Cognition, Vol. 4, No. 3

Cogzition, 0 Hsevier 4 (1976) 215-230 Sequoia S.A., Lausanne - Printed in the Netherlands Skills of divided attentio...

Author: J. Mehler | T. G. Bever & S. Franck (Editors)

85 downloads 1423 Views 6MB 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

Cogzition, 0 Hsevier

4 (1976) 215-230 Sequoia S.A., Lausanne

- Printed

in the Netherlands

Skills of divided attention*

ELIZABETH

SPELKE,

W!LLIAM

HIRST,

and ULRIC NElSSEf?** Cornell

University

Abstruct Two subjects read short stories while writing lists of words at dictation. After some weeks of practice, they were able to write words, discover relations arnong dictated words, and categorize words for meaning, while reading for comprehension at normal speed. The performance of these subjects is not consistent with the notion that there are fixed limits to attentional capacity. The study of divided attention has a long history. Most early psychologists, like their contemporary counterparts, believed that consciousness could only be directed to a single activity at a time. Conscious attention to two different actions performed at the same time was thought to be possible only if they were coordinated into a single, higher-order activity, or attended to in rapid alternation. Otherwise, it was assumed that at least one of them was being carried out ‘automatically’, without conscious control (James, 1890; Woodworth, 1921). Early investigators attempted to explore the limits of consciousness by combining diverse tasks while introspecting on their performance. Paulhan (1887) recited one poem while writing another, or while executing mathematical calculations. Solomons and Stein (1896) and later Downey and Anderson (1915) practiced reading stories while writing at dictation, and noted the changes that occurred in their conscious awareness of the act of writing. These studies did not always support the view that consciousness is unitary. Experimenter/subjects variously reported that one activity was *Some of these results were reported at the American Psychological Association, Chicago, August, 1975. **The names of the two senior authors are listed in a randomly-chosen order; they contributed equally to this work. Reprint requests may be addressed to any of the authors at the Department of Psychology, Uris Hall, Cornell University, Ithaca, NY 14853.

2 16 Elizabeth Spelke, William Hirst and Ulric Neisser

performed unconsciously (Solomons & Stein, 1896), that attention alternated between the two activities (Paulhan, 1887), and that a genuine division of attention was accomplished (Downey & Anderson, 1915). Modern studies of attention have avoided the dependence on introspection which characterizes the early work. In addition, however, they have usually divorced attention from action. Division of attention has not been defined by simultaneous directed activity, but by concurrent processing in two distinct ‘channels’. In experiments on selective listening, for example, subjects are usually asked to shadow only one of two verbal messages; the other is to be ‘ignored’. Processing of the secondary input may be assessed by testing memory for the words on the ‘unattended channel’ (Glucksberg & Cowen, 1970; Norman, 1969), by measuring autonomic responses to those words (Corteen & Wood, 1972), or by observing the facilitory and inhibitory effects of the secondary message on the focal task (Lewis, 1970). Only a few studies have required subjects to perform two simultaneous tasks (e.g. Allport, Antonis & Reynolds, 1972; Shaffer, 1975; see also Welford, 1968). None of these have examined changes in dual task performance with practice (but see Underwood, 1974). Our research revives the tradition of earlier experiments on divided attention. Specifically, it replicates and extends the work of Leon M. Solomons and Gertrude Stein at the Harvard Psychological Laboratory (Solomons & Stein, 1896). We have studied the development of skills for attending to and acting on two simultaneous messages. Two subjects, Diane and John, participated in this three-part study. As they read short stories to themselves, John and Diane first practiced writing unrelated words at dictation. When their reading speed stabilized, they were asked to detect semantic relations among the dictated words. Finally, they were asked to categorize words in a manner which forced them to use semantic information. By giving the subjects extensive practice, while gradually increasing the demands of the writing task, we were able to produce very substantial increases in their ability to perform two complex and meaningful activities at the same time.

Method Diane and John, respectively a graduate student in Biology and a Cornell Hotel School undergraduate, were recruited through the Cornell Student Employment Office. They worked for five one-hour sessions a week over a period of about seventeen weeks, paid by the hour. In each session, they read short stories while writing at dictation. The stories ranged in length

Skills of divided attention

2 17

from 700 to 5000 words, and were selected from collections of works by American, English, and translated European writers. Words for the dictation lists were selected randomly without replacement from the norms of Kucera and Francis (1967). The principal dependent variables were reading speed, reading comprehension, dictation rate, and recognition memory for the dictated words. The procedure varied considerably in the different phases of the experiment, and will be described phase by phase. A full chronology of the study appears in Table 1.

Table

Chronology of the Study

1.

Practice: 14 trials per week of reading while writing at dictation mental trials, 4 recognition trials, and 1 control trial.

Sessions

l-29

Sessions

30-35

Controlled

Sessions

36-43

Dictation with embedded lists of related words: Sentences, categories, words from syntactic classes, or rhymes. Subjects that the dictated words would be structured in any way.

Sessions 44-46

testing:

1 full experimental,

1 recognition,

Dictation with embedded lists of related words: and report the occurrence of any structured followed session 46).

Sessions 47-49

Retraining

(comprehension

Sessions 50-55

Controlled the stories.

testing

Sessions

Dictation of categorizable lists, in which or the name of its category.

56-6 1

comprehension

Continuation

Sessions 69-74

Continued practice of reading 56-6 1. (Diane only).

75-80

Controlled

testing

of reading

while

at dictation

while reading

Sessions

Writing

at dictation

while shadowing.

Reading

by means subjects

categorizing

while categorizing

Writing

I: Simultaneous

Subjects sublists.

trial per day.

words from semantic were not forewarned

were asked to look for (a one-week vacation

either

of free and cued recall of wrote

the dictated

word

of sessions 44-46.

Session 8 1 82-85

and 1 control

trials only).

of reading

Sessions 62-68

Sessions

- 10 full experi-

dictated

dictated

words,

as in sessions

words.

aloud.

and Writing

After two pre-experimental sessions to be described below, the first phase of the experiment was devoted to practicing the dual task. Diane and John participated together in 29 one-hour sessions spread over six weeks. In each session, they silently read three short stories while writing words dictated by the experimenter (WH or ES). As soon as both of them had finished writing

2 18 Elizabeth Spelke, WilliamHirst and Uric Neisser

a given word, the next word was dictated. The average rate of dictation was about 10 words per minute. They wrote on plain paper, moving their hands vertically down the page for each new word. On reaching the bottom of the page, they turned to a new sheet of paper and continued to write. Except when they changed sheets, the subjects rarely looked at their writing. There were three kinds of reading trials in this phase, given in random order. In a control triuE (one each week), Diane and John each read one short story from beginning to end without any concurrent writing. At the end of the story they received a comprehension test. Comprehension questions were prepared by the first two authors. Memory for the important details of plot and character were assessed by 8 to 1.5 short answer questions (e.g., “What did Laura say to the dead man at the cottage?” was a question pertaining to the story, “The Garden Party”, by Katherine Mansfield). In a full experimental trial (ten each week), the subjects copied dictated words while reading stories; on the average about 60 words were dictated during a single story. As in the control trials, they read the stories to completion and were given comprehension tests. In a recognition trial (four each week), reading was interrupted after exactly 40 words had been dictated, and there was no comprehension test. Instead, a test of recognition memory for the dictated words was immediately administered. Recognition tests consisted of 20 randomly selected words from the dictated list, and 20 other words, (which were never dictated) from the same norms. The lists were read aloud by the experimenter; Diane and John indicated (in writing) whether they recognized each item as having been on the dictated list. Throughout the experiment, instructions emphasized the importance of writing all the dictated words, of comprehending the stories, and of reading as rapidly as possible. On the other hand, we did not encourage John and Diane to try to remember the dictated words. They were never told in advance whether reading comprehension or word recognition would be tested. At the end of each week, they were shown how much they had progressed and were encouraged to read still more rapidly. In two pre-experimental sessions, we assessed the subjects’ normal reading speed and comprehension as well as their recognition memory for dictated words. In each of these sessions, conducted before Diane and John knew the nature of the main experiments, they read two short stories and copied two lists of 40 words from dictation on separate, alternating trials. John read at an average of 483 words per minute (wpm) and answered 73% of the comprehension questions correctly; Diane read 35 1 wpm and correctly answered 90% of the questions. John correctly recognized an average of 87.5% of the dictated words, with a false alarm rate of 2.5%; Diane recognized 77.5% of them, with 5% false alarms.


2 19

The levels of comprehension manifested in the pre-experimental sessions were little affected by the simultaneous dictation task introduced in the main experiment. Comprehension was high even in the first session. Both Diane and John’s comprehension improved somewhat over the course of the practice sessions (Table 2, line 1). The rate at which words were written (and hence the rate at which they were dictated) showed no systematic change. Recognition of the words dictated on experimental trials also showed little change with practice. Recognition memory was somewhat poorer than in the pre-experimental sessions, especially for John (Table 2, lines 2 and 3). The quality of the subjects’ handwriting deteriorated rapidly in the first week of practice and then improved, appearing normal by the fourth week. Omissions and misspellings were rare throughout.

Comprehension I-35).

Table 2.

and Recognition

Memory

on Experimental

Trials (Sessions

Sessions l-5

26-29

30-35 (Testing)

99.6 89.5

97.6b 84.3b

99.2’ 86.3”

Diane John

0.72 0.61

0.76 0.66

0.88 0.68

0.70 0.71

0.80 0.68

0.82 0.72

0.76 0.70

Recognitiond p (false alarm)

Diane John

0.23 0.02

0.18 0.04

0.19 0.09

0.25 0.10

0.26 0.15

0.28 0.12

0.33 0.12

is is is is

the the the the

mean mean mean mean

of of of of

10 trials, except 8 trials. 12 trials. 4 trials.

100.0 82.2

21-25

Recognitiond P (hit)

score score score score

86.5 71.6

16-20

Diane John

Each Each Each Each

86.8 70.3

11-15

Comprehensiona (% Correct)

a b ’ d

83.4 75.0

6-10

as noted.

Reading speeds dropped sharply on the first full experimental trials, as was expected, but soon began to increase. By about the fourth week, they began to approach normal levels (Figure 1). There was a great deal of variability from one trial to the next. In part, this must have been due to the varying strategies and motivations of our subjects. A more obvious source of variability, however, was the relative difficulty of the stories being read. In particular, some authors seemed to demand slower reading than others. In the seventh week, a different procedure was adopted to confirm that Diane and John could indeed read just as fast while taking dictation as on

220

Elizabeth Spelke, WilliamHirst and Uric Neisser

Figure 1.

Reading speeds during the practice phase: weekly means and interquartile ranges. 500

-

John

0 0

460

0

420380 2

-

340-

.;

300-

$

260-

/’ O----d

$

5

220

’ 1

I

I

2

3

, 4

L 5

I 6

Week

m

I

Control

0 0 0

220 180

(x, ’ I

IO”

/’ I 2

I 3

I 4

Week

5

I 6

Control

control trials. This second phase of the experiment involved six sessions. Each day the subjects read three stories by the same author: one in a full experimental trial, one under control conditions, and one for a recognition test. The stories which Diane read on control trials were read by John on experimental trials, and vice versa. Summary results for this phase appear in the last column of Table 2. A day-by-day comparison of experimental and control trials, presented in Table 3, reveals no systematic differences. Diane and John read as quickly, and apparently as effectively, while taking dictation as when they read alone. Some weeks later, in sessions 50-55, we attempted a stricter test of reading comprehension. In each of six sessions, five control and five full experimental trials were followed by a demanding probed-recall test of memory for selected episodes from the story read on that trial. The


Table 3.

22 1

Controlled Testing of Reading Speed and Comprehension

Reading

Reading

speed (wpm)

comprehension

(% correct)

Experimental

Control

Experimental

Control

30 31 32 33 34 35

336.1a 365.8 302.1 322.2 358.2 303.6

331.0 354.9 330.8 297.6 325.2 332.4 -_

100 100 100 100 100 95

100 100 100 100 100 95 --

x

331.43

328.65

30 31 32 33 34 35

485.5 412.1 573.5 471.6 468.0 450.0

593.3 502.0 555.0 471.6 380.4 441.8 .~-

x

411.18

490.68

Session Diane

99.17

99.17

John 100 90 82.5 65 100 80 86.25

100 100 80 95 100 100 95.83

a Each score is based on 1 trial.

episodes, which ranged in length from 192 to 410 words, were divided into ‘idea units’: 14 to 43 idea units per episode. For example, from the sentence, “When he heard the whistle of the northbound train arriving from Los Angeles, he led the girl to the window” we extracted the idea units, “when he heard the whistle of the northbound train”, “the train arrived from Los Angeles”, and “he led the girl to the window”. After each story, Diane and John were first asked to give a written account of the episode in as much detail as they could. Then they answered probing questions about all the ideas that had been left out of their recalls. One question served as a cue for each omitted idea unit. For example, the last cue for the sentence above was “Where did he lead the girl?” This procedure revealed no decrement in comprehension or memory that could be attributed to the added task of writing from dictation. John’s mean probed comprehension, in terms of the proportion of ‘idea units’ recalled, was 0.90 on experimental trials and 0.88 on control trials. Diane’s probed comprehension was 0.94 on experimental and 0.95 on control

222

Elizabeth Spelke, William Hirst and Uric Neisser

trials. Their initial free recall scores (the percentage of idea units recalled before the probing questions) were about 20 percentage points lower in all conditions.

II. Detection

of Structured

Sublists

The observations reported so far establish that John and Diane could copy dictated words while reading with normal speed and comprehension, but they give little indication of how much information the subjects picked up from the dictated words. In the second part of the study, we explored the degree to which they analysed and understood the words they wrote. In these sessions, the subjects were observed individually and no recognition or control trials were administered. Instead, John and Diane were asked to report any of the dictated words, or any ‘general properties’ of the list, which they remembered. They were also asked why they thought they remembered what they did. Lengthy stories were used (three per session) so that we could dictate lists of 80 to 100 words without interruption. Unknown to the subjects, the lists were no longer entirely random. Each included a sublist of 20 consecutive words that were interrelated in one of four ways. On the first day, the words of the sublist all came from the same superordinate category: the three trials used the categories ‘furniture’, ‘vehicles’, and ‘dwellings’, respectively. On the second day, the sublist words all came from one of three syntactic classes: plural nouns, past tense verbs, and adjectives. On the third and fourth days, consecutive words in the sublists formed sentences. These six 20-word lists each included two to five sentences, three to ten words long. On the fifth day, the words of the sublists rhymed: each consisted of 20 words rhyming with the words ‘board’, ‘bee’. and ‘bean’, respectively. The category sublists were taken from the Battig and Montague (1969) norms; the others were constructed ud I?oc by the authors. Each sublist appeared after the first 35 to 45 random words in the longer list. Of the several thousand words dictated in this phase, only 35 were spontaneously recalled. The subjects gave several reasons for these recalls. In six cases, the word had some personal significance: Diane recalled ‘diameter’, which she at first thought was her own name, and John recalled several words related to his studies, such as ‘luncheon’ and ‘finances’. In ten cases, the word recalled was semantically or phonetically related to the story being read. John noticed ‘ecumenical’ while reading a story about a priest, and ‘aversion’ while reading the word ‘version’. In six cases, one of the subjects was uncertain about exactly what word the experimenter had said, and he


223

“had to think about it”. No reasons were given for the recall. of the remaining 13 words. The subjects seemed completely unaware of the presence of the sublists on the first four days of this phase. Neither of them noted the existence of the categories, the consistent syntactic classes, or the sentences. Neither recalled more than two words from any sublist. The single exception was the phrase ‘muddy water’ from the sentence ‘Dogs drink muddy water’. Both subjects reported this phrase, but assumed that juxtaposition of words was accidental. This failure to notice the list structures is quite striking. As a control, we asked each of three naive subjects to copy one of the 80-word lists from dictation without looking at it, and subsequently to report such words and general properties as he could remember. Each type of list was read to one subject. Those who were given category and sentence lists noticed the structure immediately, though the subject who was given 20 words from the same syntactic class, plural nouns, did not. The effect of the rhyming list, given on the fifth day, was very different. Both John and Diane noticed the rhymes on the first trial (as did another naive control subject). After these sessions, we showed Diane and John the 15 sublists they had copied and asked if they remembered noticing anything about them. They confirmed that they had not. Indeed, they were not easily convinced that these lists had actually been dictated. They found it hard to believe, for example, that they had copied “trolley, skates, truck, horse, airplane, tractor, car, rocket, bike, taxi, scooter, jet, trailer, subway, tank, feet, cab, ship, tricycle, van” without noticing the category. In the final ten sessions of this phase, we determined whether the subjects, now alerted to the possible presence of structure in the dictated lists, could detect it on request. Each day, they read two very long stories (4500 to 7000 words); 200 or more words were dictated during each story. Five tenword sublists were embedded in each (otherwise random) dictation list. One such sublist consisted of words from a particular category, one of words from a given syntactic class, one of rhyming words, and two of sentences. The order of sublists, and their positions in the 200 word list, were randomly determined. As always, Diane and John were encouraged to read at their normal rate with full comprehension. In addition, they were asked to indicate whenever they had noticed a sublist by interrupting the experimenter and telling him the basis of the relation among the words (e.g., ‘sentence’, ‘clothing’). These final sessions were originally planned to take only three days, which immediately followed the earlier sessions in phase II. Five weeks later, in sessions 62-68, we returned to this task to obtain more information about performance under these conditions.

224

Elizabeth Spelke, William Hirst and Uric Neisser

The subjects proved to be quite good at detecting the structured sublists once the task had been set for them. Rhymes were always, and superordinate category lists nearly always, detected. Diane identified rhyming sublists after only 3.2 words had been dictated, on the average, and category lists after an average of 5.0 words. John detected rhyming and category lists after 2.3 and 3.1 words, respectively. Sentences were detected most of the time (42 of 69 times by Diane, 41 of 55 by John), and syntactic class lists about half the time. Diane and John were slightly outperformed by two control subjects, who each copied three of our lists from dictation under the same instructions but without simultaneous reading. The reading speed of both subjects dropped when this phase began, and again when it was resumed (Table 4). John’s reading speed recovered rapidly, while Diane’s increased more gradually. By the final sessions they read at rates comparable to those exhibited during the controlled testing of sessions 30-35. Diane’s comprehension was high throughout these sessions; John’s declined and then recovered. The initial decline in reading performance indicates that the demand to report structure from the dictated list was not fully compatible with the reading and copying skills that the subjects had developed in the preceding sessions. Table 4.

Reading

Speed

and Comprehension

Diane _

_____

While Detecting

Structured

Sublists.

John

Sessions

Speed

Comprehension

Speed

Comprehension

44 45 46

252.3a 339.4 409.5

92 100 100

388.5 500.1 442.8

42 78 100

(...) 62 63 64 65 66 61 68

283.0 299.4 326.4 310.5 360.6 342.6 325.5

100 95 100 100 100 100 95

385.7 531.6 403.0 474.0 520.8 655.2 448.5

72 60 98 98 100 78 100

a Each score is the mean of 2 trials.

