Search

James says the history of philosophy is largely the history of a certain clash of temperaments. Boring suspects that a methodological difference he finds among psychologists stems from temperamental differences—but he leaves such differences largely “unreduced.” Murphy speaks of the “basic temperamental or emotional incompatibility of the promachine and antimachine theorists.” Following their lead, the present paper asks whether anything would be different about the history of Western thought, including the theorizing of contemporary scholars and scientists, if nobody to this moment had produced the equivalent of a Rorschach movement response. The paper's answer involves (1) suggesting differences between movement and all other Rorschach response dimensions, (2) illustrating the significance of these suggested differences, (3) applying the foregoing to theorizing in scientific psychology, and (4) suggesting commonality among the intellectual controversies of ancient and contemporary thinkers.

Much of the literature on learning disabilities which has dealt with the child's perception of space has seemed to view problems in spatial perception as primarily lack of accuracy or veridicality, e.g., poor organization or distortion of assumed spatial “givens.” This paper examines other factors which are important in the child's spatial functioning, and which can be helpful in understanding and remediating problems in the learning problem child. Specifically, this paper outlines a distinction between topological versus Euclidean-geometric conceptions of space, action-function space in contrast to perceptual space, and impersonal space in contrast to interpersonal space.

Once in each of the last three decades, at least one personality theorist has colnmenced on psychology's general indifference to cognition concerned with the f~~cure (Murphy, 1936; Rapaporc, 1946; Allporr, 1955). In this paper I should like to suggest that the picture is changing-that there is in progress a redefinition of intelligence to include foresight and planning. I should especially like to suggest in addition chat the Rorschach test has from its inception provided us with a crude measure of apric~tde in this area. Much of the paper will deal with cognitive attributes of individuals scoring high in Rorschach movement categories. Porteus, of course, has long used the words foresight and planning in connection with his Maze Test. But since his 1950 pronouncement chat "it is time to come out boldly for the inclusion of planning in OLI~ statement of the nature of intelligence" (p. lo), he has been echoed, intentionally or not, by workers in diverse areas. Singer and Herman, for example, suggest that "the ability to defer immediate motor response . . . with the consequent resort to . . . fantasy gives man a control of his future through imagery and planning" (1954, p. 329). Shoben has included "the capacity for foresight and the self-imposed control of behavior through the anticipation of its ouccorne" within his "basic

Previous research with the variable spontaneous change (Means, 1965) has demonstrated a moderate inverse relationship (-.43) with California F Scale for 23 college students as well as a moderate positive relationship (.46) with the number of activities 26 VA hospital mental patients felt free to perform. The present study (Means & Harper, 1968) attempted to see whether spontaneous change was also related to jndependently sustained cognitive activity, measured by voluntary time in a think tank. Spontaneous change is thought to reflect a natural tendency to change ideas and concepts relatively independently of external contingencies such as rules, schedules, and a complex changing environment. Thus, it was reasoned that persons scoring high on spontaneous change would voluntarily keep themselves reinforced for cognitive activity when given the opportunity to think about anything they liked in a large "empty" room (think tank) for as long as they wished. For 20 undergraduate students from the introductory psychology class who fulfilled a 5-hr. course requirement, spontaneous change was measured as follows. Ss were asked to arrange 4 unequal-length lines on 10 separate 10-in. circles of filter paper. They were instructed to arrange these lines in any manner they wished, to have fun, and to arrange lines so that they liked each arrangement. The arrangements were scored by taking the differences in displacement of angles and distances from the sample lines, summed across drawings for a single spontaneous change score for each S.' After Ss had finished the line arrangements they were taken to the think tank, which was a large lecture hall (seating capacity of 100 students). Ss were instructed to think about anything they wished for as long as they liked, up to 5 hr. Each S was in the room alone and was free to leave if he became bored at any time after a M-hr. commitment. Paper was provided for those who chose to use it. There was a moderate positive product-moment correlation of .58 (p < .01) between time spent in the think tank and spontaneous change on the line-arrangement task. This result, although based on a small N of 20, suggests the importance of cognitive styles with respect to independently pursued activity in a rather unstructured situation.

Summasy.-University srudent volunteers, 47 women and 46 men, took a bartery of G spatial tests and checked for problems in a parallel form of each test a list of possible problem-solving strategies. Significant sex differences on reported strategy used for 3 of the tests were found. Evidence suggests that the women were less efficient than the men in their use of frequently used strategies and used more guessing and concrete solution styles. Sex differences in spatial ability, especially after puberty, have been welldocumented (Sherman, 1967; Maccoby, 1966). Both a genetic (Stafford, 1961, 1963, 1965; McClearn, 1967; Garron, 1970) and an environmental basis for such differences have been postulated. The question, though, of che exact nature of these differences has not been answered. It is apparent that tests as complex as the spatial tests may have functional indeterminacy, i.e., they may elicit different problem-solving styles among different people, and these problem-solving strategies may be differentially efficient. The observed mean difference may indicate that men and women may be using different problem-solving methods or men and women use the same methods, but men are more successful in solving prob- lems. The present study was undertaken to exp!ore sex differences in strategy use on spatial tests in an effort to understand the underlying basis for observed differences. Ss (47 women, 46 men), all volunteers from an introductory psychology course at a major university, took a battery of spatial tests taken from the Kit of Reference Tests for Cognitive Factors (French, et al., 1963) representing three major spatial factors: Spatial Orientation, Spatial Scanning, and Spatial Visualization. Tests were given in two 1%-hr. sessions. The tests were: Card Rotations Test, Cube Comparisons Test, Choosing A Path, Map Planning Test, Paper Folding, and Surface Development Test. Immediately following each test Ss were given a brief, untimed set of similar problems chosen from the parallel form of each test followed by a list of from 14 to 20 possible suategies to be checked if used on specific problems in the brief test. The questionnaires were adapted for these tests from a similar set developed by Frederiksen (1969) and under his direction. Possible strategies ranged from very test-specific methods, e.g., "Rolling Over" on the Card Rotations Test, "I thought of the cubes as objects and mentally 'rolled them over' in making comparisons," to very general strategies that could be used on any test, e.g., guessing or using a "systematic approach."

