Chapter 9

THE SYMBOLIC FRAME BATTERY

INTER-CORRELATION ANALYSIS IN CROSS-CULTURAL RESEARCH

by Hugh M. Lewis

The symbolic frame battery was designed during the course of the field work with the intention of standardizing the elicitation of a series of different symbolic frame tasks, with the aim of implementing basic controls over the administration/elicitation of response of these tasks, and in order to simplify and ensure greater reliability of analysis between individuals and of different tasks by the same individual.

It was deliberately designed for the purpose of eliciting and analyzing cross-cultural differences in the profiles of response patterns of these different tasks across different samples of persons, examining the hypothesis of whether significant differences of average response pattern based on cultural or subcultural levels of analysis do in fact exist, and if so, then of exploring the extent and implications of such difference in terms of cross-cultural comparison, analysis and research.

Given a range of objects to choose from--a miscellaneous collection or assortment of things placed upon a table--the random likelihood that any object chosen should in theory be the same. The fact that individuals will choose the same objects consistently at frequencies that are greater than random points to the influence of a "structural" patterning at some level which, though not directly understood, has in a sense been "indirectly elicited" by the response patterning.

This similiarity and consistency of patterning may indeed be of a relatively low frequency and therefore not clearly salient. But a similar basic design and pattern of response consistency which reveal a normal range and frequency pattern of content, can be carried over into many other kinds of framing tasks.

A great number of correlations have been run on the different samples and this number nowhere exhausts all the possibilities. An "intercorrelational" search was conducted to find those sets of highly positive and negative correlations among a number of dimensions, between different tasks, items and scores on tasks and across the different American, English and Chinese samples and subsamples.

The raison d'etre of many of these correlations are possibly chance or a matter of small, unrepresentative samples. Virtually anything may be correlated with anything else, whether there is an actual relationship or not. Even apparently strong correlations may be rather quite spurious. On the other hand, even relatively low and insignificant correlations can in fact represent important patterns, especially if these recur in certain matrices or intercorrelational structures.

When especially high positive or negative correlations are found, especially over large samples or ranges of data which are quite large, and thus have little likelihood of chance correlation, and when these data sets are united by dimensions or analytical categories, then the resulting "matrix" is regarded as meaningful in some way, even if its significance is not clear. When specific sets of correlations were found consistent across samples, the likelihood of these representing nonrandom associations increases significantly. But what they may still imply remains something of a mystery.

A number of such intercorrelational "matrices" were found to occur in a number of tasks at several levels of correlational analysis--especially at the third or fourth level of analysis when categorical dimensions or cross-dimensions began to be collapsed between the samples. Intercorrelational analysis begins properly at the second level removed from the actual data, in the correlation of correlational scores. While it is possible to go from a primary matrix of relatively low average correlation, to a second or third derivative matrix which exhibits a higher correlation, this "normal" structure may in fact be only spurious--there is no reason in claiming a structure to exist among relationships which in themselves are non-existent.

On the other hand when analysis moves from higher to lower subsequent intercorrelations, as would be expected, we can expect that derivative structure to retain some important aspects of hidden relationships between different dimensions, variables and data.

From such intercorrelational analysis, an alternative means of graphic representation of these structures, as well as a means of testing statistically for the significances and directions of their patterning, have been devised. This alternative form of statistical description should be seen as "possibilistic" rather than probabilistic in the more conventional form of statistics. Patterns which may be of low significance or correlation in some respects may nevertheless reveal hidden relationships in other ways. Together, this form of intercorrelational analysis has proven to be heuristically productive, and allows some feedback and inference to be made between different levels of analysis--even to the level of predicting the likelihood of certain specific types of response, or of infering the higher order significance of such specific responses.

Overall, these correlational patterns suggest the degree and extent of overall similarity, fit and difference between the different samples on the SFB, across its tasks, as well as the ways in which these patterns of similarity and difference occur. It is held that correlational patternings and differences between the samples may represent structural differences between cultural samples.

Within the theoretical design of the Symbolic frame protocol, such associations and complexes of associations are held to be tied to a partial validation/falsification of the theory, as well as to other unknown, and possibly hitherto unexplored relationships.It is felt that this "search" for underlying structure in the response patterns of the symbolic frame protocol can be usefully extended through more sophisticated techniques such as factor analysis. When large numbers of correlations were collected, across the sample "correlations of correlations" were tried out to find out where larger and less obvious relationships may lie in the patterning of response. Such correlational searches were done with almost every task in the battery, as well as between many of the different tasks.

In general it is felt that underlying correlational "matrices" may exist which are not obvious in a superficial examination of the data patterns and which provide evidence for symbolic "structuring" at different levels. Significant differences as well as similarities between these matrices across the different samples may exist which can be clearly discerned primarily through the use of representations in complex multidimensional space. The shared design feature of all these symbolic frame tasks, and their rationale, is that the individual response in "completing" the frame is its systematic "disambiguation" of an essentially "incomplete" and therefore intrinsically ambiguous figure/field/ground relationship in a manner that is acceptable to the individual.

The usefulness of the battery is not in terms of any single set of scores derived from any its tasks, so much as it is in the overall "socio-grid" generated from the unique patterning of profiles of the concatenated scores of all the different tasks. It is felt that if this battery is fine tuned enough, it may allow for the systematic discrimination of response patterns at a number of basic levels.

The scores are to be used as a composite of different samples. Individuals can be expected to vary substantially within different tasks, but the overall profile of the all the scores are expected to show less variation. While the battery is primarily designed for the analysis of grouped differences of response pattern, they have a secondary possibility of detailed individual analysis of response patterning as well. The battery thus allows for an number of multiple forms of analysis of the patterns of response, which may lead to a number of productive interpretations and patterns based on the derived data.

More generally, this battery is designed to solve several problems simultaneously. First it can be considered to be a supplementary research method which can be in its finished form incorporated into the fieldwork agenda of any number of researchers. Secondly, it resolves the inherent paradox that has long plagued anthropological theory and methodology, by pointing up in a statistically significant, empirically repeatable and predictable manner. The patterns of reponse may be unique to a particular category or grouping, hence relative to that grouping, and at the same time may be shared between groups, and hence, comparision of group scores should enable an alternative form of cross-cultural analysis and research which is systematic and relatively etic in form.

Furthermore, the design of the symbolic frame battery itself rests upon certain basic presumptions about the nature of human cognition, the symbolic structure and organization of reality, and of culture which its implementation and refinement may help to resolve and potentially validate at several levels, and which may be productive of new theoretical insights and perspectives as well.