The fact that the subjects did not read with normal speed or with full comprehension on some of these trials suggested a further analysis. Were they more sensitive to relations among words on trials in which they read


225

more slowly, or more superficially? We made a separate tabulation of the structure-detection data for those trials on which normal speed and 90% comprehension were achieved, and for the remaining trials. No systematic differences appeared. III. Reading

While Categorizing

Words

Judging from their ability to detect the structured sublists in the final sessions of the second phase, Diane and John appeared able to read and write simultaneously while understanding both the stories they read and the words they wrote. In order to obtain clearer evidence about the ability to extract meaning from the dictated words, a new task was introduced. On some trials, Diane and John were asked to write the names of superordinate categories to which the words belonged, rather than the words themselves. In this phase, every dictated list consisted exclusively of words from one or the other of two semantic categories, such as ‘animals’ and ‘furniture’. Different categories, either from Battig and Montague (1969) or devised by the authors, were used on each trial. We announced the names of the two categories immediately before the start of each trial. The first six sessions consisted of four kinds of trials. On ‘word trials’, John and Diane wrote the words that were dictated. On ‘category trials’, they wrote the name of the superordinate. Word and category trials both used the fully categorizable lists described above. Every trial was followed either by a reading comprehension or a word recognition test. Recognition tests consisted of 20 randomly selected items from the dictation list and 20 distracters, never dictated, from the same semantic categories. Each of the six sessions consisted of one category trial with a comprehension test, one category trial with a recognition test, and one word trial. The word trial was followed equally often by a recognition or a comprehension test. Reading comprehension and recognition memory were unaffected by the new categorization task (Table Sa). Reading speed dropped markedly for both subjects on the first few sessions of the categorization trials (Figure 2). By the end of these six sessions, only John appeared to have reached normal speed; Diane was given additional practice, with categorization trials only, for six sessions. By the end of her additional practice sessions, Diane too appeared to achieve normal reading speed while categorizing words (Figure 2). The final six sessions of the categorization phase attempted to verify that both subjects were reading as well while categorizing as they did normally. In each two hour session, the subjects read seven stories by the same author. Each session consisted of six category trials with comprehension tests and

226

Elizabeth Spelke, WilliamHirst and Ub-icNeisser

Comprehension and Recognition Memory: Categotization Phase

Table 5.

a. Sessions 564

1 Category

Trialsa

Comprehension (70 correct)

Diane John

Recognition

Diane John

0.91 0.86

0.83 0.86

Diane John

0.23 0.12

0.28 0.17

P (hit)

Recognition P U:A)

b. Sessions

98 88

Word Trialsb

75-80 Category

Comprehension (% correct) t Each Each i Each Each

100 80

score score score score

Diane John is is is is

the the the the

mean mean mean mean

of of of of

100 81

TrialsC

Control

Trialsd

100 88

6 trials. 3 trials. 36 trials. 6 trials.

one control trial (with no writing at dictation). Reading comprehension appeared little affected by the writing task (Table 5b). Diane read with full comprehension, both on categorization and on control trials. John’s comprehension on control trials exceeded his comprehension on categorization trials slightly, but both sets of scores were within his usual range. Reading speed also appeared unaffected by simultaneous categorization (Figure 2). After sixteen weeks’ practice, Diane and John were able to categorize words semantically while reading at normal speed, and probably with normal comprehension. Discussion Diane and John appear able to copy words, detect relations among words, and categorize words for meaning, while reading as effectively and as rapidly as they can read alone. What accounts for their surprising abilities? We conclude this report by considering several possible descriptions of the attentional skills that they acquired. Following Paulhan (1887), one might suggest that Diane and John rapidly ‘alternated their attention’ between the tasks, making use of redundancies


Figure 2.

221

Reading speeds during the categorization phase. A----A o---o

category trials word trials control trials

Sessions

Sessions

in the stories they read to avoid any decrement in performance. This hypothesis is not directly tested in our work, and indeed it may not be testable at all. Our results do show, however, that this hypothetical alternation would have to occur so rapidly as to take no measureable amount of time. Paulhan never predicted (or achieved) this degree of efficiency in any of the task combinations he studied. The other traditional explanation for our results was first offered by Solomons and Stein (1896). These authors suggested that one learns to read and write simultaneously by training attention away from one of the tasks: one learns to write ‘automatically’. Automaticity is a widely used concept in the literature on human performance, and it has been assessed by a number of different criteria. Solomons and Stein judged the automaticity of their behavior introspectively: they considered their writing to be ‘automatic’ when they ceased to be aware of it. Introspections do not always agree, however, and Downey and Anderson (19 15) reported no full loss of consciousness when they read and wrote together. The introspective reports

228

Elizabeth Spelke,

William Hirst and Ulric Neisser

of Diane and John were no more decisive. They sometimes reported that they thought clearly about each dictated word, repeating it to themselves while copying it. On other occasions, however, they said that they were unaware of even writing. A more objective “operational indicant” of automatic processing has been suggested in a recent theoretical discussion by Posner & Snyder (1975). An activity or a mental process might be called ‘automatic’ if it caused no interference with a concurrent attentive activity. By this criterion, Diane and John’s writing would seem to be ‘automatic’ by definition. An interference criterion of automaticity becomes more interesting when we ask if our subjects’ writing at dictation would interfere with concurrent activities other than the one on which they were trained. We did explore two transfer tasks in the final week of study: the subjects wrote at dictation while reading aloud (for one day) and while shadowing prose (for four days). Writing at dictation caused a decrement both in reading aloud and in shadow prose, but not if they shadow single letters (Shaffer, 1975). An the interference began to decrease with practice. We do not regard these transfer tasks as a definitive test of the automaticity of writing by the interference criterion. Indeed, we doubt that any definitive test will be possible. Whether or not a given response interferes with a given task depends on the nature of the response and the nature of the task. Typists appear to type ‘automatically’ from written copy if they shadow prose, but not if they shadow single letters (Schaffer, 1975). An examination of subjects’ performance on a wide range of dual tasks need not converge in any simple way on a unitary conception of attention or capacity. A third conception of automatism, which we prefer, would term behavior ‘automatic’ if it did not involve certain high-order attentional skills. We suggest that attention be regarded as a matter of extracting meaning from the world, and perceiving the significance of events. Attention is involved in comprehending what one reads or hears, or in following any meaningful event over time. Our results suggest that the writing skills developed by John and Diane in the first eight weeks were not of this kind. Since they failed to notice sentences and categories in the dictated lines, they were evidently copying the words without much semantic analysis. In this sense, their writing might be called ‘automatic’. As the demands of the experiment changed and the subjects were given additional practice, however, they gradually learned to analyze the dictated words semantically and to detect simple sentential relations among them. Finally, both subjects succeeded ;n categorizing dictated words with no loss of reading speed or comprehension. By our definition, their writing was no longer ‘automatic’, as it had been in earlier stages of practice.


229

Since we did not dictate connected discourse to our subjects, we do not know whether they would have become able to read normally while following another meaningful sequence over time. That achievement remains to be demonstrated. It seems clear, however, that they understood both the text they were reading and the words they were copying. In at least this limited sense, they achieved a true division of attention: they were able to extract meaning simultaneously from what they read and from what they heard. Our results suggest that attention itself is based on developing and situation-specific skills. Particular instances of attentive performance should not be taken to reflect universal and unchanging capacities. Performance necessarily depends on one’s knowledge about a particular set of tasks and situations, and one’s skills for coping with them. Although individual strategies may have their own limitations, there are no obvious, general limits to attentional skills. Studies of attention which use unpracticed subjects, and infer mechanisms and limitations from their performance, will inevitably underestimate human capacities. Indeed, people’s ability to develop skills in specialized situations is so great that it may never be possible to define general limits on cognitive capacity.

References Allport,

D. A., Antonis, B., & Reynolds, P. (1972) On the division single channel hypotheses. Q. J. exp. Psychol., 24, 225-235.

of attention:

A disproof

of the

W. F. & Montague, W. E. (1969) Category norms for verbal items in 56 categories: A replication and extension of the Connecticut Category Norms. J. exp. Psycho/. Monograph, 80, No. 3, Part 2. Corteen, R. S. & Wood, B. (1972) Autonomic responses to shock-associated words in an unattended channel. J. exp. Psychol., 94, 308-3 13. Downey, J. E. & Anderson, J. E. (1915) Automatic writing. Amer. J. k.vchol., 26, 161-195. Glucksberg, S. & Cowen, C. N., Jr. (1970) Memory for nonattended auditory material. Cog. PsychoL, I, 149-156. James, W. (1890) The Principles ofPsychology, Vol. 1. New York, Henry Holt and Co. Kucera, H. & Francis, W. N. (1967) Computational analysis of present-day American English. Providence, Brown University Press. Lewis, J. L. (1970) Semantic processing of unattended messages using dichotic listening. J. exp. Psychol., 85. 225-228. Norman, D. A. (1969) Memory while shadowing. Q. J. exp. PsychoI., 21, 85-93. Paulhan, I:. (1887) La simultan&t& des actes psychiques. Revue Scientifique, 13. 684-689. Posner, M. I., & Snyder, C. R. R. (1975) Attention and cognitive control. In R. Solso (Ed.), Information Processing and Cognition.. The Loyola Symposium. Potomac, MD, Lawrence Erlbaum Associates. Shaffer, L. H. (1975) Multiple attention in continuous verbal tasks. In P. Rabbitt & S. Dornic (Eds.), Attention and Performance V. New York, Academic Press. Battig,

230

Elizabeth

Spelke,

William Hirst and Uric Neisser

Solomons, L., & Stein, G. (1896) Normal motor automatism. Psychol. Rev., 3, 492-5 12. Underwood, G. (1974) Moray vs. the rest: The effects of extended shadowing practice. Q. J. exp. Psychol., 26, 368-372. Welford, A. T. (1968) Fundamentals ofskill. London, Methuen. Woodworth, R. S. (1921) Psychology: A study of mental life. New York, Henry Hold and Co.

RPsumP On a demand6 i deux sujets Apres quelques semaines de entre ces mots et les classer qu’ils lisaient. La performance de ccs sujets

de lire des petites histoires tout en ecrivant sous dictee des listes de mots. pratique les sujets ont pu Ccrire les mots dictes, dccouvrir des relations selon leur sens tant en lisant i une vitesse normale et en comprenant cc est en disaccord

avec l’idee que la capacitk

d’attention

a des limites fixes.

Cognition, 4 (1976) 231-280 o Elsevier Sequoia S.A., Lausanne

- Printed

in the Netherlands

Sex differences in cognition *

HUGH FAIRWEATHER** Universities

of Oxford

and

Bologna

Abstract Sex dtfferences in cognitive skills, grouped into three areas - motor, spatial, and linguistic ~ are assessed in the context of current notions of cerebral lateralization (Bujrery and Gray, 1972). There are few convincing sex differences, either overall, or in interactions with (putative) functional localization. There are several qualifying criteria (nature of further interactions with age, birth order, culture, sex of experimenter, sex role pressure) which would have to be met, but these are as yet inadequate1.v documented. Serious caution is urged on the proliferating number of researchers in this area.

Introduction Historical

Psychological differentiation of the sexes has been a topic of concern to modern psychology ever since the subject laid claim to its title at the turn of the century. In England, in parallel with the rise of the suffragettes, Burt (1909, 19 12) was busy comparing the performances of boys and girls in secondary schools in Oxford and Liverpool, and finding that girls excelled in memory, in judgments of pitch, and had an ‘amazing’ advantage on two-point touch discrimination. In America too, reviews began to appear regularly in the major journals. Woolley (1914) noted the female propensity for card dealing and clerical tasks; and also made the suggestion that perhaps the availability of alternative employ (prostitution, drudgery) artificially lowered the proportion of female defectives apparent in mental institutions. *Research supported by Society. **Present address: Istituto

the

Social

di Psicologia,

Science

Research

Universit$

Council

di Bologna,

of Great

Britain,

V.B. Pichat 5, 40127

and the Royal Bologna,

Italy.

232

Hugh Fairweather

Goodenough (1927) found that girls were more intelligent at 3 and 4 years, but not at 2; and that girls had better memory and verbal abilities, whereas boys were better at imitating movements, discriminating right from left, and in some tests of visual perception. Subtly, a dichotomy was being distilled. Allen (1930) noted: “Boys are generally conceded a superiority in things mathematical and concrete; girls in language work, memory, and abstract reasoning;” but added “Yet the case is not at all clear”; and “Several studies, however, present conflicting data”. Such caution was echoed elsewhere in studies of children by Schiller (1934), Garrett, Bryan and Per1 (1935) and Reichard (1944); but by this time reviewers as such had hibernated: either, one supposes, the dichotomy was set, or else there was no dichotomy at all. The topic trickled back to the text book surface of the 1950’s, but often only as reviews of earlier reviews. Among the better of these was Terman and Tyler (1956) which also took the debate into the developmental court, whilst being more interested in social behavior as both symptom and aetiology. Both the approach and the arena were extended in Maccoby (1966), the appendix to which contained a valuable source of annotated references to studies of sex differences in childhood. The next, and thus far, most comprehensive, was provided by Garai and Scheinfeld (1968) although here neither age ranges nor critical reading of sources were very salient considerations: a bibliography sine annotation. The advent of the 1970s brought the eruption of a new and outspoken phase of the women’s liberation movement. Claims that there were no substantial inherited sex differences in ability (Greer, 1971) evoked not the earlier dispassion displayed by Burt ( 19 12), but a closing of psychological ranks, more perhaps in reaction to the manner than the message. Many of the new commentators were educated in the animal laboratories where sex differences were certainly both more circumscribed, and more open to precise physiological correlation, than in humans. Analogies were easy to see, and to those committed to the construction of animal models of human behavior, undeniably seductive. Not only was there a dichotomy of human behavior, but this was to have both a good and necessary physiological foundation. The dichotomy was characterized variously as modal (Komer, 1973) emotional-hormonal (Broverman, Klaiber, Kobayashi and Vogel, 1968) or emotional-cerebral (Buffery and Gray, 1972); but above all, biological (Hutt, 1972). Females have been seen, modally, as more receptive (generally) within the tactile and auditory domains, although retaining particular high class discriminative abilities such as that involved in face recognition; and more responsive with the finer musculature - fingers, faces, mouths, the

Sex differences in cognition

233

vocal apparatus. A recipe for instant social success? Emotionally more dependent, they are ‘sympathetic’ both in nature and nervous system. As a result less exploratory, they fail to develop the independence of immediate surrounds necessary for orientation in large spaces, or the manipulation of more immediate spatial relations. Cerebrally, they live with language in the left hemisphere. Males on the contrary are characterized as brashly visual, preferring simple responsive stimuli, responding best with grosser movements; fearless and independent; parasympathetic, and right-hemispheric; and, ultimately, successful. But analogies and likewise dichotomies are dangerous. What had before been a possibility at best slenderly evidenced, was widely taken for fact; and ‘fact’ hardened into a ‘biological’ dogma. But how accurate was the appraisal of evidence on which this new dogma was based? Do males and females differ in their abilities? Are these differences irreducible, or merely attributable to modes of social training? Do we have a biological theory? Current

status: Maccoby

and Jack&

19 75

The ‘facts’ about sex differences have lately been dealt, in the form of Maccoby and Jacklin’s “The Psychology of Sex Differences”*, the severest and most discursive reappraisal they are every likely to attract. This statement is not made, however, without reservation: the very size of the enterprise alone will, one suspects, mitigate against its effectiveness. The book as a whole is unlikely to be read at length except by specialists with particular vested interests. The tables summarizing studies in individual topic areas, as convenient ready-reckoners, however, will certainly be scanned and tallied. The format here though is deceptive: good and bad studies alike get equal weight; one positive finding gets (at least in the ‘difference’ column) as much mention as numerous negative results in the same or follow up studies; table headings themselves may denote ‘parcel’ areas without in reality admitting of the corresponding ‘parcel’ answers they tempt; some areas (reading is a notable example) may fall in the gaps between tables; and finally, even allowing the accuracy, such a numerous ‘no difference’ is a difficult story to remember, and may in turn obscure the few clear findings. Methodological problems are also given short shrift. Tendencies to use the simplest of statistics, or none at all; to argue from single-sex studies; for one well-published positive finding to be remembered where a score of subsequent failures-to-replicate are forgotten; and the retreat-to-neutrality inherent in standardizations of many widely-used test batteries (omission of *Hereafter

referred

to, in parcnthescs,

as ‘M & .I’.

234 Hugh Fairweather

items yielding large sex-differences): all these are alluded to, but rarely dealt with in depth. One looks in vain for a solid discussion of sex-of-experimenter effects. One final, perhaps forgiveable affliction is that curious North American disease ‘Myopia U.S.A.‘. Much of the European, and notably British literature is missing; regrettably so, since studies here seem much less prone to spectacular findings, especially in childhood. The biological

candidate:

brain lateralization

The question of whether or not we have a biological theory is to an extent evaded by Maccoby and Jacklin, who excuse themselves on the grounds of not being biologists (and also, one suspects, because of the prematurity of the question in the face of so much negative evidence). Three possibilities are in fact discussed briefly at the end of Chapter 2. Of them, two (genetics, hormones) we shall only touch on tangentially. The third, (brain lateralization) affords a convenient and heuristic framework for the present review since it both dovetails the Maccoby and Jacklin conclusions about sex differences in cognition; and also stimulates the most active current theoretical debate in the area. Cognition occupies a large portion of the book (38 of 86 tables) and what we learn at the end is that Maccoby and Jacklin are prepared, at least in adulthood, to subscribe to that earliest and most persistent of cognitive dichotomies - namely that females excel on verbal skills, males on spatial. In the face of the now considerable evidence (see, for example, Dimond and Beaumont, 1974) indicating that such a functional dichotomy may be mirrored across the cerebral hemispheres (linguistic skills located primarily in the left, spatial in the right) one suspects they might also be prepared to subscribe to some formulation of the sex difference in these terms. They do indeed entertain such a possibility briefly (pp. 125 127) and refer the reader to Buffer-y and Gray (1972) “for detailed documentation”. Buffery and Gray outline a developmental and neuro-psychological model in which an hypothesized earlier cerebral lateralization in girls leads particularly to an enhancement of left-hemisphere ‘linguistic’ function, and also, to a lesser extent, of right-hemisphere ‘visuo-spatial’ function, relative to the comparatively undifferentiated state of boys’ hemispheres and respective functions. The riders to this are that a high degree of lateralization promotes the better overall linguistic performance, whereas ambilaterality is conducive to better visuo-spatial and motor functioning. The documentation supporting these propositions was, however, inadequate (note particularly the tendency to argue by regression from adulthood, and by analogy from the animal literature, for a thesis based in childhood) not to say inaccurate (note discussion of the work of Kimura et al.,

Sex differences

in cognition

235

and see later section on Laterality and Linguistic Skills). The possibly extenuating argument - that some male superiorities may be hidden by maturational lags in childhood - is untrue in the particular case of the grosser musculature (e.g., grip strength: Ingram, 1975a, for young children; also Barnsley and Rabinovitch, 1970, for adults); irrelevant (even if true) for ossification, dentition and attainment of maximum height and weight; and empty, given the lack of evidence actually correlating cognitive performance with the parameters of females’ supposed accelerated maturation. Where there is some evidence, it fails to assist (e.g., Gilbert, 1973, acquisition of vowel sounds correlates equally with both bone and chronological age). Even when a maturational factor is apparently isolated, its effect may wash out with appropriate control. Poppleton (1968) for instance found that early-maturing girls scored higher than late-maturing contemporaries on tests of IQ and attainment, but a far more potent finding was that they tended to come from smaller families. The present review will thus be structured as follows. First we shall examine such evidence as there is for cognitive precursors in infancy; subsequently turning to childhood and adulthood, and the critical areas of motor skills (essentially ignored by Maccoby and Jacklin), spatial skills, and linguistic skills, updating the findings on each where appropriate. With respect to each area we shall briefly review the evidence locating such skills within the cerebral hemispheres, and ask whether each sex plays an important interactive role in such location.

