets will depend upon a range of environmental factors, the effects of home country-culture on the advertising strategies within foreign markets is often overlooked. Most of the literature examining intemational advertising, attempts to compare advertising activities in two different home countries and suggest that differences are explained by underlying differences in national culture. An exception is the research conducted by Graham et al. (1993), who examined variations in advertising strategies. By examining foreign firms' behaviour in host countries, however, they fail to compare these overseas strategies with advertising inside the companies home country. This paper attempts to partially fill this gap by examining the information content of Australian, Japanese and U.S. firms' print advertising for high involvement products within Australia. Comparisons of information are made between host and foreign firms' advertising to identify whether there are differences in the information used. Previous research has suggested that U.S. and Japanese advertising styles, with respect to information content and other factors, are very different and thus if home culture were to have an effect it would be expected that there should also be differences in the information content within Australia. Japanese culture displays "more" differences to the Australian culture than occur between the U.S. and Australian culture. However, it could be suggested that the advertisements of Japanese firms will in fact be more similar to Australian firms as Japanese firms are often considered to be polycentric (i.e. adaptive) in their approach to marketing than other countries (Schneider, 1989). Although others have recently found that Japanese automobile firms are choosing to standardize activities in all international markets rather

266

CAMERON JAMES HUGHES AND MICHAEL JAY POLONSKY

than adapt to each specific market (Grein et al., 2001), it is thus unclear if they are polycentric in relation to each market. This study attempts to examine the degree of polycentric behavior of U.S. and Japanese firms by comparing the information usage of these firms in a host market (Australia) to the information usage in advertising within Japan and the U.S., and by comparing the results of this study to those previously reported in the literature. If Japanese firms are indeed polycentric, it would be expected that greater differences in information usage would occur between Japanese firms' Australian advertising and advertising in Japan, than U.S. firms' Australian advertising and advertising in the U.S. CROSS-NATIONAL

INFORMATION

CONTENT

RESEARCH Extensive research has been conducted over the past twenty years into the cultural determinants of various types of information types and levels of information within advertising (A1-Olayan & Karande, 2000; Biswas et al., 1992; Dowling, 1980; Hong et al., 1987; Madden et al., 1986; Maynard & Taylor, 1999; Norton & Norton, 1988; Onksist & Shaw, 1999; Resnik & Stem, 1977; Taylor et al., 1997; Weinberger & Spotts, 1989). The majority of these studies have examined communication strategies in one home country and/or compared multiple home country communication strategies. Other than Graham et al. (1993), few studies have examined the effects of home country-culture on companies' advertising strategies/tactics within foreign markets. Existing research efforts have validated the proposition that different cultural groups have distinct means of communicating information through advertising (A1-Olayan & Karande, 2000; Biswas et al., 1992; Dowling, 1980; Di Benedetto et al., 1992; Graham et al., 1993; Hong et al., 1987; Lin, 1993; Madden et al., 1986, 1989; Maynard & Taylor, 1999; Taylor et al., 1997; Weinberger & Spotts, 1989). Both the level of information contained in advertisements, and the types of informational cues, vary according to the cultural context of advertising. As a result of these research efforts, there has been an accumulation of empirical data helping to establish where certain national cultures are positioned with respect to the information content of their advertising. For example, it has been demonstrated that Japanese print advertisements are generally more informative than U.S. print advertisements (Hong et al., 1987; Madden et al., 1986), Chinese print advertisements are more informative than those in the U.S. (Rice & Lu, 1988), that U.S. print advertisements are generally more informative than French print advertisements (Biswas et al., 1992), as well as more informative than British television advertisements (Weinberger & Spotts, 1989).