What follows is an analysis of three samples of the first revised form of the SFB as it was given to Chinese people, mostly from the Jetty (N = 35), a small sample of British students (N =14) and of Americans (N =14). The English sample was much more educated than the others, as it was taken from a set of university students in a Southeast Asian studies exchange program, followed by the American and then the Chinese samples, which was relatively uneducated. Though small, the samples are suggestive. They have been split into subsamples of males and females of each group. The Chinese sample was large enough to be split into four subsamples of males over and under 17 and females over and under 16. Two young girls of the American female sample skewed the curve of an otherwise adult sample, and so both sets of scores are included.

The SFB is ordered in a sequence of 9 sets of tasks. The sequence was deliberate, for the purpose of moving from more perceptually oriented tasks through a variety of drawing, color ranking tasks, and basic item tasks, to the final set of inkblots.

The first task were the six MPDT figure-frame images given in reverse order, following "form B" or the "parallel" version of the orginal task (Fuller, 1982: 101-113). Presenting the figure-frame relations in reverse order does seem to make a substantial difference in terms of the sequence of individual scores--subjects in general tend to move from a more difficult to easier figure-frame relationships, as indicated by progressively lower scores, which is a inverse pattern from the original task.

The figure-frame relationships were presented printed on a paper in the upper 1/3 of a paper, with the subject asked to draw directly beneath the image. Several problems of control are resolved in the orginal task involving cards, namely, maintaining alignment to a perpendicular axis of the presented image to the paper, placement and possible crowding or copying of the figures on the same field, and removing any confusion or possibility for error in the presentation of the order or direction of the cards. Results show good correlations with the original samples with the cards, and though requiring revised norms, the efficacy of this mode of administering the task warrants further work.

Responses were scored for enlargement/reduction of the figure, for distortions and separations as well as for minor distortions, for overall direction of rotation across the six drawings, and for the raw score.

Average reduction/enlargement score was highest for the the Chinese male sample (N=6) with an average of 106.25, followed by British female (N=5) and American male (N=5) scores of 98.33, followed by the British Male score of 87.037, and then by the Chinese adult females (N=11) with a score of 83.7, the American female's score of 83.33 (total American female score inclusive of 2 children is 75.5, while the average score of the two American children was 47.92), and then Chinese child female sample (N=12) with an average of 67.0138 and a Chinese child male sample (N=5) of 64.1667.

In general, there is a pattern that Adult males score higher than adult females, and that adults score substantially higher than children. Chinese males scored highest, followed by American males and then English females. British females equal the American males, and scored higher than the Chinese and the American females.

It is not known at this time was reduction/enlargement of the drawn figures in comparison to the presented image may mean, although there is the sense that the "correct" response approaches 100. American Males and British females both most closely approximated this score.British in general most closely approximate this score, followed by the Americans and then the Chinese. Anything above (Chinese males) or below this score represents a form of distortion which may be related to the loss of information between perception and final completion, to the problem of "control" and perhaps to anxiety or shyness.

In terms of combined major distortions/separations of the figures, the English average (N=14) was lowest, .2857, compared to a Chinese average (N=34) of .647 and an American average (N=14) of .57. If we eliminate the two American girls, the average American score (N=12) is .417. Otherwise, there is no clear difference between average scores of adults and children under 17. British males scored higher than females.

A similar pattern appears for the scores for minor distortions, with the British score of 1.643, an American score (N= 14) of 2.714 (discounting the two girls 2.542) and a Chinese score of 2.853. Chinese males with a score of 3.4 did noticeably poorer than Chinese females, with a score of 2.56. American males with an average score of 2.7 scored more poorly than American females with a score of 2.428 (with the two girls 2.72).English males with a score of 1.833 scored more poorly than English females with a score of 1.3. Children 17 and under (N =21) with a score of 3 did more poorly than adults over 17 (N =41)with an average score of 2.56.

The significance of minor distortions is not well known, and the reliability of their scoring is less than that of the major distortions, though there is a strong correlation between these scores and certain forms of major distortion as well as raw scores. Minor distortions may be related to attention to detail (relative carelessness) in drawing, as well as to nervous anxiety.

In terms of direction of rotation, there appears to be little overall difference in either male or female subsamples, with 109 right rotations and 95 left rotations for the women, and 67 right rotations and 69 left rotations for the men. The Chinese males (N=11) show greater left rotations (39) than right rotations, (21). English and American male subsamples individually and combined (N=14) show greater right rotations (37) than left rotations (19)--these differences appear irrespective of age. The chi square test for significance of these total left/right rotations between Chinese men and English/American men is 11.186, which is significant past the .001 level. Women of all the subsamples show similar numbers of left and right rotations.

In terms of salience of total degrees of error in either direction, Chinese males scored a total of 101.5 degrees rotation to the right, and 142.5 degrees rotation to the left.

American men scored 30.5 degrees rotation to the right and 26 degrees rotation to the left, while British males scored 71 degrees rotation to the right and 41.5 degrees rotation to the left. British females have total of 56.5 degrees rotation to the right and 24.5 degrees rotation to the left. American females have a total of 60.5 degrees rotation to the right and 225 degrees rotation to the left. Chinese females under 17 have a total of 120 degrees rotation to the right and 187.5 to the left. Chinese females 17 and over have a total of 112 degrees to the right and 123 to the left. It appears that English overall have more rotation error to the right over the left (127.5 right, 66 left), and Chinese have overall more rotation error to the left than to the right (333.5 right, 453 left). American have 91 right and 252 left, though this difference is largely represented by the female subsample--a distortion largely the result of inclusion of the two young females. Discounting these two females, the total is 70 right and 148.5 left. Chi square tests between British and American totals, English and Chinese and Chinese and American all reveal significant differences between the .001 level.

These differences are represented by the following average scores:

Right Left

English 9.107 4.714

Chinese 9.808 13.323

Americans 5.833 12.375

Unadjusted average raw scores for the different subsamples are: For Chinese as a total sample, 22.817. For English as a total sample, 14.35. For Americans as a total sample (minus the two young girls) 17.05. Chinese males 17 and under, 21.7. Chinese males over 17, 22.583. Chinese females 16 and under, 25.625. Chinese females over 16, 21.36. English males, 12.5; English females, 16.2. American males, 11.3, American females (less the two young girls) 23.07 (with the two young girls, it is 31.72). Chinese males average score is 22.1415. Chinese females average score is 23.4925. These differences can largely be accounted for on the basis of educational achievement, as the American male sample is clearly the most strongly represented in the total number of years in school.

The following represent the correlations between the different subsamples based on the distribution of raw scores per performance the six different items.

Table Appendix 8-1

Item correlations between all of the subsamples reveal the following table of correlations between performance on the different items:

Table Appendix 8-2.

The second set of tasks as presented involved systematic testing the limits of the MPDT responses (Fuller, 1982: 5-24) with a series of yes-no questions for each of the six images drawn to test for perception, a 12 item multiple choice tasks to test for possible problems of perceptual integration, and a 6 item tracing task to test for possible hand-eye coordination problems.