Cognition

in infancy

Experiments in early childhood are both difficult to perform and to interpret. Neonates are only very occasionally receptive to outside influence: Berg, Adkinson and Strock (1973) for example rated only 77 out of 5,400 observations as ‘alert’. In particular test situations, many children are simply untestable because they fail to attend. The proportion may be as high as 50% and may vary with sex, but the pattern is rarely consistent. Berg et al., (1973) found girls more alert, but Komer (1970) did not. Moss and Robson (1970) found no difference in wakefulness, drowsiness and crying, only that boys ‘fussed’ more at 3 months; the findings are precisely counterposed by Lewis (1972). Then there is the problem of ‘fitness’. The excess of boys over girls is highest at conception, and subsequently diminishes (Taylor, 1974). Boys are seen to be less ‘fit’: they evidence more low Apgar scores at birth, and of neonatal abnormalities with an uneven distribution, almost three times as

236

Hugh Fairweather

many are more common among boys (Singer, Westphal and Niswander, 1968, for a population of 15,000 American children). Boys are also more prone to convulsions in the first year of life (Pringle, Butler & Davie, 1966, from their longitudinal study of 1 1,000 British children), and many neonate studies fail to include boys at all because of the disturbances attendant upon circumcision. This evidence is rarely reflected in test norms: there are no differences in Bayley Scales’ scores between one and fifteen months (Bayley, 1965); and Solomons and Solomons (1964) also find no differences at four months, though there is a tendency for boys to score higher amongst first borns, girls amongst second borns. Similarly, in later childhood, the sex ratio for children in remedial reading classes, heavily biassed to boys (e.g., Cashdan, Pumfrey and Lunzer, 1970-71) contrasts with the lack of sex differences in tested language skills (see later section). In early childhood, then, one is often faced with small numbers of highly selected children amongst whom the individual differences will far outweigh any more specific sex differences that may appear. Perception

Not surprising, then that Maccoby and Jacklin provide, in Chapter 2, abundant evidence for the null hypothesis in early perception (vision, audition, touch). One may note, additionally, that girls have shorter cortical evoked response (ER) latencies to visual stimulation (Engel, 1967); and that significant correlations have been reported between these latencies and scores on the mental, line motor and gross motor scores on the Bayley scales, although the latency alone accounts for only 10% of the mental score variance. The picture of better correlation with motor rather than mental performance has been reinforced more recently by findings regarding the ability to walk unsupported at 1 year (Jensen & Engel, 1971) and articulatory abilities at 3 years (Engel & Fay, 1972). In contrast, neonatal ER latencies bear no relation to a variety of intelligence test results and a finger recognition test at 7 years (Henderson & Engel, 1974); nor, for that matter, do latencies taken at 7 years (Engel & Henderson, 1973). In addition, the sex difference in the latency of the visual ER is not matched by findings for visual tracking (Korner, 1973; see also Miller, Cohen & Hill, 1970, for infants). Although there are no early sex differences in the auditory evoked potential (Engel, 1967; Ohrlich and Barnet, 1972; Buchsbaum, Henkin and Christiansen, 1974); or in the behavioral response to a burst of 80 db noise (see Korner, 1973) girls have been found more responsive using cardiac acceleration as the response measure (Vranekovic, Hock, Isaac and Cordero,


237

1974; contra Simner, 1971). Girls had a higher resting heart rate, and also took longer to accelerate, and, subsequently decelerate. Whilst “Tactile sensitivity” appears somewhat optimistically under “Open questions” in the concluding chapter, more negative evidence comes from Caldwell and Leeper (1974), and there remains no ready explanation for the confinement of greater female sensitivity to breast- rather than bottle-fed babies (see references to work of Bell et al.). Bell and Darling (1965) stress the need for longitudinal study, consideration of parental age and occupaGullickson and Crowell (1964) tion, family size, pre-natal medication; stress the factors of stimulus intensity and the spacing of trials. Learning

There is slight evidence that boys attend more initially but habituate more readily to visual stimuli (M & J, Table 2.4), but the explanation that this is due to their retarded maturation (older children in general attend less) becomes circular in the absence of relevant independent evidence (cf. M & J p. 18). Probably the most hard-headed series of studies in the area of early learning have been those by the Carons and their collaborators, who consistently caution against over-hasty generalizations. Their most ambitious project (Caron, Caron, Caldwell and Weiss, 1973) comprised an initial sample of 586 children at 16- 18 weeks of age, of whom 300 proved testable; their footnote (p. 390) commenting on the reasons for such attrition, is an important one: “... it is unclear whether the remainder, those excluded because of failure to habituate, are indeed perceptually unique, as some have maintained (McCall, 1971) or are comparable to their peers. Our own observations indicated that many infants did become satiated for the repeated stimulus in terms of ‘protest’ behaviors, although their fixation scores did not show it. What we may be indexing, then, is not deviant development, but different rates in which particular response systems are coming under discriminative control.” The stimuli used were faces, and boys looked longer in all phases (prehabituation, habituation, and recovery) though not always significantly more during the middle phase (particularly in the further experiment at 20-22 weeks, for which one may also note the much higher attrition rate for males). The authors again point to the confounding findings on spontaneous fixation and habituation which hardly justify claims for precocious development in females based only on looking times. Their list of cautionary riders extends elsewhere to sampling artifacts, adaptation to the test situation, short versus long term familiarization, and the modality of reinforcement. Other recent failures to find clearly interpretable sex differences

238

Hugh Fairweather

include: Banikiotes, Montgomery and Banikiotes (1972) and Vietze, Foster and Friedman (1974). Faces, perhaps the most popular stimulus in infant research, have inevitably thrown up a spectacular claim about sex differences. Fagan (1972) is widely quoted as having demonstrated superior discriminative ability in girls at 5 to 6 months, a finding naively interpreted as the substrate for their later greater sociability. Fagan (1972) reports sex differences in 2 of 7 studies; they are confined to photographs (not line drawings or masks) and, curiously, singletons (not twins). A subsequent, and equally extensive study (Fagan, 1973) using similar materials, but interpolating delays between the familiarization and recognition phases, fails to substantiate the earlier finding. Haaf (1974) at 5 and IO weeks; Cornell (1974) for groups from 18 to 24 weeks; and Lasky, Klein and Martinez (1974) for Guatemalan infants at 5 months, all similarly find no sex differences for facial discrimination. And indeed, if this were to be a base for later social behavior, one might expect corroborative evidence in older children and adults. There is none (for children 3 to 14 years, recognition of inverted classmates’ faces: Brooks & Goldstein, 1963; for adults, see Going and Read, 1974, and citations in Ellis, 1975: only same-sex asymmetry slightly greater in females).

Motor skills The pattern of evidence in infancy is extended in childhood: a predominance of males among those with poor general motor development, yet a failure to substantiate this sex difference in specific tests. Singer et al., (1968) report more males with abnormally low scores on both Fine and Gross motor tests, significantly so on the former. Whilst the authors claim that females did better than males in each of a variety of co-ordination tests in a smaller and somewhat atypical population (7 1 Buffalo private patients) this does not seem necessarily in accord with the mean scores as presented, the actual methods of scoring going unmentioned. Shapiro (1974) also reports an overall female superiority in Fine motor skills, for a small and ethnically heterogeneous group of 3 year olds, many of the test items deriving from the Denver Development Screening Test (Frankenburg & Dodds, 1967). Frankenburg and Dodds themselves, in commenting on the norms, conclude: “A comparison of the performance of boys and girls did not reveal any although some minor differences were marked, systematic differences, noted. In the gross motor items some of the differences noted were that boys could kick, throw, and catch a ball earlier than girls; and girls could hop and walk heel-to-toe earlier than boys”. Much earlier Gutteridge (1930),


239

observing a large sample of children mostly between 4 and 6 years, had noted boys’ comparative advance in climbing, jumping, sliding, skipping, and ball throwing; and girls’ in tricycling, galloping, hopping, and ball catching. Certainly, no reason to disturb this balance derives from laboratory measures such as free tapping and peg-moving speeds, which regularly appear in early studies of the mental and physical development of schoolchildren (Gilbert, 1894; Burt, 1909; Bickersteth, 19 17) and have more recently been employed as indices of handedness (Annett, 1970; Barnsley and Rabinovitch, 1970; Bruml, 1972). Tapping and peg moving

In general, tapping speed improves rapidly between 6 and 13 years. Spreen and Gaddes (1969) quote figures of about 200/min up to 300/min for an electric tapper, with the range reduced for the non-dominant hand, and with a manual tapper. There is poor correlation with overall intelligence, though this may be improved with subtests such as Block Design, Comprehension and Information (Boll, 1971). No significant sex differences for pencil and stylus-tapping have been found at 2, 3, 4 and 5 years (Goodenough, 19355; Goodenough & Smart, 1935). The latter did find the girls superior in needle-threading and in the three hole test (thrusting a stylus into each of 3 holes in turn as fast as possible). Though 5 year old girls were superior in various measures of finger tapping, also involving fine musculature, this finds no support from either Kinsboume and McMurray (1975) or Ingram (1975a). The latter also reports no sex difference in the ability to lift individual fingers, in 84 children between 3 and 5 years; interestingly, girls were superior in a more complex task which involved copying hand postures, for which the nonpreferred (L) hand was better. In groups of 5, 6 and 7 year olds, whilst there were no sex differences in ‘repetitive’ tapping of the index finger against thumb, girls were faster at ‘successive’ tapping of each finger against thumb in turn (Denckla, 1973). The right hand was faster for repetitive tapping, whilst there was no hand difference for the successive task. Knights and Moule (1967) report no differences for combined hands/feet in finger/foot tapping from 5 to 14 years. There also appear to be no consistent sex differences in Spreen and Gaddes’ (1969) norms for finger and hand tapping from 6 to 15 years, although there is good evidence for a female advantage in both accuracy and consistency for paced tapping to a metronome over the same age range (Wolff and Hurwitz, 1976). Eckert (1970) found boys and girls performed equally on peg-moving at 3 years and when followed up at 4 years, as did Annett (1970) over the

240 Hugh Fainveather

range 3% to 15 years; but Tiffin and Asher (1948) report a female superiority in adulthood. Boys between 3 and 5 years were better on a ball rolling task, with social reinforcement having no notable effect (Martens, 1970). Girls were consistently superior at moving marbles from 6 to 16 years (Moore, 1937) and though, according to the author, the null hypothesis could be rejected at the 95% level, this was reckoned insignificant, as was a racial time differential averaging 1O%! Kohen-Raz (1965) finds Israeli boys non-significantly better at tapping, dotting, hole punching, over the range 7 to 14 years, whereas Stachnik (1964) finds girls significantly superior at similar tasks at 9 years, but not at 8 or 10. Again, at about 9 years, neither Bolton (1903) nor Bresnahan, Hillard and Shapiro (1973) find any substantial differences, for tapping and peg moving respectively. Bruml (1972) tested 180 black children between 5 and 11 years on a variety of bimanual and unimanual preference tasks including bead-threading, thread-winding, drawing, putting beads in a bottle, picking up a ball, and throwing a ball: there were neither consistent nor significant sex differences. For 360 Ss between 4 and 40 years, Briggs and Tellegen (1971) report no differences for rate of ballistic tapping between two targets, or for the moving of either small or large pegs. Females however were significantly superior on a task involving the pickup of thin nails; and on an armsteadiness task (holding a stylus in a hole without touching the sides: see also Barnsley and Rabinovitch, 1970). For an exclusively adult sample, Lomas and Kimura (1976) find male advantages on tapping and dowel-balancing. Targeting and tracking

Using an alternate targeting task, Connolly, Brown and Bassett (1968) found girls significantly faster at 6, 8 and 10 years, though not more accurate. This was felt to correlate well with girls’ greater ‘cerebral inhibition’ as evidenced in 658 children between 4 and 15 years by their lesser spread of required digit or limb movements to adjacent digits or limbs (Connolly and Stratton, 1968; contra Ingram, 1975a). Generally, when the movement involves grosser musculature (as in the targeting of Connolly et al., 1968; and rotary pursuit) and/or additional spatial co-ordination (as in the Toronto Complex Coordinator (TCC), where a joy-stick controls the trajectory of a green light across a display towards a target red light), one would expect a male superiority (Garai & Scheinfeld, 1968; Buffery & Gray, 1972). Boys certainly tend to excel at throwing (Gesell, 1940; Gesell & Ilg, 1946). In the TCC, males have also been found generally superior at 5, 10 and 20 years (Cook & Shephard, 1958) and in


241

most measures except latency, from 5 to 70 years (Shephard, Abbey & Humphries, 1962). However, on closer inspection, much of the rest of Buffer-y and Gray’s ‘developmental’ evidence, for instance, is seen to be argued by regression from adulthood, and in those studies specifically concerned with childhood is rarely clear cut. Thus in rotary pursuit Davol, Hastings and Klein (1965) found no sex differences in groups from 5 to 9 years; neither did Davol and Breakell (1968) from 7 to 11 years; nor McManis (1965) at 11 to 12 years. This is countered by Ammons, R. B., Alprin and Ammons, C. H. (1955) whose males at 9 years achieve slightly more time on target than girls; this nonsignificant differential is maintained until 15 years, at which point the sexes diverge as males continue to improve whilst females’ performance deteriorates. The importance of practice is stressed by Husband and Ludden (193 l), who found that whilst men were initially superior over 5 trials, the difference was abolished over 30. Jones and Ellis (1962) find boys superior between 14 and 16 years. Although Simensen (1973) finds males of 12.04 years better than girls, this is for racially mixed groups in which the boys heavily outnumber the girls. His claim that this evidence ‘refutes’ that of Davol et al., stems again from the almost universal failure to recognize that the direction of sex differences may change with age. The emerging pattern is that the expected male superiority does not appear until the teens, and then initially at higher rotation speeds. As Davol and Breakell (1968) note: “One can only conclude that later sex differences are a product of bio-social changes which occur during adolescence.” Simple reaction

time

The simple reaction time (SRT) is the time taken for a unique response (typically a manual button press or release) made to a unique stimulus (e.g., flash or tone). Trials are discrete, the inter-stimulus-interval (ISI) being of the order of seconds. This may include a warning signal given at a time before the stimulus referred to as the foreperiod or preparatory interval (PI). Both PI and IS1 may be constant or variable within an experiment. Whilst experimenters have tended to make, and reviewers to collect, assumptions of a continuous male superiority (Hohle, 1967; Garai and Scheinfeld; 1968), the literature is by no means unequivocal on this point. In childhood (see Table 1, from Fait-weather, 1974), age ranges have often been narrow, and boys and girls not always well matched within them; differences in experimental procedure (modality, preparatory interval) make generalization difficult. Such consistent evidence as there is over a

n=6 G n.d. B 6 8 10

Fairweather

and Hutt (1972)

Visual-Manual

IS1 = 2 sets

n=S

G B 6 8

N.M.

(1969)

(1963)

Jones and Benton

Bakker

Hodgkins

(1894)

(1968)

Visual-Manual Auditory-Manual

n=6 Similar results for auditory and tactile modalities (Pers. Commun. 1972) G n.d. B 627 8 9,10,11

IS1 ‘irregular’

Visual-Manual

n = 80 n.d. B

6%(5-8) 12

N.M.

Visual-Manual

n = 50. Less consistent sex differences with discrimination RT B

6-17

N.M.

Gilbert

Visual-Manual

n = 20

(Bl

25, 20)

n = (30,50,

B

5-8

N.M.

Visual-Manual

6-12

N.M.

Visual-Manual

Hill (1971)

Spreen and Gaddes

(1969)

10 test)

n = 50. 6 days practice removed slight absolute male advantage

(5 trials practice,

n = 5, Little practice

n = 25

n = 25

n = 20,24,30,80

Remarks

n.d. (B)

5

N.M.

Auditory-Manual

Pomeroy

(1938)

Sex

(B)

n.d. B B

(G)

B

n.d. (B) B

n.d. (B)

Superior

n = 10. Boys (7 yrs 3m), Girls (5 yrs 4 m)!

a. 46, 7,9,11 b.4-11

8,

(4-l 0) (4-l 0)

PI = 1 to 3 sets

7,9

3-6;

N.M.

b. Tactile-Bite

a. Auditory-Bite

N.M.

3-l 2 12-15

2,3,4,5

Age(s) in years

Visual-Manual Auditory-Manual

and Bianchi

Siegenthaler (1968)

Auditory-Manual

PI regular and irregular l-16 sets (self initiated)

PI/IS1

Bellis (1933)

(1935a)

Goodenough

Auditory-Manual

Auditory-Manual

and Smart (1935)

(1970,1973)

Goodenough

Elliott

Mode (S-R)

Simple reaction time in children (ranked by age of youngest children in a particular study)

Study

Table 1.

KEY:

PI=l,2or3secs

12

n.d.

CR)

n = lOO! Lack of difference remains when followed yearly to 15.

n = 20 n = smaller

n = lOO! (India)

n.d.

B

9-14 15,16

8-l 1

n= 20 Boys 10 ms faster (E 1 yr)

B

7-14

Remarks n = 20

Sex

n.d.

Superior

7,9,11

Age(s) in years

Age: x-y indicates yearly samples, (x-y) indicates range of single sample (B) = differences small, consistent, but statistically insignificant (where mentioned) in favor of boys. B = differences large, consistent, statistically significant (where mentioned) in favor of boys. n.d. = no difference, or inconsistent, or statistically unreliable n = number of subjects, each sex, each year. N.M. = No Mention (of ISI/PI)

Auditory-Manual (lift)

Jones (1937)

>

Manual

Unwarned IS1 = 6-20 (X = 10 sets) Warned PI = l-2 sets)

N.M.

Visual-Manual

Visual, Auditory

(1964)

Philip (1934)

(196 1)

Mathur

Geblewiczowa sets

ISI=l.S,2or3secs (irregular)

PI/IS1

PI = l-2.5

Mode (S-R)

Auditory + Visual Manual (mixed) 1 (lift) Auditory-Manual

~-

Busby and Ilurd (1968)

Study

Table 1 (continued)

244

Hugh Fairweather

substantial range (Elliott, 1973) appears to favor males only after puberty. Fulton and Hubbard (1975) for example, find no sex differences in reaction time between 9 and 17 years, but a male advantage in movement time that gains ground notably during the pubertal period. Even in young adulthood, where findings have tended to look more robust, both for simple reaction speed (Elliott and Louttit, 1948; Clement, 1962) and movement speed (Henry, 196 1) such a sex difference has disappeared with practice (Seashore and Seashore, 1941), or been notable only with threshold or subthreshold stimuli (Botwinick, 1971). Certainly, whatever the difference, it is in real terms exceedingly small. Choice reaction

time

The simplest choice task is that of 2-choice discrimination, only one stimulus being imperative. As for SRT, there is little evidence of a consistent sex difference (e.g., Gilbert, 1894). More recently, and for a 4-choice discrimination task (2 imperative stimuli), Noble, Baker and Jones (1964) note that “the two sexes appear quite similar in performance level and rate of growth until the age of about 16 years, after which the females begin a fairly linear decline (p. 942). For a variety of sorting tasks with children between 7 and 15 years using a pegboard, Whitman (1925) finds “no marked sex differences . . . though at seven of the nine year levels the scores for boys are higher” (p. 121). Pyle (1925) reports girls consistently better at card sorting from 8 to 18 years. More modern choice reaction tasks resolve into two main classes. The classic choice reaction time (CRT) requires unique verbal or manual (keypress) responses typically to each of an array of visual stimuli (lights, numbers), presented randomly and individually. Using the numbers to keys version, Fair-weather and Hutt (1972) and Fair-weather (1974) have found girls significantly faster at each year from 5 to 12, the difference increasing with the number of alternative stimuli (2, 4, 8). Over the same age range, girls have also been found faster at naming colors and random objects, although, curiously, lzot for naming numbers, letters, animals and the uses of various objects (Denckla & Rudel, 1974). Fairweather and Hutt (1976, in submission) argue that there may be something special about the response selection involving choices between individual fingers in the CRT task: the sex difference in speed is not found for those responses which happen to be repeats of the immediately preceding response (see also Fair-weather, 1974, and in preparation). A recent popular variant of the CRT task requires initial memorization of a ‘positive’ set of stimuli (e.g., numbers, figures); subsequently, individual


245

stimuli of the same type which may or may not be members of this positive set are presented for corresponding classification. Unfortunately, in the two studies using this paradigm with groups of boys and girls (Hoving, Morin & Konick, 1970; Silverman, 1974) the relevant statistics have also included the results of adult groups, in whom the direction or degree of a possible sex difference would not necessarily be the same. In Silverman’s study, boys were faster on all responses at 7.7 years, but only on negative responses at 1 1.0 years; girls were consistently faster at 13.9 years. In adults, Husband and Ludden (1931) found females far superior on a 4-choice numbers-to-keys task, remaining so over 500 trials. This contrasts both with the Noble et al. (1964) data for 4-choice discrimination, and for the Simon series of 2-choice auditory RT tasks (e.g., Simon & Rudell, 1967). In the latter, the sex difference may be amplified by incompatibility (Simon, Hinrichs & Craft, 1970), but in a task involving lever movements, there was no sex difference in initial latency, although males did have faster movement times (Simon, 1969). Comparing the performances of men and women pilots on matching instrument information, Drinkwater (1968) found small initial differences (men were faster but made more errors) washed out with practice. I. Q.s: Coding and Clerical Speed

If appropriately and properly collated, data from intelligence tests would provide the single most productive source of material open to investigators of sex differences. However various problems obstruct this provision. First is the sheer multiplicity of tests, not all of which have been standardized by sex, or clearly factor-analyzed. Initial standardizations have often resulted in the specific exclusion of items yielding large sex effects; and subsequent ‘normative’ studies have tended to be limited to abnormal groups (psychiatric in-patients, children with reading problems). Early tests, and/or tests directed principally to an educational audience, have tended to skimp on statistical analysis, and to be idiosyncratic. And last, I.Q. tests have, on occasion, yielded large sex-of-experimenter effects (e.g., Quereshi, 1968; Bradbury, Wright, Walker and Ross, 1975). Which said, what subtest on the various Wechsler scales yielding the most consistent, and often significant sex difference, falls squarely within the perceptual motor domain. This is the Digit Symbol or Coding subtest, and significant female advantages have been demonstrated from 4 years on by Herman (1968) and Yule, Berger, Butler, Newham and Tizard (1969) for the WPPSI equivalent, Animal House; Darley and Winitz (196 lb) and the Scottish Standardization (1967) on the WISC; Matarazzo (1972) for the


WAIS; Quereshi (1968) and Dye and Very (1968) for Perceptual Speed; and Bennett, Seashore and Wesman (1966) for the similar DAT Clerical Speed and Accuracy, amongst adults. Lyle and Johnson (1974) isolate factors on which the sex difference loads with writing speed, and again with paired associates (although this latter test elsewhere spectacularly fails to reveal a sex difference: M & J Table 2.1 I). As always, however, there are riders: Koch (1954) finds girls better on Perceptual Speed, but only those girls with closely-spaced sibs; similarly, Laosa and Brophy (1972) find the difference confined to first-borns at 5 to 6 years. And amongst adults for WAIS Coding, there are also exceptions: Sarason and Minard, (1962); Jacobson, Berger and Millham (1970). In sum, the evidence from which the distinction gross-fine has been distilled to characterize male-female differences in motor skills is weak, and weaker still in childhood. The distinction is echoed once again in the conclusions of the adult normative studies of Barnsley and Rabinovitch (1970), though the distribution of significant differences (Table I, p. 358) is ample refutation. Here, as elsewhere, the often minute number of test trials would raise a blush in most experimental psychologists.