267

IMPORTANCE OF HOME COUNTRY-CULTURE Theoretical and empirical evidence suggests that cultural values have profound impacts upon consumer and corporate behavior (Daft & Weick, 1984; Dutton & Duncan, 1987; Erickson et al., 1984; Littrel & Miller, 2001; Mueller, 1986 1987; Schneider, 1989), including a firm's ability to coordinate activities internationally (Hewett & Bearden, 2001). Empirical evidence has indicated that cultural values may even determine behavior as complex as the selection of a particular brand within a product class category (Pitts & Woodside, 1983). Schneider's (1989) research, into the effects of national culture upon the organizational strategy formulation process, has illustrated that cultural values permeate the operational and strategic decision-making functions of firms. Schneider (1989) contends that assessments of the environment and of the company are not objective but are a function of subjective perceptions and interpretations. National culture has profound effects upon the strategy formulation process as it derives from assumptions regarding relationships with the environment as well as relationships among people (Frederickson, 1984; Lyles & Mitroff, 1985; Schneider, 1989). In the area of home country-culture, some theoretical and empirical evidence suggests that home country-culture may have a significant impact upon the strategy formulation process of companies operating in foreign markets. For example, Tse et al. (1988) investigated the impact of home country-culture on the international marketing decisions of managers from China, Canada and Hong Kong, and found that home country-culture had a significant effect on decisions. Meissner (1986) also found that home country-culture affected marketing strategies in foreign markets. He found that the German approach was productfocused, concentrated on niche markets and personal selling, whereas the Japanese approach emphasised mass marketing, advertising and special attention to customers and their needs. Graham et al. (1993), attempted to empirically validate Meissner's work by examining the intemational advertising practices of Japanese and German MNEs operating in Spanish, Indonesian and U.S. markets. The results were not totally consistent with Meissner (1986), as Graham et al. (1993) found that within foreign markets both German and Japanese companies modified the levels of information content as well as the facial expressions of models in advertisements and the number of metaphors contained in the advertisements. This suggests companies in both countries were polycentric, i.e. they varied their advertisements in the host market. Graham et al. (1993) found that Japanese companies were more proficient in their adaptive behaviours, thus supporting the view that home country-culture has varying effects upon the strategic decision-making

268


process of firms. Thus, strategic decision making, such as the decision whether to use an advertising agency based in the overseas market, home market or a global agency, may in fact be effected by the firms national culture (Lafayette, 1992; Marshall & Na, 1994; Rosen et al., 1988), although this issue has not been explicitly examined in the literature. CULTURE

DIFFERENCES BETWEEN AND USA

THE JAPANESE

There are a number of reasons for comparing the advertising behaviour of U.S. and Japanese firms in a third country. Firstly, there has been extensive literature discussing the advertising activities within these two home markets (Belk et al., 1985; Belk & Pollay, 1985; Maynard & Taylor, 1999; Mueller, 1987; Ramaprasad & Kazumi, 1992), especially information content (Hong et al., 1987; Lin, 1993; Madden et al., 1986). In relation to information content, most of this literature suggests that Japanese print advertisements are more informative, and that the advertisements contain different types of informational cues. For example, print advertisements in Japan tend to have more price cues than print advertisements in the U.S. (Hong et al., 1987; Madden et al., 1986). Secondly, these countries are very different culturally. For example, extensive differences may be illustrated using Hofstede's (1980) cultural dimension scores for individualism (U.S.-50, Japan-29); power distance (U.S.-16, Japan-21); uncertainty avoidance (U.S.-ll, Japan--44) and; masculinity (U.S.-36, Japan-50). In general, the Japanese culture is said to have developed "an attitude of receptivity and high adaptability to changing conditions" (Maruyama, 1984, p. 106). In Japanese firms, uncertainty and ambiguity is actively managed by engaging in information-generating activities. Reflecting these cultural idiosyncrasies, Japanese marketing is predominantly described as polycentric (Kagono et al., 1985; Keegan, 1983; Koichi, 1983; Lazer et al., 1985; Maruyama, 1984; Nagata, 1981; Schneider, 1989). This entails an extensive examination into the peculiarities of markets and an attempt to understand the nature of consumers in international markets (Graham et al., 1993; Koichi, 1983; Lazer et al., 1985; Meissner, 1986; Nakaishi, 1981; Yokota, 1980). Conversely, U.S. marketing practices are often stereotyped as being ethnocentrically oriented. The main criticism of U.S. marketing practices is that of an oversimplification of reality and an "over-reliance on cognitively linear explanation of events" (Pascale, 1984, p. 57). Firms in the U.S. generally set out to exert control over the external environment (Draft & Weick, 1984; Schneider, 1989; Trompenaars, 1993), and to explain marketing phenomena through statistical modeling and the development of 'lawlike generalizations.'