In regard to the first perception task, questions as to the similarity of form and axis of rotation for each of the drawings was asked, and if a negative response was given on the first question, elaboration of the difference of perceived form between the drawn and presented image were sought. Of all the subsamples, only three Chinese males showed any score at all on this set of tasks. There were qualitative differences in the elaboration of response, as both British Males and Females showed much greater care and attention to the detail of the form, indicating not only good perception, but perhaps a heightened sense of critical perception that is akin to being a perfectionist. This qualitative difference was also stronger on average for the females than for the males, indicating a greater cooperativeness and patience in the performance of the task.

The task for indicating a problem with perceptual integration of information reveals an average score of 1.4167 for Chinese females under 17 years, and an average score of 2 for Chinese females 17 years or older, an average score of 1.6 for males 17 years or below, and an average score of 1.167 for males over 17 years of age. This gives a total average score of 1.708 for Chinese females and 1.3835 for Chinese males, and a total of 1.546 for the total Chinese sample. English males have an average score of .667 and English females have an average score of .2 for the females, yielding an average score of the total English sample of .43335. American males have an average score of .5, and American females have an average score of 1.61, with a combined average of 1.056.

The task indicating possible problems with hand-eye coordination reveal the following score distributions, all scores relating to the tracing of only the circle-diamond figures.The average score of the American male subsample was 2.7. That for the American female subsample was 1.44. The average score for the English male subsample was 1.33, and for the English female subsample it was .9. For the young Chinese male subsample it was 1.6, while for the adult Chinese male subsample it was 1. For the young Chinese female subsample it was 1.542, while for the adult Chinese subsample it was 1. The american sample clearly showed the highest overall score, with an average of 2.07, compared with an average Chinese youth score of 1.57, an average English score of 1.115 and a Chinese adult average score of 1. The significance of these kinds of differences are not clear, except that perhaps Americans are spending a little less attention or care for detail. These scores are overall low and minor--to be significant of hand-eye motor coordination problems scores would have to be at least above 3.5.

The third task was a revised version of the rotating frame task, which shows a clear bipolar pattern of response, especially for the British, and to a lesser extent, the American samples, but much more of a continuum for the Chinese sample. Interpretation and design of this task is not unproblematic, and exactly what the response patterns by individuals may indicate is not entirely clear. Revision of this task in future versions of the SFB will resolve some of these difficulties.

For the Chinese males, 4 out of 11 had a score of 12 or less (out of a possible score of 24), with three of these scoring 3 or less, while 7 had a score above 12, with five of these scoring 22 or above.The bipolar nature of this response pattern is clear, with only three intermediate scores. The average score of the young Chinese males was 19.1, while that of the Chinese adult males was 11.667, and an average of 15.383 for the total sample. Fourteen out of 23 Chinese females scored 12 or less, of which 6 scored 3 or less, and 11 scored less than 6. Of the remaining 9 who scored more than twelve, 6 scored 22 or more. The average score of the young Chinese females was 10.5, while that for the adult Chinese females was 11.545, while the total female average was 11.02. The average score for the total Chinese sample as 13.2.

For the English, males 4 out of 9 scored less than 12 and less than 4, while all of the remaining 5 scored 21 or above.Only one of the five English females had a score of 1, while the remainder had a score of 20 or above. The average English female score was 18, while the average of the English males was 12.667. The average score of the total English sample was 15.335.

For the American males, 3 had a score of 1 or less, while two had a score 23 or above. The American male average was 9.6. For the American females, two out of 9 had a score of 12 or less, and only 1 had a score less than 6. Of the remaining 7, 4 had a score of 21 or above, while one had a score of 15 and two of 18. The average for the American females was 16.33. The average of the total American sample was 12.967.

The totals of the females and male averages are compared in the table below:

Table Appendix 7-4

Males Females Total Average

Chinese 15.383 11.02 13.2

English 12.667 18 15.335

Total Average 12.55 15.1167 13.834

Correlations of these averages on the basis of male versus female and total male/female averages reveals the following table:

Table Appendix 7-5

The following table shows nearly perfect positive an negative correlations of average scores between English, American and Chinese subsamples. Though the sample sizes are too small to be genuinely significant, they remain suggestive of possible differences between the subsamples. Chi square comparing raw scores of American and Chinese males and females reveals a significant difference past the .001 level.

Table Appendix 7-6.

There is a sense that many subjects who scored highly on this task viewed the problem as an intellectual challenge of a puzzle of attempting to rotate the rod in the same way as the frame to create correct alignment with the presented figure-frame relation. Regardless of the rationalization of this task in any manner, there appeared to be consistent failure to ignore the context of the frame, or alternatively to view the rod as independent of the rotating frame.

Scores of the degree of enlargement and reduction of the MPDT figures were correlated with the MPDT raw scores, the Rotating Frame raw scores (RF), with cumulative minor/major distortion and separation scores (Error), percetual integration scores (Per.In.) and with hand/eye coordination scores (E/H).

MPDT-E/R. R/F-E/R Error-E/R Per.In.-E/R H/E-E/R

En.Males -.1021 .1565 .165 .385 -.1305

En.Females .59 .1465 .6389 -.25 -.5833

En. Total .2045 .1533 -.0091 .1331 -.2397

Am. Males -.3618 -.3009 .5021 -.4962 -.0591

Am. Female -.3055 -.4216 -.604 .5175 -.4537

Am. Total -.6786 -.5025 -.3325 .1328 0.0181

Yo.Ch.Males -.4241 .1596 -.853 -.49 -.7159

Ad.Ch.Males -.0097 .5158 .3288 .7572 -.6417

Ch.Male Tot. -.0391 .1925 -.0543 -.0543 -.3437

Yo. Ch.Fem. -.4711 -.3685 -.4357 -.555 -.5911

Ad. Ch.Fem. .3742 -.25 -.2324 .7832 -.5196

Ch. Fem.Tot. -.2626 -.1785 -.3087 .342 -.4795

The following sets of correlations are of the intercorrelations of the correlations of these different dimensions:

The following table represents the correlations of these dimensions of correlation of reduction enlargement with other scores across the samples:

Correlation of scores between the MPDT and the Rotating Frame task (RF), between age and the MPDT and the RF, and between education and the MPDT and the RF are represented in the following table:

MPDT-RF Ed.-MPDT Ed.-RF. Age-RF Age-MPDT

English Males .28 .76 .49 .58 .38

English Females .08 ....-.33 .27 .04 -.52

English Total .24 -.39 .44 .23 -.62

American Males .511 -.62 -.047 -.04 -.28

AmericanFemales .367 -.68 -.33 .001 -.45

American Total -.35 .307 .347 .778 -.47

Young Chin. Males -.13 -.76 -.43 -.56 -.3

Adult Chin. Males -.63 -.8 .264 -.14 -.3

Chin. Males total -.92 -.86 .849 .88 -.88

Young Chin. Females -.16 .17 -.42 -.4 .17

Adult Chin. Females .547 -.47 -.56 .78 .56

Chin. Females Total .-.38 -.14 -.38 .356 .097

From the table above there are noteworthy differences of pattern between the different samples. Though it can be seen that there is some agreement in certain areas, there is also some disagreement of scores. Correlations of correlations between the different relationships of scores yields the following table:

Correlations of the correlations between the different samples yields the following composite pattern:

Correlations were also run across the samples between the average Form score (F) per item on the inkblot task and the other measures of education, age, MPDT and the RF.