Motor skills and laterality Hand preference

For an ethnically-mixed sample of 100 eight-month-old infants, Cohen (1966) reports that 43% have right hand preference for reaching, but he finds no sex difference. The protocols of Gesell and Ames (1947) would in any case seem to indicate that hand-preference is not firmly established until a much later date, possibly as late as 8 years. HiIdreth (1949) summarizes seven studies of handedness in schoolchildren: whilst there are generally very slightly more boy lefthanders (cf., Clark, 1970, for a large Scottish sample at 7 years), she attributes this to “outdoor and sports activities” demanding more bimanual activity, whereas “delicate motor skills . .. favor the development of right-handedness”. In parenthesis, Ingram (1975a) found not only that boys were stronger at 3 to 5 years, but that their hand strength scores were also significantly less asymmetrical. Social rather than genetic factors are also emphasized by Suchenwirth (1969) investigating 1920 children between 2% and 6 years, amongst whom there were in fact more extremely right-handed and ambidextrous boys, there being no sex differences in incidence of left-handedness. Since the proportion of ambidextrous individuals as a whole was much higher than in

Sex differences

in cognition

247

adult groups, Suchenwirth concludes that handedness is largely culturallyacquired (see also Kretz, Suchenwirth & Ferner, 1970). The method of measuring preference is, of course, always problematic, and Suchenwirth admits that his (including card-mixing, tie-tying and various tool using tasks) may well be culture-biased. Failures to find sex differences with other batteries of tests in childhood are reported by Belmont and Birch (1963). For adults, Briggs and Nebes (1975) report that “No sex differences are immediately apparent from inspection of these distributions” although the figure they refer to clearly indicates twice as many females scoring at the very extreme of right-handedness. This is a common finding, and owes more to a male preponderance amongst ambidex trals rather than amongst sinistrals (Newcombe, Ratcliff, Carrivick, Hiorns, Harrison and Gibson, 1975; Thompson and Marsh, 1976). The oft-mooted relation between hand preference and intellectual abilities finds no support in studies of large groups, and certainly not in the direction of poorer performance on the part of left- or mixed-handers (Newcombe et al., op. cit.). Hundedness

forskills

Ingram (1975a) reports no significant sex/hand interactions for finger tapping, lifting and spacing, nor for hand postures, in 3, 4 and 5 year olds. The right hand was significantly superior for finger tapping and lifting, but the left hand for spacing and posturing. The finding for finger lifting runs counter to the considerable evidence from Connolly and Stratton (1968) which shows an enhanced female superiority for the non-dominant hand, in 4 to 15 year olds. On the other hand, females also show a more pronounced dextrality for peg-moving in both children (Annett, 1970) and adults (Tiffin and Asher, 1948). In adult right-handers, single or paired finger flexions are performed better by the left hand; left handers produced no consistent hand differences (Kimura & Vanderwolf, 1970). The authors suggest that “the neural mechanism for hand preference is not based on an asymmetry in the control of fine movement” (p. 773) but on “performance in a rapid sequence” (p. 772). A subsequent study (Lomas & Kimura, 1976) finds better right hand performance for dowel balancing, sequential finger and arm tapping, and single finger tapping (the tapping tasks using a key as response mechanism). There are two overall sex differences - males balance a dowel longer, whereas females are faster at finger tapping. However there is little evidence of sex/hand interactions; the authors report no interpretable interactions for the sequential finger tapping, and unfortunately collapse the data over sexes for the two tasks not yielding an overall sex effect. For the dowel-balancing

248

Hugh Fairweather

task, concurrent verbal activity interferes for males, but not females, and particularly for the right hand, where both silent and spoken verbal activity disrupt balancing. Clinical researchers seem hesitant to locate motor skill primarily in the left hemisphere, the right, or somewhere between. Wyke (197 la) found two reasons for distinguishing between various hand-tapping tasks and the Purdue peg-board: (a) controls tapped better on all occasions except for right lesionright hand; and (b) subcortical bilateral invasion further impaired performance on the tapping but not the peg-moving. Results for this latter task compare well with those for the Pursuit Rotor, which, it may be noted, loads heavily on a ‘Manual Dexterity’ factor. The right hand tends always to be better, but right and left lesions impair equally (Heap and Wyke, 1972). One clear and somewhat surprising piece of evidence is that for a bimanual star-tracing task, left lesions impair, but right do not (Wyke, 1971 b). Such a finding clearly does not fit with Buffery’s notion of bilaterally-organized skills, nor does the general tendency for left lesions to be more disruptive (see also Kimura and Archibald, 1974, for copying hand movements). Specifically, lack of data on the effects of bilateral lesions, across sexes, renders interpretation uncertain. For the one test yielding a consistent sex difference, and that in favor of females for Digit Symbol or Coding, the two hemispheres appear equi-potential (McFie, 1960). Thus far, neither the normal nor the clinical literature on sex differences in motor skills would tempt any explanatory recourse to the cerebral hemispheres.

Spatial skills MacFarlane Smith (1964) is an oft-quoted reference work for studies in spatial ability. Five references to studies concerning sex differences are made, none of these concerning children under the age of 14 years. The first two (p. 122-123) report boys’ superiority in geometry at high school. The third (p. 209) reveals that 14 year old boys outscore girls on drawing tests if proportionality is the measure, but that this sex difference is reversed if accuracy is measured. The fourth (p. 235) reports that a spatial factor correlates with gregariousness in men but not women. The fifth (p. 255) actually does not consider men and women as separate groups though here MacFarlane Smith manages to report “a marked sex difference” in favor of males; whilst simultaneously failing to quote his own study (1948) of Scottish schoolchildren from 12 to 14 years, which did find an overall difference in favor of boys. Nine tests included some essentially similar to


249

those now called Block Design, and embedded figures, and others concerning the matching and classifying of sizes and shapes of objects. Elsewhere may be found the following remarks: “... the well-known fact that girls are inferior to boys in visualizing ability” (P- 123); “ . . . it is well known that tests with high spatial loading show a marked sex difference in favor of boys” (p. 2 10); and “ . .. men doing better than women as nearly always occurs with spatial tests” (p. 255). To none of these statements is there attached even a single reference: stuff indeed to make a myth. Similar statements may be found in Tyler (1956, p. 253) who spotlights performance on the form board as an index of the stereotype, even though Sylvester (1913) had much earlier concluded in his classic monograph: “Sex differences are of no importance in the form board test.” (p. 38). Five hundred children from 5 to 14 years were tested and although boys have a marginal 0.4s advantage, this is compensated for by the fact that they make more errors. Garrett, Bryan and Per1 (1935) find girls superior at 9 and 12 years, and boys at 15; whereas for Reichard (1944) boys have the advantage at all three ages, none of the differences in either study approaching statistical reliability. Even with thus brief an introduction it is clear that the term ‘spatial’ has been used to connote a motley collection of skills, and it is perhaps this very quality that renders the myth, if so it be, so difficult to dislodge. Accordingly, we shall look first to where the evidence is purest, to items on IQ tests, before turning to other tasks where the operation of a spatial factor, though invoked, is less definitive.

1.Q. tests On Wechsler Performance subscales Darley and Winitz (196lb) at 5 years, and Brown and Bryan (1955) between 25 and 39 years, report female advantages. They may be considered rare exceptions perhaps largely promoted by the female advantage in the Coding or Digit Symbol subtest (see section on Motor Skills). Newcombe et al., (1975) report a significant male superiority in a large adult sample average age 40 years, but for a threeitem composite only (Block Design, Object Assembly, Digit Symbol). The recent standardization of WISC-R (Kaufman and Doppelt, 1976) reveals almost perfect parity between the sexes across the range 6 to 16 years. Significant subtest differences are confined, in childhood, to male advantages on Picture Completion and Block Design, and then only at 13 years, on

250

Hugh Fairweather

the Scottish standardization of the WISC (1967), which is replicated precisely in adult scores on the WAIS (Matarazzo, 1972). Cube counting, it may be noted, is a test excluded from all Wechsler scales; and similarly Mazes only appears on the WPPSI, where it generates large male superiorities at 4 to 6 years (Fan-weather and Butterworth, 1977; Wilson, 1975; Yule et al., 1969). The difference on maze performance more generally is also found in adults (Davies, 1965) but has been found to disappear in children: with practice at 3 to 5 years (McGinnis, 1929), and with any kind of reinforcement schedule at 10 and 11 years (Wright, 1968). The American standardization of the WPPSI (Herman, 1968) reiterates the finding for Mazes, and also reports that girls score higher on Geometric Design and Block Design, though here the very large numbers involved (1,200 children between 4 and 6% years) increase the possibility that a small and elsewhere insignificant difference will attain statistical respectability. Koch (1954) finds no sex differences on PMA Space amongst 360 fiveand six-year olds balanced over birth orders and sibling age spacings. Emergence of a male superiority on the defined spatial scales appears to coincide almost exactly with adolescence (Meyer and Bendig, 1961; Bennett et al., 1966, for DAT Mechanical Reasoning and Space Relations). The failure of DAT Abstract Reasoning to generate a difference is notable since the discriminanda are of a fundamentally spatial nature (cf., similar findings with the essentially equivalent Raven’s Matrices: Coie and Dorval, 1973; Wilson, 1972). The male advantage appears confined to tasks involving transformations of visual stimuli, and particularly in three dimensions. Tasks lacking either the operational or dimensional component tend to produce much less convincing evidence (embedded figures, matrices, drawing, etc.). Perhaps the most convincing catalogue of post-adolescent male superiorities is contained in a recent report on 2508 San Francisco high school students (Yen, 1975). Highly significant differences were obtained for: PMA Spatial Relations; Form Board Test; Mental Rotations Test; and, less strikingly, Paper Folding. A considerable proportion of this superiority is attributable to the fact that significant improvements with age are found only amongst males. There is however no simple attribution to the influence of a sex-linked major-gene as assessed by sibling correlations. Druwing

and Copying

There is little evidence that one sex or the other excels consistently in drawing and copying abilities. On the Draw-a-Person task, Keogh (1972) reports a male superiority at 4 and 5 years, whereas Ford, Stern and Dillon (1974) find the opposite over a similar age range (see also Corah, 1965).

Sex differences

in cognition

25 1

Girls may be better on the ‘Incomplete Man’ task at 5 to 9 years (Ames & Ilg, 1964) but have only marginal advantages in a variety of tasks for 8 10 children at 8 to 9 years (Brenner & Gillman, 1966). Schiller (1934) had much earlier reported a reliable superiority for girls on the Goodenough Draw-a-Man test at 9 years. Keogh (197 l), however, reports no significant differences again at 9 years, whereas boys have been found better on a star tracing task at the same age (Ong and Rodman, 1972). Struempfer ( 197 1) finds no differences for figure drawing in 11 year olds, and emphasizes the problem of rating. In an intriguing pair of experiments, Keogh (1971, 1972) provides the clue that the expected male superiority in tasks with a high spatial-visualizing content may appear if one takes the drawing or copying into a somewhat larger arena. Thus, a male superiority on pattern copying appears if the patterns are walked instead of drawn, and particularly if walked in sand, providing additional visual information. This laboratory simulation of a field situation may be seen to correspond to earlier evidence of boys’ superiority in defining compass directions, indicating the relative positions of sites in the community, and maintaining orientation whilst travelling (Lord, 1941). Lord, however, also points out that “The boys respond more frequently than do the girls, and, taken as a group, are superior. In most cases, however, the girls who do respond are nearly as accurate as the boys” (p. 504). Awareness

of body parts

There is no evidence of sex differences in finger localization (Benton, 1959) or in the ability to locate a fixation point on a table from underneath, i.e., when the locating hand is hidden (Sandstrom and Lundberg, 1956; Smothergill, 1973). In terms of lateral awareness (right-left discrimination) there is also little evidence of a sex difference (Benton, 1959; Keogh, 1972; Long and Looft, 1972) though Albert (1975) has most recently demonstrated an adult female superiority for a task involving pointing to ownbody parts (pro Croxen and Lytton, 1971, for children; contra Bakan and Putnam, 1974, for adults’ naming of body parts). Fairweather (1975) failed to find differences in the speed of tachistoscopic matching of the position of a dot relative to the word ‘right’ or ‘left’. Field dependence

Much of the evidence of sex differences in spatial skills has centred on the results of various tests of ‘perceptual field dependence’, including the Embedded Figures Test (EFT) in which simple patterns are required to

252

Hugh Fairweather

be isolated from complex backgrounds; and the Rod-and-Frame Test (RFT) in which a rod is required to be adjusted to the vertical within a tilted frame. The oft-assumed continuous male superiority (e.g., Buffery and Gray, 1972; Hutt, 1972a) clearly does not in fact manifest itself until late childhood (M & J: Table 3.8). Hereafter there are both exceptions (Mayo and Bell, 1972; Deich and Hodges, 1973; Sherman, 1974*); and confoundings, with culture and more specific visuo-spatial abilities. In cultures where females are encouraged to be more exploratory, e.g., Eskimos, sex differences in field dependence disappear (Berry, 1966; MacArthur, 1967). By the same token, where females adopt a particularly submissive role, e.g., India, the sex difference in favor of males is accentuated (Pande, 1970), although, curiously, Indian women have been found less field dependent than American men (Parlee and Rajagopal, 1974). Sex differences o&y appear in cultures typed as highly stratified, i.e., high food-accumulating, non-hunting ‘Western’ cultures (Berry and Annis, 1974). Within such a culture Saarni (1973) for instance finds males less field dependent than females in groups of children between 10 and 15 years. Paradoxically, easily the most field dependent group were those girls rated as the most cognitively mature. The explanation was that these girls were also the most sensitive to sex-role expectations (see Vroegh, 197 1). Perhaps more importantly, sex differences tend to disappear with practice (Change and Goldstein, 1971) and/or be subsumed under a more powerful and more clearly spatial factor (Sherman, 1967; 1974). Thus Wolf (1971) using a modification of Witkin’s test in which no simple figures recur, thereby confining performance to the short term, found no sex differences in children between 6 and 12 years of age (though failing to discuss the possibility of sex differences in short term memory itself). Wolf did find males superior in a separate group at 17 years, and this difference remained significant after analysis of covariance had removed the effects of overall IQ and of vocabulary. However covarying out scores on the GuilfordZimmerman Spatial Perceptual Test rtmoved the sex differences entirely. Conservation Whilst almost all conservation tasks involve the inspection of visual material, not all make great demands on the ability to manipulate that material spatially, perhaps only those of substance, quantity, and three-dimensional *Perhaps not so surprising in this last case: Sherman (p. 1223) acknowledges that the “Women’s Research Institute of Wisconsin were helpful particularly in making available a pool of Ss”. Exeat male chauvinism along with spatial superiority?


253

space. In addition, the tasks were exclusively designed for use with preadolescents. Maccoby and Jacklin(Table 3.11) list over thirty studies of various kinds of conservation: about one-third find evidence of sex differences with a small majority favouring males. Recently Brekke and Williams (1973) find girls at six years significantly better at conservation of substance, noting that the sex of experimenter (female in their case) may have provided a crucial difference. This echoes earlier findings by Bittner and Shinedling (1968) who found girls superior at conservation of substance in grade 1, but that this difference was reversed in grade 3 where it also interacted significantly with sex of experimenter. Using the same task Fogelman (1970) also found girls superior at 7 years, but only for a ‘passive’ condition including more verbal cues; there were no differences for an ‘active’ condition allowing manipulation of the plasticine. Amongst related Piagetian tasks, Willemsen (1974) finds no difference in judgments of the vertical in 6 to 8 year olds; and Coie and Dorval (1973) find boys superior on spatial perspectives between 8 and 10 years. Notwithstanding the original designation of the tasks, a considerable number of both high school students (Elkind, 1961) and adults (Rowell and Renner, 1976) fail to conserve volume. A significant majority of females amongst this number is attributed, at least by Elkind (1961), to sex role differences.

Laterality

and spatial skills

Buffer-y and Gray (1972) propose bilateral cerebral representation of spatial skills in males, as opposed to a less appropriate unilateral (right-sided) representation in females. This would predict parity across the hemispheres for the effects of unilateral lesions on spatial abilities in males, as opposed to maximal disruption by right-sided lesions in females and/or bilateral invasion in males. Thus far, the theory is not well-supported by the experimental data. As might be expected from the previous section the most entertaining possibilities for a laterality theory emerge from IQ testing in clinical groups. In Lansdell’s (1968) series, for instance, men with right- rather than leftsided operations scored lower on a composite measure of ‘non-verbal’ items (including, be it noted, Arithmetic, a renegade from the WAIS Verbal subscale, and Digit Span, both tests which can lay good claim to a left-sided location - see, for example, McFie, 1960); whereas the reverse was to a lesser extent true of females. McGlone and Kertesz (1973) found that right


hemisphere males distinguished themselves (albeit in fact insignificantly) from other lesion groups by their poor Block Design performance, a test for which there is certainly good evidence for a primary right hemisphere involvement (Taylor and Warrington, 1973). There was no such sex/hemisphere trend on a language test, left lesion patients of both sexes performing consistently worse. Only for left lesion females did both the visuo-spatial and language scores correlate, suggesting more verbal mediation in ‘nonverbal’ tasks by females, and the sort of female ambilaterality noted by Lansdell. McGlone and Kertesz emphasize again that this tendency is often maladaptive: unpublished evidence shows that females lateralized ‘the wrong way’ for dot enumeration are worse on more general visuo-spatial tests. Buffery and Gray point out quite rightly that many clinical reports may be difficult to interpret since subjects were not used or able to be used as their own controls by pre-lesion or pre-operative testing. The original epilepsy or damage may well have led to the re-location of functions and it may be the facility for this that differs between the sexes, and not location per se. They may take little heart from the way in which the clinical data begins to dovetail reports on information processing in the intact brain. One of Annett’s (1973) most surprising findings was that familial sinistrality tends to increase the severity of language disability in all groups except left hemiplegic females. The speculation that this involves females with a potentially dominant right hemisphere compares interestingly with the report from McGlone and Davidson (1973). This study on dot-enumeration in normal adults revealed a clear majority of superior scores following left field presentation amongst males, as opposed to an equal distribution of such scores across fields amongst females. The additional finding that only lefthanded females with higher left ear scores for reporting dichotic digits were worse than males on PMA Space (although not WAIS Block Design) suggests that females are more ambilateral in the specific sense and that there are more individual females with their cognitive functions in the ‘inappropriate’ hemisphere. The rider to this is that having visuospatial functions in the left is more disadvantageous than language in the right. This view is diametrically opposed to that of Buffery and Gray (1972), who assert that visuo-spatial skills improve with bilateral representation. Earlier findings by Kimura (1969) for dot localization had similarly indicated a sex/hemisphere interaction: in two of four experiments scores were equal across hemispheres for females but showed clear right hemisphere advantages for males (using circular rather than square pre-exposure fields removed this interaction in two further experiments ~- field articulation?). However neither the overall laterality effect nor the interaction with