269

These generalizations or 'universal business principles' are then applied to all situations regardless of the cultural context. Finally, while there are cultural differences between the U.S. and Japan, there are also extensive similarities between the two countries. That is, both countries have similar levels of economic and socio-economic development, which may influence advertising strategies (Renforth & Raveed, 1983). For example, per capita income in the two countries is somewhat similar (U.S. $31,500; Japan $24,500), literacy rates in both countries are high (U.S. 97%; Japan 99%), as is the years of schooling required (U.S. 12 years; Japan 14 years). In addition the general media penetration in both countries is also similar for TV (U.S. 89%; Japan 79%) and radio (U.S. 83%; Japan 86%). All of these factors, plus the fact that U.S. and Japanese firms compete across a range of high involvement products globally, including within Australia, make them ideal for examination. Given the general environmental similarities, any differences in information content contained within advertising of foreign firms in Australia should, therefore, be explained by cultural differences between the two countries home culture. Three hypotheses will be examined in this paper. The first examines whether the levels and types of information used by Australian, U.S. and Japanese firms are the same and is included for methodological completeness. The second relates to the fact that Japanese firms should, according to the literature, most likely be more polycentric and U.S. firms more ethnocentric. Thus, the advertising of Japanese firms should be more similar to the host country firms' advertisements. The last hypothesis further considers this issue, by examining whether firms from Japan or the U.S. adapt their information content more in relation to advertising within their home country. The three null hypotheses are as follows: HI: Firms from Japan, the U.S., and Australia will use similar levels and

similar types of information within advertisements. H2: Advertisements from Japanese firms will be more similar to advertise-

ments from Australian firms than will be U.S. firms' advertisements. H3: Advertisements in Japan and U.S. will use the same types and levels of

information content as appears in Japanese and the U.S. firms' Australian advertising. METHODOLOGY This study utilised a multi-stage methodology. The first set of activities is related to the content analysis process and includes identifying, collecting the stimuli

270


(i.e. advertisement) and coding the data. The second phase involved the analysis of the data (i.e. information content) in advertisements used by Australian, Japanese and U.S. firms within Australia using ANOVA's and paired t-tests. The third phase used paired t-tests and ANOVA's to compare the results in phase two with those of Maden et al. (1986) to identify whether the information content of advertising by Japanese and U.S. firms in Australia, differed from the information content of general advertising within these firms' home country (i.e. Japan and the U.S.). Madden et al.'s (1986) work was chosen for comparison because it examined 1440 Japanese and U.S. print advertisements, contained high involvement product types and listed individual cue results. It also used three judges to code the information content, who were trained prior to conducting coding procedures, and resulted in 97.6% reliability. It is assumed there is no time bias, as Abernethy and Franke's (1996) meta-analysis into information content found there was no time effect on information levels. Although the fact that the information content is collected across samples collected 10 years apart may be one limitation of this study. Previous literature has identified that information content in advertising varies according to the product type being advertised (Dowling, 1980; Resnik & Stern, 1977). In addition some researchers have suggested that high involvement purchasing decisions are more complex (Johar, 1995; Stayman & Batra, 1991), which would infer that consumers require more information in evaluating high involvement products (Bolting, 1988; Holmes et al., 1987; von Keitz, 1990; Mazursky & Schul, 1992; O'Cass, 2000). To address this fact, the advertisements selected for examination within the study discussed in this paper were further restricted to high involvement products. High involvement products were defined as those that are purchased infrequently, relatively highly priced and require extensive information search prior to purchase (Mueller, 1987). There are of course many products that fit this definition and for the study reported in this paper three sub-groupings of high-involvement products were examined - cars; computers, electronics, and telecommunications (CET); and airlines. This refinement was undertaken to ensure there were a sufficient number of advertisements in each category for firms from the three countries, Australia Japan, U.S. This resulted in 424 different advertisements being included in the analysis. See Table 1 for a breakdown of advertisements by country and product category. In terms of the stimuli this research used all half page or larger advertisements (color or black and white) from 1996 and 1997 of a weekly popular general interest magazine and a weekly business type magazine. Both of these magazines were compatible in terms of circulation and target audiences and as

271

Advertising Communication in Australia Table 1.