F-Ed. F-Age F-MPDT F-RF

English Males -.26 ..-.24 -.18 -.2

English Females .95 .99 -.46 .1

English Total .34 .13 -.16 -.11

American Males -.22 -.46 .258 -.76

AmericanFemales .665 .505 .15 -.15

American Total .249 .354 0 -.56

Young Chin. Males .009 .04 .041 -.19

Adult Chin. Males -.23 .302 .037 .099

Chin. Males total -.28 -.24 .044 .212

Young Chin. Females .25 .21 .32 -.46

Adult Chin. Females .086 -.13 -.24 -.32

Chin. Females Total .188 .024 .035 -.327

Correlations of correlations between the different scoring dimensions yields the following table:

Correlations of across the different samples for the different scoring dimensions yields the following table:

The fourth set of tasks consists of a eight color Luscher like rank order in which the 8 colors (red, yellow, blue, green, purple, brown, grey, black)are presented in a pyramid form before the subject who is asked to rank the colors from 1-8, one being the most favorite color.This process is repeated twice, the presented pattern of the colors randomly juxtaposed each time, and the informant advised to make the second rank order independently of consideration of the first.

For the American sample for the first rank order, the most salient first choice was shared by yellow and purple, and second choice color was yellow, followed by blue as the third choice, red as the fourth choice, green as the fifth choice, brown as the sixth choice, grey as the seventh and black as the final choice. In the second rank order, the most salient first and second choices was clearly yellow, with second choice shared by green, followed by blue as the clear third choice, purple and red sharing fourth choice, green as the clear fifth choice, brown sixth, grey seventh and black last. Purple and green were the most ambivalent colors, while red was overall the flattest.

For the English sample for the first rank order, the most salient first choice color was purple, followed by blue and purple in second place, red in third place, green and purple and blue sharing fourth place, red and black sharing fifth place, grey in sixth, brown in seventh and eighth place. For the second rank order, Red and purple were in first place, followed by yellow, blue and purple in second place, red in third, yellow, blue and green in fourth, red and green in fifth, black in sixth, grey in seventh and black in last place.

For the Chinese male subsample in the first rank order, the following order was: 1, purple; 2. red; 3. yellow; 4. green; 5. green; 6. brown; 7. grey; 8 black. The second rank order was: 1. blue; 2. red and yellow; 3. green and purple; 4. green; 5. green; 6. brown; 7. grey; 8. black.

For the young Chinese female subsample in the first rank order, the following order was: 1.blue and purple; 2. blue and purple; 3. red; 4. yellow; 5. green, 6; brown; 7. grey; 8. black. The second rank order was: 1. purple; 2. blue; 3. purple. 4. yellow; 5. green; 6. brown, 7. grey; 8 black.

For the older Chinese female subsample in the first rank order, the following order was: 1. purple; 2. purple; 3. yellow; 4. blue; 5. grey; 6 green; 7. brown; 8 black. The second rank order was: 1. purple; 2. purple; 3. yellow; 4. blue; 5. green; 6. grey; 7. brown; 8. grey.

Table Appendix 7-7.

Americans yell. yell. blue pur. grn. brn. grey black

grn. red

English red yell. red yel. red black. grey. black.

pur. blue blue grn.

pur. grn.

Chinese Males blue red grn. grn. grn. brn. grey. black

yel. pur.

Chinese Females < 16 blue blue red yel. grn. brn. grey. black

pur. pur.

Chinese Females.=> 16 pur. pur. yel. blue grn. grey brn. grey.

It was apparent that the adult Chinese females had the greatest consensus, followed by the Americans and the other Chinese subsamples, while the English had the least amount of agreemnet. The collapsed scores of the three subsamples are represented by the following set of graphs which depict the color landscapes of the collapsed scores (rank 1 + rank 2) of the three sets of samples--Americans, English, all Chinese.

Graph Appendix 7-1. Comparison of Color Landscapes

While these combined scores conflate subsample differences, they also serve to highlight the implict differences between the alternative response patterns of the three cultural samples. The following table shows the color preference sequences of the three samples derived from the graph above:

Table Appendix 7-9. Color Salience Order.

American yell. yell. blue red grn. brn. grey black

English pur. pur. red blue red grey grey brown

grn. black

Chinese pur. pur. yell. blue grn. brn. grey. black

Eight color correlations of the three major samples are represented by the following table:

The following table shows the correlation of correlations of the different color correlations between the three main samples. It is clear that English and Chinese color patterns are clearly correlated, but the American's pattern is quite distinct from either the Chinese or the English.

The fifth set of tasks involves a modified version of the Symbolic Profile drawing task. The different samples were inventories as to the frequency of common items drawn.

The symbolic profile consist of six small squares on the paper, each containing a different little symbol. The subject is asked to draw a picture within each square utilizing the symbol. Since this is a Jungian type task, universal meanings have been somewhat arbitrarily ascribed to each symbol. Since these samples are cross-cultural, this can be considered to be something of a test of the Jungian presupposition of the universality of certain symbols--a suggestion somewhat beyond disproof.

The response patterns of the Americans, English, Chinese Male and Chinese female subsamples are presented for each of the squares, and correlations of response between items and between samples are then sought.

There is a sense of the possibility of sharing beyond random chance alone in both the frequency of the most salient items drawn, in terms of the range of different kinds of items drawn, and possibly in terms of deeper thematic content and underlying basic forms of the objects. There are clear cultural differences in this regard. There are also evident cultural differences in the form and detail of basic things. British draw wheel barrows and rabbits, cottage style houses, Big Bens, action stick figures and sailing craft with portholes. The Chinese draw typical Chinese-style houses mostly without chimneys, as well as fish, rivers, flags and faces with hair and bangs. The Americans draw American "cabin" style houses with chimneys, cars and roads, American style "happy" faces largely without hair, many different small animals, Christmas trees, and snow or rain or hail. There are also apparent qualitative differences in the manner and overall gestalt of the drawings which are difficult to analytically describe--for instance the British draw enlarged figures incorporating the entire square, with largely cartoonish figures, a "scene," a simile or metaphoric description and a sense of space. Chinese drawings are more often smaller, single objects or simple, often disparate, sets of objects with strong stereotypical stylization but without great use of space. American drawings appear more dominated by the underlying sense of geometric form, especially rectangular--they tend to be simpler but more "symbolic" and less "poetic" than the British, but more sophisticated than the Chinese in terms of elaboration and use of space.