255

sex have found support in recent replications (Bryden, 1973; 1976). Bryden (1973) additionally failed to find any significant influence of either handedness or familial sinistrality, but did note that a right field superiority emerged on the second half of (a very small number of) trials. Neither the presence of a frame of reference nor the location of the response card were found to be relevant controls in Bryden (1976). A detection paradigm indicated a right hemisphere bias for false alarms (see also Dimond and Beaumont, 1973), and a tendency for females to be poorer detectors and to produce more false alarms. However, much more substantial studies than those presently in vogue, both in terms of subject numbers and methodology, would be needed for such possibilities to generate much more than their present passing interest. Elsewhere amongst the motley collection the evidence is both sparse and inconclusive: hemispheric ascription is either unclear or inappropriate, and few experiments have compared the sexes. Ratcliff and Newcombe (1973), for male missile wounded patients, found that only bilateral posterior damage impaired performance on a locomotor maze (essentially a two-dimensional, map-reading-and-following task involving whole-body orientation) but that right posterior cases were also impaired on a stylus maze (similar, but lacking the orientational component), and cube counting. Ratcliff (1970) had earlier failed to show impairment on a right-left judgment task in which the stimulus, a manikin, changed orientation in only two dimensions; changes in three dimensions, however, selectively produced deficits with right posterior damage. Fairweather (1975) similarly failed to show hemispheric differences in reaction times for right-left judgments, but noted a tendency for males to perform better in the right visual field, females in the left. Although Ong and Rodman (1972) find no interaction with handedness for performance on a star-tracing task, Buffery (see Buffery & Gray, 1972) does, using the more sensitive ‘Conflict Drawing Test’ (CDT). This technique requires simultaneous drawing with both hands ~ a circle with one, and a square with another. For a large sample of children between 3 and 11 years, the CDT revealed that girls had a greater tendency to draw better squares with their left hand. The degree of this superiority increased with age for both sexes, though not appearing in boys until 7 years. The sex difference was less marked in older than in younger children, and in slight qualification it may also be noted that Buffery’s females also showed a much greater degree of right-handedness than the boys. And lastly, embedded figures appears to be a ‘spatial’ skill for which the left hemisphere dominates. Cohen, Berent and Silverman (1973) investigated 36 female, right-handed, adult depressives: left-sided ECT produced greater

256

Hugh Fainveather

field independence, whereas right-sided ECT worsened performance. The result was explained by assuming that ECT attenuated responsiveness to the more peripheral parts of the stimulus field, which in turn may be the more specific spatial prerogative of the right hemisphere. Similarly, Pizzamiglio (1974) has found greater field independence in those Ss with evidence of greater left-sided laterality: right handers, and those with a marked right ear superiority on dichotic digits. There is, tinally, good evidence of a clear, adult, male superiority for a small nucleus of definitive spatial skills. The qualifications are important: there are rarely significant findings in childhood, and differences tend to increase thereafter; and though the kernel of definitive skills may generate differences in related tasks they are everywhere more dilute and less decisive. The question turns to the aetiology of this difference. There is little doubt that high spatial performance is strongly influenced by a sex-linked major-gene, but attribution of the sex difference to such an influence is neither simple nor complete (Bock and Kolakowski, 1973; Yen, 1975). Bock and Kolakowski conclude that since “... sex differences in spatial ability are difficult or impossible to observe in pre-pubertal children.... The possibility that spatial ability is testerosterone-limited as well as sex-limited is clearly an area for further study.” Subsequent correlative studies between either observed or induced androgen levels and cognitive performance can only, and even then marginally, be explained on the assumption that spatial ability is optimal at average levels of androgen, an explanation stated independently of sex differences per se (Bock, 1973). Certainly non-genetic, presumably environmental, factors also influence performance (Yen, 1975). Culture-based explanations have always had good intuititive evolutionary appeal. Man is supposed to attain spatial skills by virtue of being the prototype hunter, or food gatherer. This notion has even received the tine-grain accolade of a direct correlation between spatial ability (copying patterns; geometric figures; mazes) and the distance covered from home at herding time. However, the correlation holds equally for both sexes: girls who travelled more than boys had similarly higher spatial scores (Nerlove, Munroe and Munroe, 197 1).

Linguistic

skills

Development Maccoby and Jacklin (I 975, p. 75ff.) find little reason to suppose that there are any essential differences in either the development or expression of


257

linguistic skill before adolescence. The historic possibility that girls acquire sounds and words very slightly earlier than boys remains unchallenged by (the lack of) recent normative studies. Irwin and Chen (1946) for instance plotted the curves of acquisition of phonemic types in infancy: the number rose from 7 in the first 2 months to 27 at 2% years. Although the theoretical curves derived separately for the two sexes reveal a tendency for girls to outstrip boys from 18 months on, only two of the six comparisons in the last year are significant: as the authors note (p. 434) “This is not enough evidence upon which to base an assertion of the definite presence of sex differences”. Besides, given that the adult ceiling of phonemic types is 35, such a difference would indeed be short-lived. Morley (1957) investigated a large group of English children, and found no significant sex differences in the ages at which the first word, and the first 2-3 word phase appeared; whilst girls did evince ‘intelligible speech’ earlier than boys, any difference in this respect had disappeared by age 4. This study is one of four reviewed by Darley and Winitz (196 la) who conclude (p. 284): “At present there is little evidence that girls begin to speak earlier than boys as measured by the age of appearance of the first word”. Girls have more recently been shown ahead at 2% but not 2 years in the acquisition of consonants (Paynter and Petty, 1974). In his excellent review of articulatory development, though, Winitz (1969) summarizes thus the results from 11 studies in the period 193 l-1965: “ . . . girls surpass boys more times than boys surpass girls (see Table 3.3). However, very few of the differences are significant.... How can we explain, then, the fact that a greater proportion of elementary school males are most often reported as articulator-y defective ?” The possibility of a greater range of performance among males is dismissed (p. 150): “This conclusion is probably not true, since the variances for boys and girls are very similar. No further explanations for the generally accepted fact that young females outstrip their male companions in articulator-y development can be offered.” Mothers certainly report significantly more boys not talking at 2 years (7.6% against 4.8%: Pringle et al., 1966). Also significant are later differences in the incidence of speech difficulties (1 1.5% against 8.5%) and unintelligible speech (16.2% against 1 1.4%). Whilst significantly more boys are reported to stammer (7.9% against 4.5%) this difference is not observed during specific clinical examination (1.3% against 0.8%). In experimental studies of retarded readers, or dyslexia, one invariably finds a vast majority of boys amongst the sample (e.g., Critchley, 1970). Two extensive surveys at age 10 in England, however, make an important definitional point: sex ratios of 3.3 and 3.6 to 1 (boys:girls) for reading retardation (assessed in relation to

258 Hugh Fainveather

reading scores predicted from individual 1.Q.s) reduce to 1.3 and 1.4 to 1 respectively for reading backwardness (assessed in relation to population norms) (Berger, Yule and Rutter, 1975). Reading Reading is thought to be a lynchpin citation with respect to sex differences in verbal skills. In a fine review, Thompson (1975) points to several factors obscuring this conclusion. Primary among these is a cultural factor: by far the majority of studies finding significant differences originate in the U.S.A.; in England, and notably Scotland, such differences are hard to find. To Thompson’s list of British studies finding no differences on reading-related tasks, we may add Phillips and Bannon (1965); Pidgeon (1960); Ross and Simpson (1971); and Wilson (1972) for various groups between 8 and 15 years. A European review (Malmquist, 1970) also quotes two Scandinavian studies showing non-significant sex differences in primary schools. Keogh and Smith (1967), following a small group of American children between 8 and 12 years, find no differences on reading and spelling at either age. The incidence of significant findings tends to decrease after the age of 10 years even in the U.S. (pro: Moore, 1940; Broom, 1941; contra Glass, Neulinger and Brim, 1974). Thompson rejects the possible explanation of sex-typing of the early school environment (predominantly feminine) since this fails to promote equal sex differences in other subjects (e.g., arithmetic). Recent evidence is more persuasive. Asher and Gottmann (1973) report a highly significant female superiority for reading comprehension in two large groups at fifth grade followed to sixth grade (N’s = 534, 712). Interestingly, sex of teacher during the intervening year did not interact with reading development across sexes. In an intelligent small-scale follow-up, Asher and Markell (1974), noting evidence that both boys and girls perceive reading as a feminine activity, investigated interest areas in their fifth grade Ss. Subsequent testing of the comprehension of either high- or low-interest material revealed that the expected sex difference was limited to the low-interest material. Boys were apparently poorly motivated on material in which they had whereas girls performed equally across the board: a little interest, particularly neat example of the intrusion of sex-role into the achievement domain. A further and perhaps more conclusive example is provided by Dwyer (1974) who performed a multiple regression analysis for factors including IQ, sex role standards and sex role preference for 385 children between the ages of 7 and 18 years (second to twelfth grade). She concludes (p. 8 11):


“The results suggest that reading and a function of the child’s perception sex-inappropriate than of the child’s for masculine or feminine sex role, arithmetic.”

259

arithmetic sex differences are more of these areas as sex-appropriate or biological sex, individual preference or liking or disliking of reading or

I.Q. tests In accordance with the standardizations, significant sex differences are rarely reported for Verbal subscales; where they are, they favor males. Thus, boys have an overall 2% point advantage on WISC-R Verbal, a difference “though significant in a statistical sense, not meaningful in a practical sense” (Kaufman and Doppelt, 1976). Brown and Bryan (I 955) for the Wechsler-Bellevue at 9 to 11, and 25 to 39 years (but not between); and Newcombe et al., (1975) for WAIS Verbal in adults, find similar male advantages. Koch (1954) also found a male superiority on PMA Verbal at 5 and 6 years, but only for first borns widely spaced in age. With regard to subtests, the WPPSI has yielded only a female advantage on Sentences for Herman (1968). For the WISC, Darley and Winitz (196 lb) find girls superior on Similarities at 5 years; the Scottish Standardization (1967) finds boys outscore girls on Information at both 9 and 13 years, on Comprehension only at 9, and on Vocabulary only at 13 years. The WAIS American norms find significant advantages for males on Information, Comprehension and Arithmetic; for females on Vocabulary (see also Shaw, 1965). Arithmetic, or more generally quantitative ability, is indeed a noted bastion of male superiority (M & J: Table 3.5). There are again notable exceptions, even in adolescence and beyond (Meyer and Bendig, 1961; Bennett et al., 1966; Dye and Very, 1968). Pringle et al., (1966) report only very marginal differences favoring boys in a problem test involving numbers at 7 years; likewise Emmett’s (1954) data indicate no difference between 9 and 11 years. Similarly Pidgeon (1960) finds no difference on Problem Arithmetic at 10, but boys significantly better by 14 years. However Pidgeon also reports girls to be significantly better at Mechanical Arithmetic at both 7 and 10 years (although this is contradicted by Wilson, 1972, at the latter age), but not at 14 years, and notes the comparative falling-off of female ability upon entering secondary school. Phillips and Bannon (1968) find boys much better in small groups at 11 years on a local English mathematics test. Laboratory studies indicate no differential development of number concepts in 3 and 4 year olds (Estes & Combs, 1966; Siegel, 1971). Pederson, Shinedling and Johnson (1968) find a male advantage with male


examiners, and a female advantage with female examiners, for small groups of 7 year olds on WISC Arithmetic. Maccoby and Jacklin (Ch. 2) list various tasks using verbal material which fail to elicit sex differences: paired associate learning; discrimination learning; and memory for digits, although not for words, where females tend to have the edge (e.g., May and Hutt, 1974, for auditorily but not visually presented words at 8 to 10 years). Other recent failures to show sex differences in memory include: Goldberg, Perlmutter and Myers (1974) at 2 years; Marx (1972) and Shapiro and Moely (1971) in later childhood. Laboratory

studies

Thompson (1975) notes the effects of group vs. individual testing, the latter invariably tending to yield modulated differences. Thus laboratory studies of cross-modal matching (Reilly, 1971; Bryden, 1972) or the reproduction of sound sequences in a visual display (Zaner, Levee & Guinta, 1968; Calfee, Lindamood & Lindamood, 1973), often held to underpin reading ability, have all failed to demonstrate sex differences. Of nine tests on the Illinois Test of Psycholinguistic Abilities (ITPA) only one yields a significant sex difference in a British sample of 4 year olds (Mittler & Ward, 1970). Girls are better on Visual-Motor Association, a test measuring ability to relate pictures of objects to each other on the basis of their common uses. Other studies in the pre-school period fail to find sex differences in syntactic/conceptual comprehension, in tasks requiring matching pictures to words or sentences (DiVesta & Stauber, 1971; Parisi, 197 1). Weener (197 1) finds no evidence that the syntactic structure of verbal messages influences recall selectively across sexes in 5, 6, 7 and 8 year olds. Paris (1973) finds no sex differences in logical comprehension in 7 and 11 year olds. Boys between 4 and 11 years have better signal-noise ratios for the perception of speech in noise, significantly so in the 11 year old group (Siegenthaler and Barr, 1967); whereas girls excel at 6, but not at 7, 8 and 9 years, for comprehension of sentences of varying difficulty heard at various speaking rates (Nelson, 1976). In terms of smaller linguistic units, Iversen, Silberberg and Silberberg (1970), for instance, found 5 year old girls had better knowledge of letters but not numbers. Girls also better recognized pronounceable/unpronounceable trigrams presented tachistoscopically, at 7 and 9 years (Gibson, Osser & Pick, 1973), but not words in two hemifield studies also in childhood (For-gays, 1953; Olson, 1973). In the auditory mode, there are sex differences neither in the recall of words presented dichotically at 3,4, 5 and 6 years (Nagafuchi, 1970) and at 7, 10, 14 and 17 years (Friedrich, 1974),

Sex differences

in cognition 26 1

nor in the recall of consonant-vowel pairs between 5 and 13 years (Berlin, Hughes, Lowe-Bell & Berlin, 1973). There is also little evidence that one sex excels in the recall of dichotic digits: whilst Kimura (1963) found girls significantly better at 5 and 6 years out of the six yearly groups from 4-9 years, this has been replicated neither by Knox and Kimura (1970), nor Geffner and Hochberg (197 1).

Where is the difference? in what, then, lies the Maccoby and Jacklin conviction that there are female superiorities in verbal skills? Principally, one supposes, in Table 3.4, which lists the six most sizeable studies. Of the five also revealing sizeable differences there are two American and one Swedish study, pro hypothesi, balanced by one English study, contra hypothesi in the reading (vocabulary) comprehension ambit. Precise statistical significances go unmentioned in the American studies; and the Swedish study finds a female superiority on one verbal test but not another (‘Opposites’); and in a second study, in a comprehensive but not an elementary school. From the data reviewed earlier, it does indeed seem unlikely that this ambit would yield big sex effects unless they are culturally linked, which leaves Droege (1967) alone, testifying to the superiority of American females immediately postadolescence, on a wide variety of verbal tests. Elsewhere, females at the same age show advantages on DAT Language Usage (spelling and grammar: Bennett, et al., 1966); on PMA Verbal Meaning, Reasoning and Fluency (Meyer and Bendig, 1961) and the Guilford Fluency tests (Olive, 1972; M & J: Table 3.13). However, they do not excel on the DAT Verbal Reasoning, nor, always, on the other and various tests of reasoning which may load on a verbal factor (Dye and Very, 1968; M & J: Table 3.12). Color naming, held to be a close relation of verbal fluency, tends also to yield female advantages in both children (Ligon, 1932) and adults (Golden, 1974, for group testing) which are occasionally highly significant (Broverman and Klaiber, 1969), but also reduce with practice (Stroop, 1935). Data for reading the classic ‘Stroop’ stimuli (color words printed in confounding colors), originally yielding no difference (Stroop, 1935) have more recently matched both the tendency for female superiority (Peretti, 1969) and its sometimes significance (Golden, 1974; Peretti, 1971, the effect in this latter being accentuated under ‘competitive’ instructions). Curiously, reading color words printed in black never gives a significant sex difference (Ligon, 1932; Golden, 1974; Stroop, 1935).


Viewed critically, then, the evidence is not compelling. Is it coincidence, for example, that those studies showing female advantages on the Remote Association Test (M & J : Table 3.13) are also the smallest?*

Laterality

and linguistic skills

Taclzistoscopic

word recognition

Forgays (1953) found no overall sex differences in word recognition in children, but failed to make specific mention of a possible sex/laterality interaction. A significant laterality effect (better recognition in the right visual field) did not appear until about 12 years (see also Miller and Turner, 1973). The interpretation of laterality effects in terms of left-right scanning patterns acquired through reading (rather than hemispheric differences per se) is challenged by McKeever and Huling (1970) who find clear effects in both good and poor readers at the same age, using shorter exposures, and requiring report of a number at the central fixation point, to guard against eye movements. Turner and Miller (1975) offer partial resolution: clear right field advantages from 6 years for three-letter word recognition (multiple choice format) with very short exposures, disappear with identification responses, necessitating much longer exposure times; but reappear with five-letter words. Sex was nowhere a significant factor. Olson (1973) demonstrated a right field advantage in good readers from 7 years on, but not until 10 years for those children reading six months below their expected grade level. There were no sex differences. It is interesting to note the hint of bimodality in the distribution of visual field superiorities, and tempting to speculate that one sex might turn out to have a majority amongst the small number of ‘left fielders’ in a Iarger sample (cf., McGlone and Davidson, 1973). Poor readers have also been shown to be less well-lateralized than normal readers for word recognition between 7.06 and 8.07 years (Marcel, Katz & Smith, 1974). Boys were significantly better lateralized than the girls. Since this fails to match the BufferyGray formulations, Marcel et al., suppose that the matching by reading performance causes sampling from the higher end of the lateralization scale in boys, given that boys are poorer readers in general. This is a strange argument. First, it would have to suppose that there is no absolute relation between reading and lateralization (which they *I:vcn

then

one of thaw

in the Annotated

Bibliography

co listed

in the Table

(Ohnmacht

as having found no difference.

and McMorric)

is recorded,

correctly,


263

are attempting to demonstrate) but that this relation is sex-dependent (which is an interesting sex difference in itself). And, second, it is contrary to the evidence of sex differences in reading. Buffery (see Buffery & Gray, 1972) reports an experiment in which three groups of boys and girls, at 5,6 and 7 years, were required to ‘match’ words presented either auditorily or visually and to either the same or different hemispheres. In all cases, presentation of the spoken word to the left hemisphere ensured superior performance, independently of the placement of the visual stimulus. This difference increased in significance with age in girls, but only attained significance in 7 year old boys. The interpretation is one of earlier cerebral lateralization for language in girls. This is surprising in view of the fact that: (a) the best condition is never that in which both stimuli go to the left hemisphere (it being very hard to accept the explanation that two words which have a 50% probability of being the same ‘overload’ the channel capacity of one hemisphere); and (b) that the two ‘mixed’ conditions are so different from each other. A more parsimonious view is that the auditory word is received first and determines the locus of the matching process; the left hemisphere is then appropriate for making linguistic matches. The sex difference may then reflect facility for the response selection process rather than linguistic processing. Dichotic

listening

Kimura (1963) tested groups of boys and girls at each year from 4 to 9 on a dichotic digits task (digits presented simultaneously at both ears - subsequent recall test). Only two groups failed to show a significant right ear (hence left hemisphere) advantage: girls at 7 years, and again at 9 years. Kimura also failed to replicate Ghent’s (196 1) finding that girls show greater touch sensitivity on the ‘non-dominant’ (equals ‘non-writing’ only!) thumb earlier than boys, at 6 years as opposed to 11 years. This dominance in girls disappeared again at 11 years, and the relation to adult dominance is doubtful in view of the unexplained lower thresholds in childhood. Kimura (1967) reports a study based on children of lower socio-economic background, half Canadian and half Californian, in which a group of twenty 5 year old boys, in contrast to a similar group of 5 year old girls, failed to show a significant difference between ears on the digits task. But at 6, 7 and 8 years, boys showed greater asymmetry than girls. Knox and Kimura (1970) fail to find a single significant interaction between ear and sex for five different dichotic tasks in groups at 5, 6, 7 and 8 years. Digits and words were more accurately recalled from right ear presentation, and