Computers, electronics, telecommunications (CET) Airlines Cars Total

Advertisements by Product and Country. Japan Number (%)

Australia Number (%)

Number (%)

82 (56.9%)

63 (58.9%)

120 (69.4%)

51 (35.4%) 11 (7.6%)

19 (17.8%) 24 (22.4%)

19 (11.0%) 34 (19.7%)

144 (100%)

107 (100%)

173 (100%)

U.S.

such should have similar types of advertisements, thus minimising variation across finns (i.e. Australian, U.S., Japanese). Much of the previous cross-cultural content analysis literature has included advertisements from a wide cross section of product categories (i.e. high and low involvement). To achieve broad-based coverage of advertisements, researchers have collected materials from a crosssection of publications to ensure that no types of advertisements were over represented (for example see A1-Olayan & Karande, 2000). However, some of the cross-cultural content analysis work has been more focused in terms of its objectives and as such has examined a smaller sample of publications. For example, Maynard and Taylor (1999) used advertisements appearing in one magazine targeting young females in the U.S. and Japan to evaluate girlish images in these two countries. The study reported within this paper is concerned only with advertisements for high involvement products. As such it was deemed appropriate to examine two weekly business focused publications (i.e. 104 issues in total). It is assumed that these would include a representative sample of similar high involvement products marketed by Australian, Japanese and U.S. firms within Australia. A content analysis process, using Resnik and Stem's (1977) 14 informational cues, was used to evaluate the information content within the 424 print advertisements identified, although the taste and nutrition cues were not used, as they were not relevant to the high involvement categories being examined. This typology was employed because it is an objective, systematic and quantitative description of the qualitative communications content of advertising material. It has also been suggested that Resnik and Stem's (1977) typology has been extensively used within the literature (Abemethy & Franke, 1996) and thus, it allows this study to make comparisons to earlier works examining information content of advertising within Japan and the U.S.. However it should also be noted that some researchers such as Taylor et al. (1997, p. 7) suggest that the Resnik and Stem's (1997) criteria "omits several important types of information seen

272


fairly commonly in advertising" and that some of these other types of information are frequently used in other countries, i.e. non-U.S, advertisements (Taylor et al., 1997). The Resin and Stern (1997) criteria are still used in this paper because in addition to being extensively used in the literature (Abemethy & Franke, 1996), their use also allows for a direct comparison to the previous results of other researchers to be made. The content analysis coding in both the pre-test and final examination of the data involved three independent judges to evaluate the advertisements (Kassarjan, 1977), and inter-judge reliability was evaluated using Rust and Cooil's (1994) Proportional Reduction of Loss method (PRL). A pre-test of the criteria was undertaken on a sample of advertisehaents from a previous year to evaluate the inter-judge reliability, with a 0.91 level of agreement, which is greater than the 0.85 suggested by Kassarjian 1977. The substantive analysis of the data involved a combination of parametric and non-parametric statistical techniques. These tests included, multivariate analyses of variance, T-tests, Z-tests of proportions, and signed rank tests. ANALYSIS An analysis of the inter-judge reliability using PRL yielded a 0.94 level of agreement, which exceeds the 0.85 recommended by Kassarjian (1977) and thus allowed the substantive analysis to be undertaken. The first step involved an examination of whether there were differences in the overall informational levels and types of information within advertisements from Australian, Japanese and U.S. firms (see Table 2 for summary of results). The mean number of cues per country was Australia-l.73, Japan-2.43 and the U.S.-1.91, which are statistically different across samples ( F = 2.065, p = 0.0029). Paired comparisons between countries also identified that there were differences in mean cue usage, Aus-Japan = 2.49 (p = 0.03); Aus-U.S. = 2.85 (p = 0.02); U.S.-Japan = 3.89 (p = 0.00). A MANOVA was undertaken for the set of cues across the countries, which was significantly different (F = 2.25, p < 0.001). A non-parametric signed rank test was also undertaken on ordering of cues between the pairs of countries and identified that the types of cues used varied across countries as well, Australia-Japan = 27.5 (p = 0.02); Australia-U.S. = 27.5 (p = 0.02); U.S.-Japan = 22.5 (p = 0.04). This suggests that the usage of types of information varied across the three companies, which supports the previous research identifying that advertising in different cultures relays different types of information (Di Benedetto et al., 1992). As such, looking at aggregate levels of information is insufficient to understand all components of cultural differences.