There is a sense in some of the squares that the "underlying" geometric form of the symbol of the square thematically "unites" the different pictures drawn within the square, and also provides a basic form which may take many alternate shapes--a round dot can become grapes, marbles, suns, concentric circles, dots on the ends of pencils, holes in walls or tables, while small squares can become elaborated into larger rectangles--computer screens, hallways, boxes, books, puzzles or houses. Items have been highlighted where this seems to be occuring in the following lists, and also items have been underlined which reappear with frequence between different samples or within the same sample.

Square 1. Curved line.

Chinese Females: fish and flag 3 each; river, mouse, eye, hat 2 each; lips, cake, candles, face, hair, tree, sea, string or thread, caterpillar, ice cream, glass, bag with strap, shape 1 each.

Chinese Males: worm 3; eye and flag 2 each; paper, brow, face, waves, 1 each.

English: face and pairs of eyes, 3 each; lips, sea, sun, and sail boats, 2 each; flower, leaf, rabbit, skyline, life-raft, squiggly line, cigarette, 1 each.

Americans: face 5, ocean, 4; waves, sailing boat, 2 each; dophin, snake, fish, car on a road, 1 each.

Square 2--small circle

Chinese females: flower, 4; concentric circles, sun, leaf, eyes, turtles, 2 each; box, hole, grapes, nail with head, umbrella, stick, child's face, snowman, snow, fish, marbles, hills, pencil, 1 each.

Chinese males: sun, face, hair, fish, person, radio, television, flag, flower, chinese javelin, concentric circles, hole in wall, 1 each

English: stick figure 2, cat's face, snow, cloud, spider, wheel barrow, dice, flower, # 5, straw, soccer player, ball, pea shooter, pea, fox head, string of pearls, smiling face, 1 each.

Americans: happy face 3; flower 2; chesire cat, snake, circle, dog, dots, reindeer, snowman, fish, 1 each.

Square 3--diagonal line

Chinese females: house 4; pencil and triangle 3 each; table, 2; line, box, frog egg, ladder, key, umbrella, rain, pot, stirrer, river, tree, ruler, kite, kite tail, hills, tree, 1 each.

Chinese males: triangle 6; pencil 2; house, kite, fly, arrow and arrowhead, 1 each.

English: kite, kite tail, 3 each; park, swing, slide, merry-go-round, fir tree, hill, house, sail boat, water, flag, asterix, fish, big Ben clock, triangle, stick figure, Javelin thrower, javeline, volcano erupting, arrow, 1 each.

Americans: pyramids, 3, house, arrow 2 each, puzzle, slide on playground, face, book, box, square bisected with diagonal, road arrows, mountain with snow, chimney, 1 each.

Square 4--small square

Chinese Females: house, 8; box 4; cup, 2 each, rectangle, robot, umbrella, pencil, fish, face, square, ribbon, cat's face, ground, weeds, moon, 1 each.

Chinese Males: radio, tree, pencil, book, glasses, nose, bell, anchor, stick, helicopter, concentric squares, brick wall, 1 each.

English: house with chimney 4; flower 2; robot's face, face with squares, stick figure, fruit in a basket, leaves , step pyramid, glasses, television, car, throwing dice, match stick, flower pot, shoulder pole for carrying baskets, 1 each.

Americans: square inside a square, 4; checkerboard, face, hallway windows, computer, 3 rectangles of increasing size, wood, blocks, cube, house, chimney, christmas tree with presents, 1 each.

Square 5--up-turned curved line

Chinese Females: face, 10, hair, 5, bangs, 4, tongue, cup, torso, 2 each, bowl, t-shirt, mouth, eye, flower, sun, crescent, dress, round thing, cake, fish, 1 each.

Chinese Males: face, 4, banana, 2, shape, can-cylinder, head, eye, samurai armour, concentric circles, 1 each.

English: face, 5, eye, plam treees, hair, 2 each, lashes, hammock, human figure, rugby ball, cresent moon, sun, wok, stick figure, jackolantern, mug, bird, tie, 1 each.

American: face, 9, snail, reindeer, sun, volcano,1 each.

Square 6.--dot

Chinese females: head, bangs, flower, 3 each; triangle, arrow with point, tree, 2 each, umbrella, pyramid, stick, fish, tadpole, flower pot, shirt, lamp, plug, apple, hill line, clock with hour hand, star, crescent moon, grass, 1 each.

Chinese males: ice cream cone, 2, shampoo bottle, pencil, ribbon, arrow, bow, line dot, sea, fishhook, jet airplane, MPDT dots, raindrops, 1 each.

English: dots, 7; face, 2, sun, dress, concentric spirals, # 4, dice, ball, cat, boat, water, mouse, tail, whiskers, MPDT dots, 1 each.

Americans: Snow/rain/hail, 4; circle around dot, 2, right triangles, house, fire from dot, curving line, dog, four legged animal, christmas tree, dot, mouth shape, window, blob, flower, rectangle, 1 each.

From these lists basic "symbol chains" for each subsample can be constructed based on the most frequently occuring things across the six task items.

Chinese females face 13; house 12; fish and flower 6; pencil, 5; triangle 4 and flag,3.

Chinese males triangle 6, face 5; pencil 4; flag 3; concentric circles, 2

English face 13; house, boats, and suns, 5each; eyes, 4 flower 3; kite, 3.

Americans face 13; animals (dogs, cats, snails, snakes, fish, reindeer) 12; ocean 4, flowers and house,3.

The sixth set of tasks involves 5 different sets of basic items (geometric shapes, basic symbolic shapes, animals, household things, flora and fauna) which subjects were asked to select and rank from 1 to 10, and then to draw lines connecting as many items together on the page by any criteria of relationship of the subject.

Geometric Shapes

There were fourteen basic (square, triangle, circe) and irregular (trapezoid, parallelogram, upside down triangle) geometric shapes.

Americans: 1. horizontal oval, (21%); 2.triangle (21%); 3. square (21%); 4. square (21%); 5. none; 6. circle (21%); 7. pentagon (21%); 8.hexagon (21%); 9. parallelogram (35.7%); 10. octagon (21%).

English: 1. circle (35.7%); 2. none; 3. square, vertical oval and circle, (21% each); 4. hexagon and square (21% each); 5. none; 6. octagon and hexagon, (28.6% each); 7.none; 8. triangle (21%); 9. triangle (21%); 10. parallelogram (21%).

Chinese Males: 1. hexagon (36.4%); 2. circle and horizontal oval, (27%); 3. none; 4. triangle (27%); 5. pentagon (27%); 6 none; 7. none; 8. none; 9. none; 10. vertical rectangle (27%).