264

Hugh Fairweather

environmental sounds from the left. There were no sex differences either for the digits or the placing-objects-on-pictures task (two objects named dichotically). Kimura tentatively ascribes a male superiority in recall of environmental and animal sounds to a differential operation of the right hemisphere between the sexes, even though in the latter there was not even a significant ear difference. Geffner and Hochberg (197 1) replicate the Kimura findings for digits (see also Pizzamiglio and Cecchini, 197 1). Inglis and Sykes (1967) failed to find consistent ear differences between 5 and 10 years, claiming that some of Kimura’s results may be explained by a preponderance of children recalling from the right ear first, rather than recalling more accurately from that ear. The bulk of developmental improvement, not unnaturally, occurs in the second span, indicating an increase in short term storage capacity. Likewise Bakker (1967, 1970) presenting series of digits or Morse sounds monaural!y, found neither asymmetries nor sex differences (see also Nagafuchi, 1970, for limitations of this technique). Bryden (1970) persists in the Bakker habit of reporting only percentages of children having right ear dominance, this time for dichotic digits at 7, 9 and 11 years, and finds the right ear significantly superior overall even when order of recall is controlled for. In addition “... there are very major sex differences: the adult pattern emerges earlier in girls than boys, and laterality effects are much more closely related to reading ability in boys than girls.” (p. 448) In fact, the only pattern occurring earlier in girls than boys is a difference between right and left handers (left handers showing less right ear superiority) and neither this nor the correlation with reading is statistically significant. The hypothesized tendency for earlier lateralization in girls finds no substantiation in studies with even younger children. Thus, in a dichotic listening task using two- and three-syllable words, Nagafuchi (1970) found twice as many significant differences favoring left hemisphere reception in boys than in girls between 3 and 5 years; and Ingram (1975b) using the pointing-to-pictures and placing-objects-on-pictures tasks from Knox and Kimura (1970) finds boys and girls equally lateralized at 3 and 5 years, only girls at 4 failing to show a difference across ears. This last finding has been replicated for dichotic monosyllables amongst both low- and middle-class 4 year olds (Geffner and Dorman; 1976: note that this reflects a sex difference in degree of laterality alone, the distribution of individuals across laterality classes being the same; note also the 48% attrition rate amongst low-class children). Other recent failures to find sex differences include: Berlin et al., (1973) for consonant-vowel pairs, between 5 and 13 years; Friedrich (1974) for


265

words, 7 to 17 years; and Satz, Bakker, Teunissen, Goebel and Van der Vlugt (1975) for digits, 5 to 11 years. The Satz et al., study is otherwise alone both in finding a significant developmental change in the right ear advantage, and also in using as many as four digit pairs per trial, which technique may have strained memory capacities and tempted erroneous stimulus fusions to an extent untouched by earlier studies (see Porter and Berlin, 1975, for comment). The more general failure to promote age changes combined with the very preliminary evidence of pre-language lateralization (Molfese, Freeman and Palermo, 1975: greater left- than right-sided auditory evoked response amplitude to syllables and words in ten infants) at the very least complicates any simple equation between language and lateralization. Such an equation continues to be made between language and anatomy however, the early tentative findings of a relative enlargement of the left temporal planum being confirmed in the most recent and largest sample (Wada, Clarke and Hamm, 1975). A sex difference is seen to reside in a greater number of females exhibiting reversals of the normal pattern, although statistical analysis of all right-left ratios would have had greater heuristic value in assessing the real extent of the difference. Whilst the relation between size and function remains of course hypothetical, this would go some way towards explaining current findings of attenuated functional asymmetries in adult females. The U.S. Public Health Survey for instance finds only males exhibit a right ear advantage for auditory acuity (Kannan and Lipscomb, 1974). Marshall and Holmes (1974) have shown better word recognition for males in the right but not the left visual field (see also Ehrlichman, 1971); and Hannay and Malone (1976a) demonstrate a right field superiority o&y for males for remembering nonsense syllables. An appeal in the latter case to a complex interaction between sex and familial handedness (attenuated asymmetries confined to non-familial right-handed females) finds small support in a follow-up study (Hannay and Malone, 1976b) and none at all in a more comprehensive study on dichotic digits (Briggs and Nebes, 1976). An overall female superiority in this last report may be attributed to females’ higher scoring on a Verbal I.Q. test; one may speculate on the possible influence of such a bias in similar student populations. No overall sex difference is found for dichotic consonant perception by Lake and Bryden (1976), usin g a similar experimental design involving handedness and familial sinistrality. Males were, however, again more clearly lateralized than females, and this effect interacted strongly with presence/ absence of familial sinistrality (FS+/FS-). The most ‘ambilateral’ groups comprised all FS+ females, but only left-handed FS- males. Modulated

266

Hugh Fairweather

laterality effects in females for such left hemisphere tasks are paralleled by evidence for an ostensibly, though not in fact necessarily, right hemisphere task, that of recognizing unknown faces (Perez, Mazzucchi and Rizzolatti, 1976; Umilta, Brizzolara, Tabossi and Fan-weather, 1976). Taylor’s (1969) interpretation of data on the onset of temporal lobe epilepsy, on the other hand, favors greater female lateralization: girls exhaust left hemisphere risk earlier than boys as a protective mechanism for an especially lateralized and/or important function. The interpretation finds little support from clinical data in adults. A combination of 4 WAIS Verbal subtests, for instance, promoted no sex/side of lobectomy interaction; nor, for that matter, did it discriminate overall between left- and right-sided lobectomies (Lansdell, 1968). Lansdell (196 1) had earlier reported that female verbal abilities are less disturbed by left-sided lobectomy. Whilst L‘. . . left temporal lesions were disruptive only in males” for Gorham’s Proverbs Test, this hardly taps one of the cardinal tasks of language. A similar finding for subcortical neurosurgery and a word association test (Lansdell, 1973) is confounded (as indeed are many other findings) by the fact that the test otherwise yields no overall sex difference. There is, in sum, very little evidence of an overall sex difference in verbal ability. There is, as indeed also for spatial ability, very suggestive evidence that this ability is more symmetrically represented across the cerebral hemispheres in the female than in the male brain. Whilst no single study could claim categorical proof; whilst many clinical studies remain much less than ideally controlled; and whilst the relation between, for instance, dichotic laterality patterns and the global verbal ability remains obscure, there is also as yet no significant pointer to the contrary.

Conclusion With respect to sex differences in cognition one can only conclude that there are very few: too few to tempt the further longevity of popular dichotomies by listing the possibilities here. Certainly, the incidence of such differences is outnumbered by the qualifications noted in the present review: age; culture; birth order; family size; sex of experimenter; and replicability both between and within studies. These are just some of the factors we know about, but what, for example, of possible sex biases in the constitution of student populations (upon which most of the adult evidence is based), or the volunteering patterns within those populations? It is perhaps no accident that the failure to find sex differences is clearest in childhood, where the population is least selected.

Sex differences

in cognition

261

This conclusion underlines, importantly, that we are far from having an adequate data base upon which to construct theories. The theoretical candidate we have entertained here, that of cerebral lateralization (Buffer-y and Gray, 1972) fails not only on these general grounds, but also on the specific grounds that such sex/hemisphere interactions as have been found precisely contradict the theory’s predictions. Other candidates may be swiftly and similarly dismissed. Perhaps the most attractive, the hypothesis of spatial ability being carried on an X-linked recessive gene, is complicated by the failure to find a (highly circumscribed) sex difference until puberty. A recent speculation in terms of maturation rate (Waber, 1976) is embarrassed, factually, by a failure to find a sex difference. And logically, by a miscued syllogism: early maturers differ, cognitively, from late maturers; girls mature earlier than boys; therefore; sex differences in cognition are a function of maturation rate. The first ‘premiss’ awaits substantial replication; the second awaits verification for psychologically relevant indices (the ‘implicit physiological correlates’ of op. cit., p. 573); and the conclusion is not possible since it contains more information than the premisses, i.e., that there are sex differences, a statement which, as we have seen, requires considerable qualification. It must be stressed, finally, that the majority of studies reviewed here and elsewhere are both ill-thought and ill-performed. Whilst in other circumstances this may be regarded as the occupational hazard of the scientific enterprise, here such complacency is compounded by the social loadings placed upon these kinds of results. It is clearly very easy to include sex as a bonus factor in experiments which have little scientific merit. We cannot pretend that we are testing a theory of sex differences, since at present none can exist. Legitimate studies of sex differences can only grow first out of observations of clear individual differences in the investigation of salient psychological processes; and second, from the observation that the groups of individuals thus differentiated have clearly biased compositions when divided by sex. Studies with clinical populations would already tend to meet these criteria. Studies within the normal population, predicated on the assumption that discriminations are useful, can only be regarded as tempting sexism.

268

Hugh Fainveather

References Albert, M. L. (1975) Cerebral dominance and reading habits. Nature, 256, 403404. Allen, C. N. (1930) Recent studies in sex differences. Psychol. Bull., 27, 394407. Ames, L. B. and Ilg, F. L. (1964) Sex differences in test performance of matched girl-boy pairs in the 5-9 year old age-range. J. genet. Psychol., 104, 25-34. Ammons, R. B., Alprin, S. 1. and Ammons, C. H. (1955) Rotary pursuit performance as related to sex and age of pre-adult subjects. J. exp. PsychoL, 49, 127-133. Annett, M. (1970) Growth of manual preference and speed. Brit. J. Psycho/., 61, 545-558. Annett, M. (I 972) The distribution of manual asymmetry. Brit. J. Psychol., 63, 343-358. Annett. M. (1973) Laterality of childhood hemiplegia and the growth of speech and intelligence. Cortex, 9, 4-33. Asher, S. R. and Gottman, J. M. (1973) Sex of teacher and student reading achievement. J. educ. PsychoL, 65, 168-171. Asher, S. R. and Markell, R. A. (1974) Sex differences in comprehension of high- and low-interest reading materials. J. educ. Psychob, 66, 68@687. Bakan, P. and Putnam, W. (1974) Right left discrimination and brain lateralization. Archs. Neurol., 30,334-335. Bakker, D. J. (1967) Left-right differences in auditory perception of verbal and non-verbal material by children. Q. J. exp. PsychoI., 19, 334-336. Bakker, D. J. (1969) Eye asymmetry on simple RT task with children. Percept. mot. Skills, 28, 328. Bakker, D. J. (1970) Ear-asymmetry with monaural stimulation: relations to lateral dominance and lateral awareness. Neuropsychologia, 8, 103-l 17. Banikiotes, 1:. G., Montgomery, A. A. and Banikiotes, P. G. (1972) Male and female auditory reinforcement of infant vocalizations. Devel. Psychol.. 6, 476481. Barnsley, R. H. and Rabinovitch, M. S. (1970) Handedness: proficiency versus stated preference. Percept. mot. Skills, 30, 343-362. Bayley, N. (1965) Comparisons of mental and motor test scores for ages l-15 months by sex, birth order, race, geographical location and education of parents. CTtild. DeveL, 36, 379411. Bell, R. Q. and Darling, J. F. (1965) The prone head reaction in the human neonate: relation with sex and tactile sensitivity. Child Devel., 36, 943-949. Bellis, C. J. (1933) Reaction time and chronological age. hoc. Sot. exp. Biol. Med., 30, 801-803. Belmont, L. and Birch, H. G. (1963) Lateral dominance and right-left awareness in normal children. Child Devel., 34, 257-270. Bennett, G. K., Seashore, H. G. and Wesman, A. G. (1966) Differential aptitude tests. New York: The Psychological Corporation. Benton, A. L. (1959) Right-left discrimination and finger localization. New York: Hoeber-Harper. Berg, W. K., Adkinson, C. D. and Strock, B. D. (1973) Duration and frequency of periods of alertness in neonates. Devel. Psychol., 9, 434. Berger, M., Yule, W. and Rutter, M. (1975) Attainment and adjustment in two geographical areas: II The prevalence of specific reading retardation. Brit. J. Psychiat., 126, 5 IO-5 19. Berlin, C. I., Hughes, L. F’., Lowe-Bell, S. S. and Berlin, H. L. (1973) Dichotic right ear advantage in children 5 to 13. Cortex, 9, 3933401. Berry, J. W. (1966) Temne and Eskimo perceptual skills. ht. J. Psycho/., I, 207-229. Berry, J. W. and Annis, R. C. (1974) Ecology, culture and psychological differentiation. ht. J. Psychol., 9, 173-193. Bickersteth, M. E. (1917) The application of mental tests to children of various ages. Brit. J. Psycho/., 9, 23-73. Bittner, A. C. and Shinedling, M. M. (1968) A methodological investigation of Piaget’s concept of conservation of substance. Genet. Psychol. Monog., 77, 135-165.


Bock,

269

R. D. (1973) Word and image: sources of the verbal and spatial factors in mental test scores. Psychometrika, 38, 437-457. Bock, R. D. and Kolakowski, D. (1973) Further evidence of sexBnked major-gene influence on human spatial visualizing ability. Amer. J. hum. Genet., 25, 1-14. BOB, T. 3. (1971) Correlation of WISC with motor speed and strength for brain damaged and normal children. J. Psychol., 77, 169-172. Bolton, T. L. (1903) The relation of motor power to intelligence. Amer. J. Psychol., 14. 351-367. Botwinick, J. (1971) Sensory-set factors in age differences in reaction time. J. Genet. Psychot., 119, 24 l-249. Bradbury, P. J., Wright, S. D., WaIker, C. E. and Ross, J. M. (1975) Performance on the WISC as a function of sex of E, sex of S and age of S. J. Psychol., 90, 51-55. Brekke, B. and Williams, I.D. (1973) Conservation and sex. Percept. mot. Skills, 37, 14. Brenner, M. W. and Gillman, S. (1966) Visuomotor ability in schoolchildren - a survey. Devel. med. Child Neurol., 8, 686-703. Bresnahan, J. L., Hillard, J. P. and Shapiro, M. M. (1973) Reinforcement in children: received and expected. Bull. Psychon. Sot., 2, 195-197. Briggs, G. G. and Nebes, R. D. (1975) Patterns of hand preference in a student population. Cortex, II, 230-238. Briggs, G. G. and Nebes, R. D. (1976) The effects of handedness, family history and sex on the performance of a dichotic listening task. Neuropsychologia, 14, 129-133. Briggs, P. F. and Tellegen, A. (197 1) Development of the Manual Accuracy and Speed Test. Percept. mot. Skills, 32, 923-943. Brooks, R. M. and Goldstein, A. C. (1963) Recognition by children of inverted photographs of faces. Child Devel., 34, 1033-1040. Broom, M. E. (1941) A study of race and sex differences in reading comprehension. J. educ. Res., 34, 587-593. Broverman, D. M. and Klaiber, E. L. (1969) Negative relationships between abilities. Psychometrika, 34, 5-20. Broverman, D. M., Klaiber, E. L., Kobayashi, Y. and Vogel, W. (1968) Roles of activation and inhibition in sex differences in cognitive abilities. Psychol. Rev., 75, 23-50. Brown, M. H. and Bryan, G. E. (1955) Sex differences in intelligence. J. clin. Psycho/., II, 303-304. Bruml, H. (1972) Age changes in preference and skill measures of handedness. Percept. mot. Skills, 34, 3-14. Bryden, M. P. (1970) Laterahty effects in dichotic listening: relations with handedness and reading ability in children, Neuropsychologia, 8, 443450. Bryden, M. P. (1972) Auditory-visual and sequential-spatial matching in relation to reading ability. Child Devel., 43, 824-832. Bryden, M. P. (1973) Perceptual asymmetry in vision: relation to handedness, eyedness, and speech lateralization. Cortex, 9, 418432. Bryden, M. P. (1976) Response bias and hemispheric differences in dot localization. Percept. Psychophys.. 19. 23-28. Buchsbaum, M. S., Henkin, R. I. and Christiansen, R. L. (1974) Age and sex differences in averaged evoked responses in a normal population, with observations on patients with gonadal dysgenesis. Electroenceph. clin. Neurophysiol.) 37, 137-144. Buffery, A. W. H. and Gray, J. A. (1972) Sex differences in the development of perceptual and linguistic skills. In: C. Ounsted and D. C. Taylor (eds) Gender Differences: Their Ontogeny and Significance. London: Churchill Livingstone. Burt, C. (1909) Tests of general intelligence. Brit. J. Psychok, 3, 94-177. Burt, C. (1912) Experimental tests ofhigher mental processes and their relation to general intelligence. J. exp. Ped., 1. 93-l 12.

270

Hugh Fairweather

C. and Moore, R. C. (1912) The mental differences between the sexes. J. exp. Ped., 1, 273284 and 3555396. Busby, W. A. and Hurd, D. E. (1968) Relationship of auditory and visual reaction times to reading achievement. Percept. mot. Skills, 27, 447450. Caldwell, H. S. and Leeper, H. A. (1974) Temporal patterns of neonatal vocalizations: a normative investigation. Percept. mot. Skills, 38, 91 l-916. Calfee, R. C., Lindamood, P. and Lindamood, C. (1973) Acoustic-phonetic skills and reading kindergarten through twelfth-grade.J. educ. Psychol., 64, 293-298. Caron, A. J., Caron, R. F., Caldwell, R. C. and Weiss, S. J. (1973) Infant perception of the structural properties of the face.Devel. Psychol., 9, 385-399. Cashdan, A., Pumfrey, P. D. and Lunzer, E. A. (1970-71) Children receiving remedial teaching in reading. Educ. Res., 13, 98-105. Chance, J. 1;. and Goldstein, A. G. (1971) Internal-external control of reinforcement and cmbeddedfigures performance. Percept. Psychophys., 9, 33-34. Clark, M. M. (1970) Readingdifficulties in schools. Harmondsworth. Penguin Books. Clement, F. (1962) Modifications du temps de reaction simple en fonction de l’agc et de quelques autres facteurs. Revue Psychol. appl., 12, 163-188. Cohen, A. I. (1966) Hand preference and developmental status of infants. J. genet. Psychol., 108, 337-345. Cohen, B. D., Berent, S. and Silverman, A. J. (1973) Field dependence and lateralization of function in the human brain. Archs. gen. Psych&. 28, 165-167. Coie, J. D. and Dorval, B. (1973) Sex differences in the intellectual structure of social interaction skills. Devel. Psychol., 8, 261-267. Connolly, K., Brown, K. and Bassett, E. (1968) Developmental changes in some components of a motor skill. &it. J. Psych&., 59, 305-314. Connolly, K. J. and Stratton, P. (1968) Developmental changes in associated movements. Devel. med. child Neural., 10, 49-56. Cook. T. W. and Shephard, A. H. (1958) Performance on several controldisplay relationships as a function of age and sex. Percept. mot. Skills, 8, 339-345. Corah, N. L. (1965) Differentiation in children and their parents. J. Pers., 33, 300&308. Cornell, E. H. (1974) Infants’ discrimination of photographs of faces following redundant prcsentations. J. exp. child Psychol., 18, 98~106. Critchley, M. (1970) The dyslexic child. London, Heincmann. Croxen, M. L. and Lytton, H. (1971) Reading disability and difficulties in finger localization and right-left discrimination. Devel. Psychol., 5, 256-262. Darley, F. L. and Winitz, H. (1961a) Age of first word: review of research. J. Speech hear. Dis., 26, 272-290. Darley, F. L. and Winitz, H. (1961b) Comparison of male and female kindergarten children on the WISC. J. genet. Psychol., 99, 4149. Davies, A. D. M. (1965) The perceptual maze test in a normal population. Percept. mot. Skills, 20, 287-293. Davol, S. J. and Breakell, S. L. (1968) Sex differences in rotary pursuit performance of young children: a follow-up. Percept. mot. Skills, 26, 1199-1202. Davol, S. H., Hastings, M. L. and Klein, D. A. (1965) Effect of age, sex, and speed of rotation on rotary pursuit performance in young children. Percept. mot. Skills, 21, 35 l-357. Deich, R. F. and Hodges, P. M. (1973) Motion sickness, field dependence, and levels of development, Percept. mot. Skills, 36, 1115-l 120. Dcnckla, M. B. (1973) Development of speed in repetitive and successive finger-movements in normal children. Duel. med. child Neural., 15, 635-645. Burt,

Sex differences

Denckla,

M. B. and Rudel,

R. (1974)

Rapid

‘automatized’

naming

of pictured

in cognition

objects,

colors,

27 1

letters

and numbers by normal children. Correx, IO, 186-202. Dimond, S. I. and Beaumont, J. G. (1973) Difference in the vigilance performance of the right and left hemispheres. Cortex, 9, 259-265. Dimond, S. J. and Beaumont, J. G. (1974) Hemisphere function in rhe human brain. London, Elek Science. Di Vesta, 1;. J. and Stauber, K. A. (1971) Identification of verbal concepts by preschool children. Devel. Psychol.. 5, 81-85. Drinkwater, B. L. (1968) Speed and accuracy in decision responses of men and women pilots. Ergonomics, II, 61-67. Droege, R. C. (1967) Sex differences in aptitude maturation during high schoo1.J. counsel. Psychol., 14,407-ill. Dwyer, C. A. (1974) Influence of children’s sex role standards on reading and arithmetic achievement. J. educ. Psykhol., 66, 811-816. Dye, N. W. and Very, P. S. (1968) Growth changes in factorial structure by age and sex. Genef. Psychol. Monog., 78, 55-88. Eckert, E. M. (1970) Visual-motor tasks at 3 and 4 years of age. Percept. mot. Skills, 31, 560. Ehrlichman, H. I. (1971) Hemispheric functioning and individual differences in cognitive abilities. Diss. Abstr., 33B, 2319. Eikind, D. (1961) Quantity conceptions in junior and senior high school students. Child. Devel., 32, 55 l-560. Elliot, F. R. and Louttit, C. M. (1948) Auto braking reaction times to visual vs. auditory warning signals. Proc. Indian Acad. Sci., 4 7, 220-225. Elliott, R. (1970) Simple reaction time: effects associated with age, preparatory interval, incentive shift, and mode of presentation. J. exp. child Psychol., 9, X6-107. Elliott, R. (1973) Personal communication. Ellis, H. D. (1975) Recognizing faces. Brir. J. Psychol., 66, 409426. Emmett, W. G. (1954) Secondary modern and grammar school performance predicted by tests given in primary schools. Brit. J. educ. Psychol., 24, 91-98. Engel, R. (1967) Elcctroencephalographic responses to sound and to light in premature and full term neonates. Luncet, 8 7, 18 1. Engel, R. and Pay, W. (1972) Visual evoked responses at birth, verbal scores at three years, and IQ at four years. Devel. Med. child Neural., 14, 283-289. Engel, R. and Henderson, N. B. (1973) Visual evoked responses and IQ scores at school age. Devel. Med. child Neurol., 15, 136-145. Estcs, B. and Combs, A. (1966) Perception of quantity. J. genet. Psychol., 108, 333-336. Fagan, J. F. (1972) Infants’ recognition memory for faces. J. exp. child Psychol., 14, 453476. Fagan, J. F. (1973) Infants’ delayed recognition memory and forgetting. J. exp. child Psycho!., 16, 424450. Fairweather, H. (1974) Choice reaction times in children. Unpublished D. Phil. thesis, University of Oxford. Fairweather, H. (1975) Processing the terms “right” and “left”: Compatibility, practice, lateralization and sex differences. Paper presented to the Psychonomic Society, Denver. Fairweather, H. and Butterworth, C. (1977, in press) The WPPSI at four years: A sex difference in Verbal-Performance discrepancies. Brir. J. educ. Psychol. Fairweather, J. and Hutt, S. J. (1972) Sex differences in perceptual motor skill in children. In: C. Ounsted and D. C. Taylor (eds.) Gender differences: their ontogeny and significance. London. Churchill Livingstone. I:airweather, H. and Hutt, S. J. (1976, in submission) On the rate of gain of information in children.