273


©

"0

New Directions in International Advertising Research (Advances in International Marketing)

Advances in Advertising Research 1: Cutting Edge International Research

Advances in Advertising Research 1: Cutting Edge International Research

International Marketing: Emerging Markets (Advances in International Marketing, Volume 21)

International Marketing Research

New Directions in Nursing History: International Perspectives

New Challenges to International Marketing (Advances in International Marketing, Volume 20)

Go Figure!: New Directions in Advertising Rhetoric

Advances in International Accounting

New Directions in Action Research

Advances in International Accounting, Volume 18 (Advances in International Accounting)

Advances in International Accounting, Volume 16 (Advances in International Accounting)

Advances in International Accounting, Volume 17 (Advances in International Accounting)

Introduction to International Marketing (Marketing in Action)

International Marketing

New Directions in International Economic Law: In Memoriam Thomas Walde

Advances in Advertising Research, Volume 2

Small Firms in Tourism: International Perspectives (Advances in Tourism Research)

Reassessing the Internationalization of the Firm (Advances in International Marketing)

New Directions for Biosciences Research in Agriculture

New directions in wireless communications research

Progress in International Business Research, Volume 1 (Progress in International Business Research) (Progress in International Business Research) (Progress in International Business Research)

New Directions In Counselling

New Directions in Knitting

New Directions in Chenille

Advances in Electronic Marketing

Advances in Electronic Marketing

International Marketing (15th Edition)

Emerging Issues in International Business Research (New Horizons in International Business.)

An International Handbook of Tourism Education (Advances in Tourism Research)

New Directions in International Advertising Research (Advances in International Marketing)

Advances in Advertising Research 1: Cutting Edge International Research

Advances in Advertising Research 1: Cutting Edge International Research

International Marketing: Emerging Markets (Advances in International Marketing, Volume 21)

International Marketing Research

New Directions in Nursing History: International Perspectives

New Challenges to International Marketing (Advances in International Marketing, Volume 20)

Go Figure!: New Directions in Advertising Rhetoric

Advances in International Accounting

New Directions in Action Research

Advances in International Accounting, Volume 18 (Advances in International Accounting)

Advances in International Accounting, Volume 16 (Advances in International Accounting)

Advances in International Accounting, Volume 17 (Advances in International Accounting)

Introduction to International Marketing (Marketing in Action)

International Marketing

New Directions in International Economic Law: In Memoriam Thomas Walde

Advances in Advertising Research, Volume 2

Small Firms in Tourism: International Perspectives (Advances in Tourism Research)

Reassessing the Internationalization of the Firm (Advances in International Marketing)

New Directions for Biosciences Research in Agriculture

New directions in wireless communications research

Progress in International Business Research, Volume 1 (Progress in International Business Research) (Progress in International Business Research) (Progress in International Business Research)

New Directions In Counselling

New Directions in Knitting

New Directions in Chenille

Advances in Electronic Marketing

Advances in Electronic Marketing

International Marketing (15th Edition)

Emerging Issues in International Business Research (New Horizons in International Business.)

An International Handbook of Tourism Education (Advances in Tourism Research)

Recommend Documents