Adult Chinese Females: 1. octagon, hexagon, and circle (27% each); 2. none; 3. none; 4. small rectangle (27%); 5. triangle (27%); 6. triangle (27%); 7. hexagon (36%); 8. none; 10. square (27%).

Young Chinese Females: 1. hexagon (41.67%) and circle (33.3%); 2. hexagon (25%); 3. octagon (25%); 4. none; 5. upside down triangle (33.3%); 6. none; 7. none; 8. horizontal oval (25%); 9. none; 10. none.

Basic Symbols

Americas: 1. sun (35.7%); 2. 5 pt. star, (42.9%);3. Christmas tree (28.6%); 4. crescent moon (35.7%); 5. heart (21%); 6. Greek Cross (21%); 7. none; 8 none; 9 none; 10, none.

English: 1. sun (71.4%); 2. 5 pt. star (25.7%); 3. sworl (21%); 4. none; 5. 8 pt. star (21%); 6. 8 pt. star and heart (28.6% each); 7. crescent (28.6%); 8. equal sign. (21%); 9. none, 10. 8 pt. start, (28.6%)

Chinese Males: 1. 5 pt. star and sun, (36.4% each); 2. 5 pt. star (27%); 3. sun (27%); 4. none; 5. 8 pt. star (45.45%); 6. arrow and heart, (27% each); 7. none; 8. none; 9. none; 10; gold symbol (27%).

Adult Chinese Females: 1. sun (36.4%); 2. 5 pt. star (36.4%); 3. none; 4. 8 pt. star (27%); 5. none; 6. swastika (36.4%); 7. diamond; (27%); 8. none; 9; none; 10. MAS logo, (27%)

Animals

Americans: 1. horse and dog (21% each); 2. dog (21%); 3. turtle (28.6%); 4. none; 5. alligator and dog (21% each); 6. bird (42.9%) and cow (21%); 7. none; 8. shark (28.6%); 9. pig (21%); 10. none.

English: 1. gecko (28.6%); 2. elephant (21%); 3. dog and elephant (21% each); 4. cow (28.6%) and turtle (21%); 5. crab (28.6%); 6. turtle and monitor (21% each); 7. cow (28.6%) and fish (31%); 8. rooster, (21%); 9. gecko (21%); 10. none.

Chinese Males: 1. horse and turtle (27% each); 2. crab (36.36%) and cow (27%); 3. none; 4. none; 5 none; 6. elephant (36.36%); 7. fish and rooster (27% each); 8. none; 9. mouse (27%); 10. none.

Adult Chinese Females: 1. horse (27%); 2. fish (27%); 3. none; 4. cow (27%); 5. rooster (27%); 6. bird (27%); 7. none; 8. shark (27%); 9. none. 10. monitor (27%)

Young Chinese Females: 1. horse (41.67%) and dog (25%); 2.none; 3. turtle and elephant (25% each); 4. none; 5. none; 6. none; 7; none; 8. fish and cow (25% each); 9. shark and mouse (33.33% each); 10.snake (25%).

Household things

While some of the samples were probably too small to generate significant saliencies among household things, there are low frequences and a sense that some things are chose more often in any rank than others

Americans: 1. chair, sunglasses, globe, bicycle, and knife, (14.28% each); 2. sunglasses (14.28%); 3. key (21%), valentine (14.28%); 4. stapler, (14.28%); 5. tennis shoe, alarm clock (14.28% each); 6. robot, skeleton key (14.28% each) and globe (21%); 7. tennis shoe and box (14.28%). 8. sunglasses and calendar (14.28% each). 9. ribbon (14.28%); and 10. tennis shoes and book (14.28% each). Things most frequent in any rank include sunglasses (10), tennis shoe (9) chair, globe, open box, swiss army knife and television (7 each); ribbon, key and shower head (6 each); clock, calendar, skeleton key, pot and bird cage (5 each).

English: 1.bicycle (42.86%) and sunglasses (14.28%); 2. globe (21%), swiss army knife and bicycle (14.28% each); 3. globe (21%), swiss army knife and tennis shoe (14.28% each); 4. valentine, bird cage and coffee cup (14.28%); 5.globe (21%), swiss army knife and key (14.28%); 6. stamp, open box and chair (14.28%); 7. clock, fountain pen and tennis shoe (14.28%); 8. tennis shoe (21%), skeleton key, television and open box (14.28%); 9. fountain pen and book (14.28%); 10. chair, stamp and book (14.28% each).

The closed box, two tape dispensers and glue tube and push pins were the only items not chosen at all. Things most salient in any rank order include the globe (12), the bicycle (11), tennis shoe and swiss army knife (9), sunglasses (8) chair and key (7) clock, open box and coffee cup (6), and the book, fountain pen, stamp, and valentine (5 each).

Chinese Males: 1. sunglasses (27%) bicycle and swiss army knife (18%); 2. tennis shoe, bicycle and book (18% each); 3. pot (18%); 4. tennis shoe, closed box and open box (18% each). 5. sunglasses (18%); 6. globe (27%) and bird cage (18%); 7. bicycle and key (18% each); 8. spanner wrench and bird cage (18% each); 9. none; 10. valentine (27%). Things not chosen include the right angle, eraer, two tape dispensers and glue tube. Things salient in any category include the swiss army knife (9) sunglasses and bicycle (8 each), the television (7), tennis shoe (6) and the globe, valentine, spanner wrrench, key, and bird cage (5 each).

Adult Chinese Females:1. television (27%) and ribbon (18%); 2.bicycle (27%) tennis shoe, and globe (18% each); 3. tennis shoe, globe and swiss army knife (18% each); 4. none; 5. bicycle (18%); 6.globe and valentine (18% each); 7. sunglasses (27%) and ribbon (18%); 8. chair, sunglasses, ribbon and television (18% each); 9. clock (18%); 10. stamp (27%) and skeleton key (18%). Things not chose include right angle, book, the two tape dispensers, a tube of glue, pencil and push pin. Items salient in a number of rank orders include the bicycle (10, or 9%), sunglasses (9, or 8%), the globe and television (8 each or 7.2% each), the ribbon, valentine and tennis shoe (7 each or 6.3% each).

Young Chinese Females: 1. bicycle (50%),valentine and closed box(16.67%); 2. none. 3. bicycle (16.67%); 4. sunglasses, television and skeleton key (16.67%); 5. spanner wrench and key (16.67%); 6. none; 7. ribbon and swiss army knife (16.67%); 8. globe and book (16.67%); 9. stamp (16.67) and 10. stapler (25%), stamp and sunglasses (16.67%) The two tape dispensers, glue tube, pencil and push pin were things not chosen at all. Things salient in any rank order include the sunglasses (10), bicycle (9), tennis shoe (7), skeleton key, television and valentine (6 each), and the globe, book, ribbon, stamp, open box and swiss army knife (5 each).