272

Hugh Fairweather

Fogelman, K. R. (1970) Piagetian tests and sex differences - II. E&c. Res., 12, 154-155. Ford, M. A., Stern, D. N. and Dillon, D. J. (1974) Performanceof children ages 3 to 5 on the Draw-aPerson task: Sex differences. Percept. mot. Skills, 38, 1188. Forgays, D. G. (1953) The development of differential word recognition. J. exp. Psychol.. 45, 165168. Frankcnburg, W. K. and Dodds, _I. B. (1967) The Denver Development Screening Test. J. Pediuf.. 71, 181-191. Friedrich, D. (1974) Developmental analysis of memory capacity and information encoding strategy. Devel. Psychol.. 10, 559-563. Fulton, C. D. and Hubbard, A. W. (1975) Effect of puberty on reaction and movement times. Res. Q. Am, Ass. Hlth,phys. Educ., 46, 335-344. Garai, J. E. and Scheinfeld, A. (1968) Sex differences in mental and behavioral traits. Genet. Psychol. Monog., 77, 169-299. Garrett, H. E., Bryan, A. I. and Perl, R. E. (1935) The age factor in mental organization. Archs. Psychol., No. 176. Geblewiczowa, M. (1961) Reaction time as a developmental trait connected with sex. Psyehofogia Wych., 18, 3945. Geffner, D. S. and Dorman, M. F. (1976) Hemispheric specialization for speech perception in fouryear-old children from low and middle socioeconomic classes. Cortex, 12, 71-73. Geffner, D. S. and Hochberg, I. (1971) Ear laterality performance of children from low and middle socioeconomic levels on a vrbal dichotic listening task. Cortex, 7, 193-203. Gesell, A. (1940) The first five years of life. London, Methuen. Gesell, A. and Ames, L. B. (1947) The development of handedness. J. genet. Psychol., 70, 155-175. Gesell, A. and Ilg, F. L. (1946) The child from five to ten. London, Hamish Hamilton. Ghent, L. (1961) Developmental changes in tactual thresholds on dominant and nondominant sides. J. camp. physiol. Psychol., 54, 670473. Gibson, E. J., Osser, H. and Pick, A. D. (1963) A study of the development of grapheme-phoneme correspondences. J. verb. Learn. verb. Beh., 2, 142-146. Gilbert, J. A. (1894) Researches on the mental and physical development of schoolchildren. Stud. Yale psychol. Lab., 2, 40-100. Gilbert, J. H. (1973) Acoustical features of children’s vowel sounds: development by chronological age versus bone age. Lang. Speech, 16, 218-223. Glass, D. C., Neulinger, J. and Brim, 0. G. (1974) Birth order, verbal intelligence and reading aspiration. Child Devel., 45, 807-811. Going, M. and Read, J. D. (1974) Effects of uniqueness, sex of subject, and sex of photograph on facial recognition. Percept. mot. Skills, 39, 109-l 10. Goldberg, S., Perlmutter, M. and Myers, N. (1974) Recall of related and unrelated lists by 2-year-olds. J. exp. Psychol.. 18, l-8. Golden, C. J. (1974) Sex differences in performance on the Stroop Color and Word Test. Percept. mot. Skills, 39, 1067-1070. Goodenough, F. L. (1927) The consistency of sex differences in mental traits at various ages. Psychol. Rev., 34,44&462. Goodenough, F. L. (1935a)Thedevelopment ofthe reactive process from early childhood to maturity. 1. exp. PsychoI., 18. 431-450. Goodenough, F. L. (1935b) A further study of tapping in early childhood. J. uppl. Psychol., 19, 309-319. Goodenough, F. L. and Smart, R. C. (1935) Inter-relationships of motor abilities in young children. ChildDevel., 6, 141-153. Greer, G. (197 1) The female eunuch. St. Albans, Granada Publishing Limited.


273

Gullickson, G. R. and Crowell, D. H. (1964) Neonatal habituation to electrotactual stimulation. J. exp. child PsychoI., I, 388-396. Gutteridge, M. V. (1930) A study of motor achievements of young children. Archs. Psychol., No. 244. Haaf, R. A. (1974) Complexity and facial resemblance as determinants of response to facelike stimuli by S- and lo-week old infants. J. exp. child Psychol., 18, 480487. Hannay, H. J. and Malone, D. R. (1976a) Visual field effects and short-term memory for verbal material. Neuropsychologia, 14, 203209. Hannay, H. J. and Malone, D. R. (1976b) Visual field recognition memory for right-handed females as a function of familial handedness. Cortex, 12, 4 l-48. Heap, M. and Wyke, M. (1972) Learning of a unimanual motor skill by patients with brain lesions: an experimental study. Cortex, 8, l-18. Henderson, N. B. and Engel, R. (1974) Neonatal visual evoked potentials as predictors of psycho educational tests at age seven. Devel. Psychol., 10, 269-276. Henry, F. M. (1961) Stimulus complexity, movement complexity, age, and sex in relation to reaction latency and speed in limb movements. Res. Q. Am. Ass. Hlth. phys. Educ., 32, 353-366. Herman, D. 0. (1968) A study of sex differences on the Wechsler Preschool and Primary Scale of Intelligence. From: Proceedings, 76th Annual Convention, APA, pp. 455-456.* Hildreth, G. (1949) The development and training of hand dominance: II. Developmental tendencies in handedness. J. genet. Psychol., 75, 221-254. Hill, J. F. (1972) Interrelations of the reaction time, movement time, motor ability, and physical fitness of children five through eight years old. Diss. Abstr., 32, 7-A, 3759. Hodgkins, J. (1963) Reaction time and speed of movement in males and females of various ages. Res. Q. Am. Ass. Hlth. phys. Educ., 34, 335-343. Hohle, R. (1967) Component process latencies in reaction times of children and adults. In: L. P. Lipsitt and C. C. Spiker (eds.) Advances in child development and behavior, Vol. 3. New York, Academic Press. Hoving, K. L., Morin, R. E. and Konick, D. S. (1970) Recognition reaction time and size of the memory set: a developmental study. Psychon. Sci., 21, 247-248. Husband, R. W. and Ludden, M. J. (1931) Sex differences in motor skills. J. exp. Psychol., 14, 414422. Hutt, C. (1972) Neuroendocrinological, behavioral and intellectual aspects of sexual differentiation in human development. In: C. Ounsted and D. C. Taylor (eds.) Gender differences: their ontogeny and significance. London, Churchill Livingstone. Hutt, C. (1972a) Malesandfemales. Harmondsworth, Penguin Books. lnglis, J. and Sykes, D. H. (1967) Some sources of variation in dichotic listening performance in children. J. exp. child Psychoi., 5, 480-488. Ingram, D. (1975a) Motor asymmetries in young children. Neuropsychologia, 13, 955102. Ingram, D. (1975b) Cerebral speech lateralization in young children. Neuropsychologia, 13, 103-105. Irwin, 0. C. and Chen, H. P. (1946) Development of speech during infancy: curve of phonemic types. J. exp. Psychoi., 36, 431436. Iversen, I. A., Silberberg, N. E. and Silberberg, M. C. (1970) Sex differences in knowledge of letter and number names in kindergarten. Percept. mot. Skills, 31, 79-85. Jacobson, L. I., Berger, S. E. and Millham, J. (1970) Individual differences in cheating during a temptation period when confronting failure. J. Pers. sot. Psychok, 15, 48-56. Jensen, D. R. and Engel, R. (197 1) Statistical procedures for relating dichotomous responses to maturation and EEG measurements. Electroenceph. clin. Neurophysiol.. 30, 437-443.

*Data obtainable

from The Psychological

Corporation,

757 Third

Ave., New York 10017.

274

Jones,

Hugh Fairweather

D. R. and Benton, A. L. (1968) Reaction time and mental age in normal and retarded children. Amer. J. ment. Defic., 73, 143--147. Jones, H. E. (1937) Reaction time and motor development. Amer. J. Psychol., 50, 181-194. Jones, R. W. and Ellis, N. R. (1962) Inhibitory potential in rotary pursuit acquisition by normal and defective subjects. J. exp. PsychoL, 63, 534-537. Kannan, P. M. and Lipscomb, D. M. (1974) Bilateral hearing asymmetry in a large population. J. acoust. Sot. Amer., 55, 1092-1094. Kaufman, A. S. and Doppelt, J. E. (1976) Analysis of WISC-R standardization data in terms of the stratification variables. ChiN Devel., 47, 165-17 1. Keogh, B. K. (1971) Pattern copying under three conditions of an expanded spatial field. Devel. Psychol., 4, 25-3 1. Keogh, B. K. (1972) Preschool children’s performance on measures of spatial organization, lateral preference, and lateral usage. Percept. mot. Skills, 34, 299-302. Keogh, B. K. and Smith, C. E. (1967) Visuo-motor ability for school prediction: a seven-year study. Percept. mot. Skills, 2.5, 101-l 10. Kimura, D. (1963) Speech lateralization in young children as determined by an auditory test. J. camp. physiol. Psychol., 56, 899-902. Kimura, D. (1967) Functional asymmetry of the brain in dichotic listening. Cortex, 3, 163-178. Kimura, D. (1969) Spatial localization in left and right visual fields. Can. J. Psycho/., 23, 445 -458. Kimura, D. and Archibald, Y. (1974) Motor functions of the left hemisphere. Brain, 97, 337.-350. Kimura, D. and Vanderwolf, C. H. (1970) Relation between hand preference and the performance of individual finger movements by right and left hands. Brain, 93, 769-774. Kinsbourne, M. and McMurray, J. (1975) The effect of cerebral dominance on time sharing between speaking and tapping by preschool children. Child Devel., 46, 240-242. Knights, R. M. and Moule, A. D. (1967) Normative and reliability data of finger and foot tapping in children. Percept. mot. Skills, 25, 7 17-720. Knox, C. and Kimura, D. (1970) Cerebral processing of non-verbal sounds in boys and girls. Neuropsychologia, 8, 227-237. Koch, H. L. (1954) The relation of ‘Primary Mental Abilities’ in five- and six-year-olds to sex of child and characteristics of his sibling. Child Devel., 25. 209-223. Kohen-Raz, R. (1965) Developmental increase of ambidexterity in target tests. J. child Psychol. Psychiat., 6, 43-54. Korner, A. F. (1970) Visual alertness in neonates: individual differences and their correlates. Percept. mot. Skills, 31, 499-509. Korner, A. F. (1973) Sex differences in newborns with special reference to differences in the organization of oral behavior. J. child Psycho!. Psychiat., 14, 19-29. Kretz, V., Suchenwirth, R. and Ferner, U. (1970) Handigkeit in Abhangigkeit vom Lebensalter: Untersuchungen an 574 3-6 jahrigen Knaben unter besondercr Beriicksichtigung der Geschlechtsspezifitat. Neuropsychologia, 8, 215-226. Lake, D. A. and Bryden, M. P. (1976) Handedness and sex differences in hemispheric asymmetry. Brain Lang., 3, 266-282. Lansdell, H. (1961) The effect of neurosurgery on a test of proverbs. Amer. Psychol., 16, 448. Lansdell, H. (1968) The use of factor scores from the Wechsler-Bellevue Scale of intelligence in assessing patients with temporal lobe removals. Cortex, 4, 257-268. Lansdell, H. (1973) Effect of neurosurgery on the ability to identify popular word associations. J. abnorm. Psychol., 81, 255-258. Laosa, L. M. and Brophy, J. E. (1972) Effects of sex and birth order on sex-role development and intelligence among kindergarten children. Devel. Psychol., 6, 409415. Lasky, R. E., Klein, R. E. and Martinez, S. (1974) Age and sex discriminations in five- and six-month old infants. J. Psychol., 88, 317V324.

Sex differences

in cognition

275

Lewis, M. (1972) State as an infantenvironment interaction: an analysis of mother-infant interactions as a function of sex. Merrill-Palmer Q., 18, 95-122. Ligon, E. M. (1932) A genetic study of color naming and word association. Amer. J. Psychol., 44, 103-122. Lomas, J. and Kimura, D. (1976) Interhemispheric interaction between speaking and sequential manual activity. Neuropsychologia, 14, 23-34. Long, A. B. and Looft, W. R. (1972) Development of directionality in children: Ages six through twelve. Devel. Psychol., 6, 375380. Lord, F. E. (1941) A study of spatial orientation of children. J. e&c. Res., 34.481-505. Lyle, J. G. and Johnson, E. G. (1974) Analysis of WISC coding: 5. Prediction of coding performance. Percept. mot. Skills, 39, 11 l-l 14. MacArthur, R. (1967) Sex differences in field dependence for the Eskimo: replication of Berry’s findings. ht. J. Psychol., 2, 139-140. Maccoby, E. E. (1966) Sex differences in intellectual functioning. In: E. E. Maccoby (ed.) The development of sex differences. London, Tavistock. Maccoby, E. E., and Jacklin, C. N. (1975) The psychology of sex differences. Oxford, University Press. Macfarlane Smith, I. (1948) Measuring spatial ability in school pupils. &cup. Psychol., 22, 150-159. Macfarlane Smith, I. (1964) Spatial ability: its educational and social significance. London, University of London Press. Mcl:ie, J. (1960) Psychological testing in clinical neurology. J. Nerv. Ment. Dis., 131, 383-393. McGinnis, E:. (1929) The acquisition and interference of motor habits in young children. Genet. Psychol. Monogr., 6, 203-309. McGlone, J. and Davidson, W. (1973) The relation between cerebral speech laterality and spatial ability with special reference to sex and hand preference. Neuropsychologia, II, 105-l 13. M&lone, J. and Kertesz, A. (1973) Sex differences in cerebral processing of visuospatial tasks. Cortex, 9, 313-320. McKeevcr, W. F. and Huling, M. D. (1970) Lateral dominance in tachistoscopic word recognition of children at two levels of ability. Q. J. exp. Psychol., 22, 600-604. McManis, D. L. (1965) Pursuit-rotor performance of normal and retarded children in four verbalincentive conditions. Child Devel., 36, 667-683. Malmquist, L. (1970) A decade of reading research in Europe, 1959-1969: a review. J. educ. Res., 63, 309-329. Marcel, T., Katz, L. and Smith, M. (1974) Laterality and reading efficiency. Neuropsychologia, 12, 131-139. Marshall, J. C. and Holmes, J. M. (1974) Sex, handedness, and differential hemispheric specialization for components of word perception. Int. Res. Commun. Syst., 2, 1344. Martens, R. (1970) Social reinforcement effects on preschool children’s motor performance. Percept. mot. Skills, 31, 787-792. Marx, M. (1972) Recall/recognition differcnccs as a function of elementary school grade. Psychon. Sci., 27, 230-232. Matarazzo, J. D. (1972) Wechsler’s measurement and appraisal ofadult intelligence. Baltimore, Maryland, Williams and Wilkins. Mathur, V. K. (1964) Sex differences in reaction time. IndianJ. Psychol., 39, (I), 31-33. May, R. B. and Hutt, C. (1974) Modality and sex differences in recall and recognition memory. Child Devel.. 45, 228-231. Mayo, I’. R. and Bell, J. M. (1972) A note on the taxonomy of Witkin’s field-independence measures. Brit. J. Psychol., 63, 255-256. Meyer, W. J. and Bendig, A. W. (1961) A longitudinal study of the Primary Mental Abilities Test. J. educ. Psychol., 52, 50-60.

276

Hugh Fairweather

Miller, D. J., Cohen, L. B. and Hill, K. T. (1970) A methodological investigation of Piaget’s theory of object concept development in the sensory-motor period. J. exp. child Psychol., 9, 59-85. Miller, L. K. and Turner, S. (1973) Development of hemifield differences in word recognition. J. e&c. Psychol., 65, 172-176. Mittler, P. and Ward, J. (1970) The use of the Illinois Test of Psycholinguistic Abilities on British fouryear-old children: a normative and factorial study. Brit. J. educ. Psychol., 40, 43-54. Molfese, D. L., Freeman, R. B. and Palermo, D. S. (1975) The ontogeny of brain lateralization for speech and nonspeech stimuli. Brain Lang., 2, 356-368. Moore, J. E. (1937) A test of eye-hand co-0rdination.J. appl. Psychol., 21, 668472. Moore, J. E. (1940) Further study of sex differences in speed of reading. Peabody J. Educ., 17, 359-362. Morley, M. (19.57) The development and disorders of speech in childhood. London, Livingstone. Moss, H. A. and Robson, K. S. (1970) The relation between amount of time infants spend at various states and the development ofvisual behavior. Child Devel., 41, 509-S 17. Nagafuchi, M. (1970) Development of dichotic and monaural hearing abilities in young children. Acta Otolaryng., 69, 409-415. Nelson, N. W. (1976) Comprehension of spoken language by normal children as a function of speaking rate, sentence difficulty, and listener age and sex. Child Deuel., 47, 299-303. Nerlove, S. B., Munroe, R. H. and Munroe, R. L. (1971) Effect of environmental experience on spatial ability: a replication. J. Sot. Psychol., 84, 3-10. Newcombe, F. G., Ratcliff, G. G., Carrivick, P. J., Hiorns, R. W., Harrison, G. A. and Gibson, J. B. (1975) Hand preference and IQ in a group of Oxfordshire villages. Ann. hum. Biol.. 2, 235-242. Noble, C. E., Baker. B. L. and Jones, T. A. (1964) Age and sex parameters in psychomotor learning. Percept. mot. Skills, 19. 935-945. Olive, H. (1972) A note on sex differences in adolescents’ divergent thinking. J. Psychol., 82, 3942. Olson, M. E. (1973) Laterality differences in tachistoscopic word recognition in normal and delayed readers in elementary school. Neuropsychologia, 1 I, 343-350. Ong, J. and Rodman, T. (1972) Sex and eye-hand preferential difference in star-tracing performance. Amer. J. Optom., 49, 436438. Ohrlich, E. S. and Barnet, A. B. (1972) Auditory evoked responses during the fist year of life. Elecfroenceph. clin. Neurophysiol., 32, 16 l-169. Pande, C. G. (1970) Sex differences in field dependence: confirmation with Indian sample. Percept. mot. Skills, 31, 70. Paris, S. G. (1973) Comprehension of language connectives and propositional logical relationships. J. exp. child Psychol., 16, 278-291. Parisi, D. (197 1) Development of syntactic comprehension in preschool children as a function of socioeconomic level. Devel. Psychol., 5, 186-189. Parlee, M. B. and Rajagopal, J. (1974) Sex differences on the Embedded-figures Test: a cross-cultural comparison of college students in India and in the United States. Percept. mol. Skills, 39, 1311-1314. Paynter, E. T. and Petty, N. A. (1974) Articulatory sound: acquisition of two-year-old children. Percept. mot. Skills, 39, 1079-1085. Pedersen, D. M., Shinedling, M. M. and Johnson, D. L. (1968) Effects of sex of examiner and subject on children’s quantitative test performance. J. Pers. sot. Psycho/., IO, 25 l-254. Peretti, P. 0. (1969) Cross-sex and cross-educational level performance in a color-word interference task. Psychon. Sci., 16, 321-323. Pcretti, P. 0. (1971) Effects of non-competitive, competitive instructions, and sex on performance in a color-word interference task. J. Psychof.. 79, 67-70. Perez, E., Mazzucchi, A. and Rizzolatti, G. (1975) Tempi di reazione discriminativi alla presentazione di materiale fisiognomico in soggetti maschili e femminili normali. Boll. Sot. Ifal. Biol. sper., 51, 1445-1450.


277

Philip, B. R. (1934) Reaction times of children. Amer. J. Psychol., 46, 379-396. Phillips, L. J. and Bannon, W. J. (1968) The Stanford-Binet, Form L-M, third revision: a local English study of norms, concurrent validity and social differences. Brit. J. educ. PsychoI., 38, 148-161. Pidgeon, D. A. (1960) A national survey of the ability and attainment of children at three age levels. Brit. J. educ. Psychoi.. 30, 124-133. Pizzamiglio, L. (1974) Handedness, ear-preference, and field independence. Percept. mot. Skills, 38, 70&702. Pizzamigho, L. and Cecchini, M. (197 1) Development of the hemispheric dominance in children from 5 to 10 years of age and their relations with the development of cognitive processes. Brain Res., 31,363-364. Pomeroy, J. E. (1938) The relation of reaction time of five-year-old children to various factors. Child Devel., 9, 281-283. Poppleton, P. K. (1968) Puberty, family size, and the educational progress of girls. Brif. J. educ. Psychol., 38, 286-292. Porter, R. J. and Berlin, C. I. (1975) On interpreting developmental changes in the dichotic right-ear advantage. Brain Lang., 2, 186-200. Pringle, M. L. K., Butler, N. R. and Davie, R. (1966) 11,000 seven-year-olds. London, Longmans. Pyle, W. H. (1925) Nature and development of learning capacity. Baltimore, Warwick and York. Quereshi, M. Y. (1968) Intelligence test scores as a function of sex of experimenter and sex of subject. J. Psychol., 69, 277-284. Ratcliff, G. (1970) Aspects of disordered space perception. Unpublished D. Phil. dissertation, University of Oxford. Ratcliff, G. and Newcombe, F. (1973) Spatial orientation in man: effects of left, right, and bilateral posterior cerebral lesions. J. Neuml. Neurosurg. Psychiat., 36, 4488454. Reichard, S. (1944) Mental organization and age level. Archs. Psychol., No. 295. Reilly, D. H. (197 1) Auditory-visual integration, sex, and reading achievement. J. educ. Psychol., 62, 482-486. Ross, J. M. and Simpson, H. R. (1971) The national survey of health and development: 1, Educational attainment. Brit. J. educ. Psycho/., 41, 49-61. Rowell, J. A. and Renner, V. J. (1976) Quantity conceptions in university students: Another look. Brit. J. Psychol., 67, I-10. Saarni, C. I. (1973) Piagetian operations and field independence as factors in children’s problemsolving performance. child Devel., 44, 338-345. Sandstrom, C. I. and Lundberg, I. (1956) A genetic approach to sex differences in localization. Acta Psyehob, 12, 247-253. Sarason, I. G. and Minard, J. (1962) Test anxiety, experimental instructions and the Wechsler Adult Intelligence Scale. J. educ. Psycho/., 53, 299-302. Satz, P. Bakker, D. J. Teunissen, J., Goebel, R. and Van der Vlugt, H. (1975) Developmental parameters of the ear asymmetry: A multivariate approach. Brain Lang., 2, 17 1-185. Schillcr, B. (1934) Verbal, numerical and spatial abilities of young children. Archs. Psychol.. No. 161. Scottish Council for Research in Education (1967) The Scottish Standardization of WLSC. London, Univ. of London Press. Seashore, S. H. and Seashore, R. H. (1941) Individual differences in simple auditory reaction times of hands, feet and jaws. J. exp. Psychol., 29, 3422345. Shapiro, A. H. (1974) Effects of family density and mother’s education on preschoolers’ motor skills. Percept. mot. Skills, 38, 79-86. Shapiro, S. I. and MOC~Y, B. E. (197 1) Fret recall, subjective organization and learning to learn at three age levels. Psychon. Sci., 23, 189-l 9 I. Shaw,D. J. (1965) Sexual bias in the WAIS. J. consult. Psychol., 29, 590-591.

278

Hugh Fairweather

Shepard, A. H., Abbey, D. S. and Humphries, M. (1962) Age and sex in relation to perceptual-motor performance on several control-display relations on the TCC. Percepf. mar. Skills, 14, 103-l 18. Sherman, J. A. (1967) Problem of sex differences in space perception and aspects of intellectual functioning. Psychol. Rev., 74, 290-299. Sherman, J. A. (1974) Field articulation, sex, spatial visualization, dependency, practice, laterality of the brain, and birth order. Percept. mot. Skills, 38, 1223-l 235. Siegel, L. S. (1971) The sequence of development of certain number concepts in preschool children. Devel. Psychol., 5, 357-361. Siegenthalcr, B. M. and Barr, C. A. (1967) Auditory tigure-background perception in normal children. Child Devel., 38, 1163-l 167. Siegenthaler. B. M. and Bianchi, P. A. (1968) Reaction time ofjaw movement to auditory and tactile stimulation in children. Percept. mot. Skills, 27, 583-588. Silverman, W. P. (1974) High speed scanning of nonalphanumeric symbols in cultural-familially retarded and nonretardcd children. Am. J. ment. Defic., 79, 44-5 1. Simensen, R. J. (1973) Acquisition and retention of a motor skill by normal and retarded students. Percept. mot. Skills, 36, 791-799. Simner, M. L. (1971) Newborn’s response to the cry of another infant. Devel. Psychol., 5, 136-1.50. Simon, J. R. (1969) Reactions toward the source of stimulation. J. exp. Psychol., 81, 174-176. Simon, J. R., Hinrichs, J. V. and Craft, J. L. (1970) Auditory S-R compatibility: reaction time as a function ofear-hand correspondence and car-response-location correspondence. J. exp. Psycho/., 86, 97-102. Simon, J. R. and Rudell, A. P. (1967) Auditory S-R compatibility: the effect of an irrelevant cue on information processing. J. appl. Psychol., 51, 300-304. Singer, J. E., Westphal, M. and Niswander, K. R. (1968) Sex differences in the incidence of neonatal abnormalities and abnormal performance in chiidhood. Child Devel., 39, 103-I 12. Smothergill, D. W. (1973) Accuracy and variability in the localization of spatial target at three age levels, Devel. Psychol., 8, 62-66. Solomons, G. and Solomons, H. C. (1964) Factors affecting motor performance in four-monthold infants. Child Devel., 35, 1283-1296. Sprcen, 0. and Caddes, W. H. (1969) Dcvelopmcntal norms for 15 neuropsychological tests age 6 to 15. Cortex, 5, 170-l 9 1. Stachnik, T. J. (1964) Cross-validation of psychomotor tests for children. Percept. mot. Skills, 18, 913.-916. Stroop, J. R. (1935) Studies of interference in serial verbal reactions. J. exp. Psychol., 18, 643 -662. Struempfer, D. J. W. (1971) Validation of two quality scales for children’s figure drawings. Percept. mot. Skills, 32, 887-893. Suchenwirth, R. (1969) Bedingungen der Handigksit und ihre Bedcutung fir die Klinik der Hcmispharenprozesse. Nervenarzt, 40, 509 -5 17. Sylvester, R. H. (1913) The form board test. Psychol. Monogr., 15, whole number 65. Taylor, A. M. and Warrington, E. K. (1973) Visual discrimination in patients with localized cerebral lesions. Cortex, 9, 82-93. Taylor, D. C. (1969) Differential rates of maturation between sexes and between hemispheres. bncet, I9July, 140-142. Taylor, D. C. (1974) The influence of sexual differentiation on growth, development and disease. In: J. A. Davis and J. Dobbing (eds.) Scientific foundations of paediutrics. London, Heincmann. Terman, L. M. and Tyler, L. E. (1954) Psychological sex differences. In: L. Carmichael (ed.) Manual of child psychology. 2nd edition. New York, John Wiley. Thompson, A. L. and Marsh, J. 1:. (1976) Probability sampling of manual asymmetry. Neuropsychologia, 14, 217--223.


279

Thompson, G. B. (1975) Sex differences in reading attainments. Educ. Res.. 18, 16-23. Tiffin, J. and Asher, E. J. (1948) The Purdue Pegboard: Norms and studies of reliability and variability. J. appl. Psychol., 32, 234-247. Turner, S. and Miller, L. K. (1975) Some boundary conditions for latcrality effects in children. Devel. Psychol., II, 342-352. Tyler, L. E. (1956) Tize psychology of human differences. New York, Appleton-Century-Crofts. UmiltZ, C., Brizzolara, D., Tabossi, P. and Fairweather, H. (1576) Factors affecting face recognition in the cerebral hcmisphcres: Familiarity and naming. Paper presented at “Attention and Performance VII” France, August. Vietze, P., Foster, M. and Friedman, S. (1974) Response differentiation in infants: a sex difference in learning. Percept. mot. Skills, 38,479-484. Vranekovic, G., Hock, E., Isaac, P. and Cordcro, L. (1974) Heart rate variability and cardiac response to an auditory stimulus. Biol. Neonate, 24, 66-73. Vrogh, K. (1971) Masculinity and femininity in the elementary and junior high school years. Deuel. Psychol., 4, 248-253. Waber, D. (1976) Sex differences in cognition: A function of maturation rate? Science, 192, 572574. Wada, J. A., Clarke, R. and Hamm, A. (1975) Cerebral hemispheric asymmetry in humans. Archs. Neural., 32, 239-246. Wccner, P. (1971) Language structure and free recall of verbal messages by children. Devel. Psychol., 5,237-243. Whitman, I:. C. (1925) A brief test series for manual dexterity. J. educ. Psychol., 16, 118-123. Willemsen, F. \?i. (1974) Judgment of the vertical as a function of stimulus characteristics, age, and sex. Percept. mot. Skills, 38, 1344. Wilson, J. A. (1972) Personality and attainment in the primary school: II. Personality structure and attainment of ten-ycar*lds. Research in Education, No. 7. Wilson, R. S. (1975) Twins: Pattcms of cognitive dcvelopmcnt as measured on the Wechsler Preschool and Primary Scale of Intelligence. Devel. Psychol., 1 I, 126-l 34. Winitz, H. (1969) Articulatory Acquisition and Behavior. NW York. Appleton-Century-Crofts. Wolf, V. C. (1971) Age and sex differences as measured by a new non-verbal visual perceptual test. Psychon. Sci., 25, 85-87. Wolff, P. H. and Hunvitz, I. (1976) Sex differences in finger tapping: A developmental study. Neuropsychologia, 14. 35-41. Woolley, H. T. (1914) The psychology of sex. Psychol. Bull., II, 353-378. Wright, D. (1968) Social reinforcement and maze learning in children. Child Devel., 39, 177-183. Wyke, M. (197 la) The effect of brain lesions on the performance of bilateral arm movements. Neuropsychologia, 9, 33-42. Wykc, M. (1971b) The effects of brain lesions on the learning performance of a bimanual co-ordination task. Cortex, 7, 59-72. Yen, W. M. (1975) %x-linked majorgenc influcnccson selected types of spatial performance. Behavior Genetics, 5, 28 l-298. YuIc, W., Bcrgcr, M.. Butler, S., Ncwham, V. and Tizard, J. (1969) The WPPSI: an empirical evaluation with a British sample. Brit. J. educ. Psychol., 39, l-13. Zaner, A. R.. Levee, R. F. and Guinta, R. R. (1968) The development of auditory perceptual skills as a function of maturation: A pilot study. J. Auditory Res.. 8, 3 13.-322.

280

Hugh Fainveather

Dans le cadre des notions courantes sur la lateialisation (Buffcry and Gray, 1972) on a regroup6 les differences sexuelles d’habilete cognitive en trois categories: motrice, spatiale et linguistique. 11 semblerait qu’il existe t&s peu de differences sexuelles qui soient vraiment convainquantes en cllesmemes ou en rapport avec la localisation fonctionnelle supposee. Plusieurs criteres doivent deja etre retenus (nature de I’interaction avec I’age, ordre de naissance, culture, sexe de I’expkimentateur, etc.) mais devront etre completes. On irmet de strieuses reserves sur la proliferation actuelle des investigations dans ce domaine.

Cognition, 4 (1976) 281-302 Q Elsevier Sequoia S.A., Lausanne

- Printed

in the Netherlands

Perceptual tuning and conscious attention: Systems of input regulation in visual information processing*

THOMAS

H. CARR

George Peabody College VERNE

R. BACHARACH

Acadia University

Abstract Literature is reviewed in support of very early perceptual input regulation which can be based on higher-order conceptual or structural stimulus properties. This learning-dependent perceptual tuning takes two forms: preconscious orientation, which operates automatically in the processing of most visual information and can be observed as a strong tendency to parse scenic information in terms of conceptually meaning$ul distinctions, and temporary tuning, in which perceptual biases are modified on a task-specific basis with respect to either scenic information or stimuli whose conceptual properties are arbitrary or conventional such as words as opposed to nonsensical letter strings. Perceptual tuning is contrasted with conscious attention, which is primarily a task selector rather than an input selector.

Attention, or the selection of some inputs and activities at the expense of others, has been documented in a variety of experimental paradigms. At what points in the course of visual information processing major selective mechanisms can operate, and on what bases selection can be made at each point, remain matters for debate. Gardner (1972), for example, has argued

*Preparation of this manuscript was supported in part by NICHD Program Project Grant #5POlHD00973 to George Peabody College. The authors wish to thank Prof. Irving Biederman of the State University of New York at Buffalo for his comments and criticisms on earlier drafts. Address correspondence to Dr. Thomas H. Carr, now at the Department of Psychology, University of Oregon, Eugene, OR 97403, USA.

282

Thomas H. Can- and Verne R. Bacharach

for multichannel or unlimited capacity perceptual processing, with selective attention operating rather late in the system, at the level of short term memory and response production. Sperling (1960), Atkinson and Shiffrin (1968), and Shiffrin and Geisler (1973) have attributed selective functions to the transfer of information out of a non-selective, high-capacity iconic store into a limited-capacity short term memory. Egeth (1967), taking a different approach, has maintained that stimulus recognition is the result of an adjustable hierarchy of tests performed on sensory input. Egeth’s position allows the possibility of perceptual as well as memorial selection processes, as do some other feature-testing models such as Sutherland’s (1959) or Neisser’s ( 1967). Haber (1966), in a review of the effects of set on perception, concluded that selective processes do occur during perceptual intake as well as during response production. Posner, Klein, Summers, and Buggie (1973) agreed, saying that set affects both the early buildup of information in the visual system and the operation of later decision-making mechanisms. Egeth and Smith (1967) and Potter (1975) have both reported perceptual priming of the identification of briefly-presented visual schemes through the advance presentation of a target scene. In addition, Potter (1975) reported priming as a result of advance topical descriptions, though Lawrence and Coles (1954) had previously failed to find such an effect in a different task situation. Finally, Erdelyi (1974), Gibson (1966; in press), and Mace (1973) have claimed that selection characterizes all aspects of information processing, from earliest to latest. Erdelyi argues from a wide range of literature, but especially from work done in the perceptual defense and vigilance paradigms (Bruner and Klein, 1960; Dixon, 1971). Gibson and Mace argue from the theoretical perspective of direct perception, which bears some similarities to the functional phenomenology of Cassirer (1946) and Merleau-Ponty (1962). In response to this controversy, the present paper reviews a varied body of literature dealing with the loci and parameters of visual selective attention. It is maintained that input selection can be found at several points in the course of processing. More specifically, a case is made for very early perceptual selection, called perceptual tuning, which can be based on higherorder conceptual or structural stimulus properties. Perceptual tuning as a preconscious input regulation system is contrasted with conscious attention, which also selects information to be used, but primarily selects tasks to be carried out. It is argued that an automatic form of tuning, dependent upon learning but essentially independent of the perceiver’s intentions, is fundamental to

Perceptual tuning and conscious attention

283

the processing of most visual information. The paradigm example of this long-term tuning or preconscious orientation is a characteristic tendency to parse visual scenes in terms of conceptually meaningful relations among stimulus components. The strength of this tendency is demonstrated by Biederman’s work on the ‘grammar’ of scenic information (Biederman, 1972; Biederman, Glass, and Stacey, 1973; Biederman, Rabinowitz, Glass and Stacey, 1974). As will be seen, the conception of preconscious orientation offered here depends substantially upon Posner and Snyder’s (1975) theory that incoming perceptual information automatically activates internal representations which are habitually associated with it. Beyond a general orientation toward perceiving visual input in terms of conceptual distinctions, more specific perceptual tuning can be instituted temporarily in response to particular conceptual information or knowledge of stimulus conditions which defines inputs needed by a task-oriented perceiver. Temporary tuning can modify the processing of scenic information, changing the relative probabilities of perceptual recognition for certain conceptual classes of stimuli or stimulus components (e.g., Potter, 1975). Temporary perceptual tuning can also modify the processing of stimuli whose conceptual structure is entirely arbitrary or conventional, such as words as opposed to nonsensical letter strings (Carr, Lehmkuhle, Kottas, Astor-Stetson, and Arnold, in press), or character arrays such as the combinations of letters and digits used by Jonides and Gleitman (1972) or E

Cognition, Vol. 4, No. 3

Cognition, Vol. 3, No. 4

Cognition, Vol. 4, No. 4

Cognition, Vol. 3, No. 3

Cognition, Vol. 4, No. 1

Cognition, Vol. 8, No. 4

Cognition, Vol. 2, No. 4

Cognition, Vol. 7, No. 4

Cognition, Vol. 6, No. 4

Cognition, Vol. 5, No. 4

Cognition, Vol. 1, No. 4

Cognition, Vol. 4, No. 2

Cognition, Vol. 2, No. 3

Cognition, Vol. 7, No. 3

Cognition, Vol. 3, No. 1

Cognition, Vol. 9, No. 3

Cognition, Vol. 8, No. 3

Cognition, Vol. 6, No. 3

Cognition, Vol. 5, No. 3

Cognition, Vol. 3, No. 2

Cognition, Vol. 1, No. 2-3

Cognition, Vol. 10, No. 1-3

Cognition, Vol. 6, No. 1

Cognition, Vol. 5, No. 1

Cognition, Vol. 5, No. 2

Cognition, Vol. 7, No. 2

Cognition, Vol. 7, No. 1

Cognition, Vol. 2, No. 1

Cognition Vol. 8, No. 1

Cognition, Vol. 1, No. 1

Cognition, Vol. 9, No. 2

Cognition, Vol. 4, No. 3

Cognition, Vol. 3, No. 4

Cognition, Vol. 4, No. 4

Cognition, Vol. 3, No. 3

Cognition, Vol. 4, No. 1

Cognition, Vol. 8, No. 4

Cognition, Vol. 2, No. 4

Cognition, Vol. 7, No. 4

Cognition, Vol. 6, No. 4

Cognition, Vol. 5, No. 4

Cognition, Vol. 1, No. 4

Cognition, Vol. 4, No. 2

Cognition, Vol. 2, No. 3

Cognition, Vol. 7, No. 3

Cognition, Vol. 3, No. 1

Cognition, Vol. 9, No. 3

Cognition, Vol. 8, No. 3

Cognition, Vol. 6, No. 3

Cognition, Vol. 5, No. 3

Cognition, Vol. 3, No. 2

Cognition, Vol. 1, No. 2-3

Cognition, Vol. 10, No. 1-3

Cognition, Vol. 6, No. 1

Cognition, Vol. 5, No. 1

Cognition, Vol. 5, No. 2

Cognition, Vol. 7, No. 2

Cognition, Vol. 7, No. 1

Cognition, Vol. 2, No. 1

Cognition Vol. 8, No. 1

Cognition, Vol. 1, No. 1

Cognition, Vol. 9, No. 2

Recommend Documents