Flora and Fauna

A similar pattern occurs for these things as for the household objects--most samples were to small to generate significant saliencies in most ranks.

Americans: 1.dog (21%) and tyrannosaurus and pine tree (14.28% each); 2.bird (21%), archaeopterix and palm tree (14.28% each); 3. pine tree and dragon fly (14.28% each); 4. dog (14.28% ); 5.spider and coconut palm (14.28%); 6.magnolia tree, horse, and frog (14.28% each); 7.clam shell and starfish (14.28% each); 8. none; 9.horse (14.28%); 10. none.The oak leaf and trilobite the only object not selected at all. Things most frequent in any category include the squirrel, pine tree and dog (7), the horse and tyrannosaurus (6), the starfish, frog, coconut palm and second pine tree (5 each).

English: 1. seaturtle (21%) and fish (14.28%); 2. starfish and clamshell (14.28%); 3. palm tree and saguro cactus (14.28%); 4. saguro cactus and starfish (14.28%); 5.lizard and dragon fly (14.28% each); 6. mushroom and seaturtle (14.28% each); 7.fish and seaturtle (14.28%); 8. magnolia, mushroom, and coconut palm (14.28%); 9. snail (14.28%); 10. crab and pterodactyl (14.28%). The beetle, flower, cactus, acorn, oak leaf, seaweed, peel bug, ant, spider, grasshopper, shark, tuna fish, dinosaur, trilobite, alligator were things not chosen at all.Items salient in any rank indlue the coconut palm (6.36%), the pine tree and the horse (6 each) follwed by the starfish, magnolia and clam shell (5 each). Things most significant in any rank order are the mushroom (10), the starfish (9), the seaturtle (8), the fish (7), palm and coconut palm (6), and the clam shell, second seaturtle, pine tree, saguro cactus, 2nd pine tree and common tree. (5 each).

Chinese Males: 1. tyrannosaurus and horse (18%); 2. fish (27%) and clam shell (18%); 3. shark (18%); 4. seaturtle (18%); 5. frog, snake and seaturtle (18%); 6. horse and fish (18%); 7. bird (18%); 8. none; 9. none; 10. clam shell (18%). Items not chosen include the peel bug, grasshopper, lizard, trilobite, beetle, flower, cactus, oak leaf and beaver cactus.Things salient in any category include the horse and fish (6), the shark, frog, seaturtle, dinosaur, clam shell (5 each).

Adult Chinese Females: 1. pine tree (27%) and coconut palm (18%); 2. magnolia (18%); 3. magnolia (18%); 4. coconut palm (27%) and pterodactyl (18%); 5. mushroom and horse (18% each); 6. none; 7. none; 8. none; 9. dog and clamshell (18% each); 10. horse, starfish and dragon fly (18% each). Things not chosen include flower, cactus, oak leaf, bee, ant spider, grasshopper, lizard, shark frog, snake, seaturtle, dinosaur, bird, alligator and beetle.

Young Chinese Females: 1. tyrannosaurus (25%): 2.dog, horse and dragon fly (16.67% each); 3. pine tree (16.67%); 4. bird and dragon fly (16.67%); 5. none; 6. bird (25%) and tuna (16.67%); 7. dog and frog (16.67%); 8. clam shell (16.67%); 9.none; 10. tyrannosaurus (16.67%). Things not chosen at all include the flower, acorn, oak leaf, beaver cactus, saguro cactus, peel bug, ant, spider and grasshopper.Things salient in any category include the dog (8), the bird and tyrannosaurus (7 each), the the fish (6) and the seaturtle (5).

Though the reasons why different people would consistently choose different things are unknown, it was felt that it may be partly explained in terms of cognitive styles and alternative ways of organizing similar fields of information. Subjects were asked to link together with lines as many things as they wanted on each set on any criteria they wanted to. What emerged are clear consistent individual differences in patterning of the linkages, individual consistencies which are carried over from one task to the next, as well as different overall tendencies for the different samples to link things together in fundamentally different ways.

There are evident at least 5 or six distinguisable patterns of linkages. The first is the replication of a single link between two different and otherwise separate objects. This pattern is repeated across a sample and what emerges is a clear 1 to 2 ratio of linkages to things linked, usually on the basis of similarity of appearance or likeness..The result is a minimum number of linkages and a maximum number of things linked The Chinese, especially the young females and males, but also adult males and females, and someAmericans consistently show this pattern of linkage. Americans sometimes extend this and the second pattern to incorporate every item on the page.

The second pattern of linkage is an extension of this pattern to include short "strings" or chains of linked objects, such that the ration of linkages to items falls from 1:2 to somewhere midway between 1:2 and N: N + 1. Such strings appear more frequently with the American subjects. Sometimes long single strings are created in such a manner, though the obvious relationships between either end of the chain is remote or not apparent. This appears to be a common pattern with the English sample, as well as the American sample.

The third pattern is the closure of such a chain to incorporate things in a close circle of relation ships such that the ratio of linkages to things linked is almost N: N + 1. This pattern is not too common but frequent enough to be noted, especially in small clusters incorporateing 3 or 4 items, sometimes in greater number.

The fourth pattern which is one step further is the creation of a "star cluster" or central "hub" from which more than 1 or two linkages extend to other objects otherwise unlinked. There may be several linkages coming from a single or several items, again the ratio of linkages to things linked falls to N: N + 1.

The fifth pattern which occurs is the sudden interlinking of such star clusters to form enclosed groupings in which everthing is related to everything else, such that the ratio of linkages to things linked suddenly reverses and the total number of linkages quickly becomes exponentially greater than those items linked.

The sixth pattern is the specification of groups by outlines in which unspecified interlinkages between all the items can be implied and summarized. This and the fifth pattern was most frequent with the English, especially the English females.

It is not sure whether minimizing linkages to maximize things linked, or to maximize linkages to minimize things linked is the more efficient strategy. They are clearly alternate approaches to the same problem and lead to different results. It is not known how these patterns may indicate differences of conceptual integration or even mental organization of the brain. There is a sense that grouping things in blanket categories in which everything is simply related to everything else on the basic principle of the grouping may be an efficient way of summarizing a lot of otherwise complex information.

Differences in frequencies of linkages, things linked, and ratios of linkages to things linked have been calculated for the different groups. British have more average linkages than the American or the Chinese, and a higher average ratio of linkages to things linked, than either the Chinese of the Americans. Also the pattern of what kinds of things are linked to others varies considerably between different samples, and shows some significant frequencies within samples.

Subjects had been asked to link things with lines on the five Basic things tasks in order to get at some patterns in how subjects may be organizing things. Clear and different patterns emerged from these elicitations which warrant closer analysis. It was possible to count the number of connections and the number of things connected in a very quantitative form of analysis, and to thereby assist analysis of the different patterns.

Linkages, things linked, and ratio of linkages to things linked were correlated between different items, tasks and samples, and some of these measures were then intercorrelated with age and years of education, as well as with other scores of the SFB, namely with the Form scores on the inkblots, MPDT raw scores, Rotating Frame scores. Correlations were also performed on the saliency patterns of the frequencies of associations to different things on each task, as well as of the relative saliencies of things associated.

Across all of the tasks, the British had the highest average number of linkages per task (12.8) compared to an American total average of 8.81, a Chinese Male total average of 8.11 and a Chinese female total average of 5.97 (Chinese total average was 7.04). These averages reflect well the simple fact of different styles of linkages, ranging from the style typical of Chinese females of 1 linkage to every two separate things, to the American tendencie to form longer "chains" of linkages, to the British pattern of forming "star clusters" and larger groupings in which everything is implicitly connected to everything else. Relatively high correlations were obtained with the scores of the different subsamples were intercorrelated across the five tasks, with .98 correlation or above within the three cultural groupings, and with the lowest correlations betweent these groupings occuring between all of subsamples and the adult Chinese males and adult American males respectively.These differences may be the result of the skewing of the Chinese and American male samples and their small size.

A similar pattern exists for the average number of things connected between the different samples, except that the Americans in toto (12.4) are higher on average than the British (11.6) and both are higher than the Chinese total average (8.425). English males have the highest average (13.2), followed by American Females (12.5). The following graph represents the correlations of numbers of things connected across the five tasks between the different subsamples:

The ratio of linkages to things linked is a better indicator of the differences of patterning between the samples, expressed by the following table:

It can be read at the bottom row that the total English ratio of .97 approximates the N to N average of total connectedness, and the English female ratio of 1.14 actually exceeds this ratio, indicating a tendency to form closed groupings. The Chinese sit at the opposite end of the continuum, with young Chinese females (.65) and Chinese Females in total (.69) tending toward the N to N X 2 pattern of single connections between two otherwise unconnected objects, whereas the American pattern is clearly similar to the Chinese, except that American males (.68) are more like Chinese Females and American Females (.72) are more like Chinese males (.71). Americans in general tend to form short chains of linkages, while these chains become a little longer with Chinese Males--these chains constituting an intermediate pattern between the English and Chinese female pattern with expected frequencies of N to (N + N/2). The correlations of the ratios across the different samples for the five tasks is demonstrated by the following matrix:

These differences mask other important correlations, especially in intersample correlations of frequency patterns of association of items within each of the different tasks.

It is evident that within sample correlations between the three groups (Chinese, Americans and English) are greater than between sample correlations. Overall correlations of individual items between the different samples across the entire 5 tasks are represented in the following table:

Inter-item correlation of the various tasks can also be done within samples as well as across the various samples, for instance the following table represents the interitem correlations across all of the samples for the basic geometric shapes:

According to this diagram, the trapezoid and parallelogram are highly correlated (.9), implying the strong possiblity of salient frequencies of interconnection between them, as are the horizontal oval and the circle (.84), as are the hexagon and the octagon (.84) and the horizontal rectangle and the vertical rectangle (.81). These across-sample correlations can then themselves be correlated with similar within sample correlations to determine the extent of variation from these samples.

Correlations of average linkages, things linked and ratio of linkages to things linked were made with MPDT scores, Rotating Frame (RF) scores, average form scores on the Inkblots and education and age.These correlations are represented by the following tables:

Linkages

MPDT RF Form Ed. Age.

Am. Males -.2 -.68 .4 .01 -.47

Am. Fem. 0 -.36 .539 .244 .432

Am. Total. .012 -.4 .448 .137 .324

Eng. Males -.37 -.43 .12 -.46 -.37

Eng. Fem. .11 .29 .77 .9 .69

Eng. Total. -.2 -.13 .36 -.19 -.14

Yo. Chin. Male .189 .082 -.86 .011 -.02

Ad. Chin. Male -.39 .75 .114 .086 .062

Chin. Male Total. .002 .408 -.35 -.04 -.26

Yo. Chin. Fem. -.411 -.246 -.027 .4817 .4358

Ad. Chin. Fem. .1266 .2483 -.094 -.132 .3429

Chin. Fem. Total -.208 -.038 -.077 .0834 .185

Things Linked

MPDT RF Form Ed. Age.

Am. Males -.19 -.62 .449 .097 -.32

Am. Fem. .125 -.15 .312 .05 .279

Am. Total. .062 -.32 .352 .069 .194

Eng. Males .12 -.42 .04 -.06 -.51

Eng. Fem. -.31 -.4 .73 .73 .72

Eng. Total. -.05 -.47 .24 -.08 .12

Yo. Chin. Male .018 .023 -.72 -.39 -.4

Ad. Chin. Male -.02 .618 .643 .003 -.23

Chin. Male Total. -.01 .366 -.03 -.36 -.31

Yo. Chin. Fem. -.534 .1188 -.095 .1575 .1726

Ad. Chin. Fem. .0808 .1172 -.561 -.287 .3315

Chin. Fem. Total -.282 .0999 -.394 -.182 .0332

Ratio

MPDT RF Form Ed. Age.

Am. Males -.27 -.76 .437 -.31 -.58

Am. Fem. -.21 -.57 .777 .487 .508

Am. Total. -.09 -.47 .571 .2 .401

Eng. Males -.34 -.23 .08 -.4 -.18

Eng. Fem. .48 .39 .48 .67 .39

Eng. Total. -.07 .05 .18 -.2 -.2

Yo. Chin. Male .22 .054 -.67 .127 .085

Ad. Chin. Male -.49 .651 -.18 .23 .078

Chin. Male Total. .011 .32 -.39 .088 -.17

Yo. Chin. Fem. -.042 -.364 .2636 .5201 .5497

Ad. Chin. Fem. .127 .236 .0273 -.086 .2994

Chin. Fem. Total .0311 -.112 .076 .0924 .1318

The following table represents intercorrelation between the different samples across the range of relationships:

Intercorrelation of the different relationships is demonstrated in the following table:

Intercorrelational searches were conducted at a "higher" level of analysis in order to discover any patterns of intercorrelational matrices within the Basic Things tasks. In general, there appeared to be few significant intercorrelations of the five sets of tasks across the differe nt samples, in terms of actual rank order distributions, as shown in the following graph:

Intercorrelation of rank order frequencies across the five samples of the last two sets of tasks combined reveal the following table:

When looked at in terms of gross saliency patterns of items which were chosen at any rank, a much stronger correlational matrix is evident across the different samples for each of the five different tasks as well as for all the tasks combined.The following five tables represent the across the sample correlations for each of the five tasks:

Basic Shapes

Basic Symbols