Previous studies of perceptual causality have involved the visual modality directly or indirectly through associative imagery. The question addressed in this paper is whether causal impressions will occur in those never experiencing vision, i.e., in the early blind. A qualified answer of “yes” may be given. Fixed sequences of haptic stimulation “causal” and “non-causal” in type evoked the same reports from early blinded as from late blinded and blindfolded but normally sighted. Also, different semantic structures appear to be applied by the early blind, as contrasted to late blind and sighted persons to phenomena that are primarily (but not solely) visual in nature. It was concluded that the reports of causal interactions given by the early blind do not represent verbal habits based on vision but reflect causal experience unrelated to vision.

If hypnotizability is essentially role-taking aptitude, as asserted by Sarbin and Coe (1972), and if the cognitive determinants of hypnotic ability are to be found in S's attitudes, motivations, and expectancies and in his ability to think along with and vividly imagine the suggestions he is given, as Barber (1972) has stated, then the use of suggestion to implement new role patterns which do not fit the traditional hypnotic paradigm should be a fairly simple procedure. If the perception of one's own awareness is a subjective phenomenon or an epiphenomenon, as behaviorists have long maintained, then the total number of possible role patterns involving "altered states" (or altered experiences) of consciousness which may be induced by means of suggestion should be roughly equivalent to the number of definitions of such states which it is possible to conceive or to imagine; for each of these imagined definitions may be communicated to a responsive S in the form of suggestions which define for him the nature of the changes which he is about to undergo in the perception of his own awareness. It is commonly known that the traditional "hypnotic" or sleeping trance behaviors were first manifested by people who were imitating the behavior of the retarded Victor Emmanuel, who was too stupid to realize that he was supposed to go into convulsions while standing under a "magnetized" elm tree on the estate of the Marquis de Puy&gur and went to sleep instead (Gibbons, 1973, p. 5). However, Ss who possess a high degree of role-taking aptitude and who are able to think along with and imagine vividly the suggestions which they are given should easily be able to adopt new forms of rrance-like behavior without requiring another Victor Emmanuel to model the appropriate responses for them! Considering the wide variety of induction methods currently employed, the absence of any reliable physiological correlates of hypnosis, and the speed with which hypnosis may be re-induced, the only essential prerequisites for a suc'An earlier version of this paper was presented at meetings of the Society for Clinical and Experimental Hypnosis, Newport Beach, Calif., December, 1973.

To test the concept of middleness, 4 series of 7 problems each consisting of 3, 5, 7, 9, 11, 13, and 15 blocks were administered to 402 male and 381 female children between the ages of 36 and 155 mo. The task required S to select the middle block and the results showed that this ability began to emerge at the 3-yr. level, developed rapidly through year 6, and became asymptotic by the eighth year. A wide variance in the ability was exhibited but perfect performance on all of the problems was not found until age 13½. The use of the task as a test of cognitive development is discussed as well as alternate variations involving blocks of different colors, asymmetrical arrangements, paper-and-pencil forms, and verbal reinforcement.

Summary.-3 experiments were performed to clarify the relationship of two principles of cognitive control to [he recall of different types of stimulus material. Exp. I shows that there is little or no relationship between leveling-sharpening and recall of thematic material and no relationship to recall of sequentially presented, highly similar designs. Exps. 11 and 111 show that, as predicted, field-articulation or field-dependence is related to the recall of similar visual designs but not to the recall of thematic material. The relationship of individual differences in cognitive control to recall has been studied only to a limited degree, so that results obtained thus far are in need of clarification. The relationship of field-articulation or field-dependence to recall, for example, has remained incompletely resolved because of apparently conflicting results (cf. Witkin, Dyk, Faterson, Goodenough, & Karp, 1962). Gardner and Long ( 1961) and Long (1962) have presented evidence that measures of this important dimension of individual differences can be used to predict accuracy of recall under particular conditions (e.g., when verbal learning materials are highly ~imilar, so that a high degree of selectiveness of attention is required for unconcarninated recall of individual items registered). The relationship of leveling-sharpening to recall (see Gardner & Lohrenz, 1961) has also remained partially unspecified. The present paper reports results of three experiments designed to clarify the relationship of field-articulation and leveling-sharpening to recall of two types of stimulus material: (a) thematic material and (b) highly similar visual designs. Ic was predicted that neither of the two cognitive controls would be related to the recall of thematic material and that field-articulation, but not leveling-sharpening, would be related to recall of highly similar visual designs. METHOD Sab jects

The research reported in the present paper took as its poinc of deparmre Heinz Harunann's theory of ego psychology (4). According to Hartmann's formulations, certain ego processes are congenital and conflict-free in origin, that is, they do not originate as adaptations to ego-id conflict. Other ego processes, while originating as adaptations to conflictual situations, may become conflict-free and "automatized." An automatized ego process functions without conscious awareness of its functioning. It was the purpose of the presenc research to investigate two such unconscious, automatized ego processes. The unconscious nature of automatized ego activity suggested that studies of behavior without awareness might provide a source of informacion and suggestions of techniques suited to the study of the present problem. A study by F~threr and Eriksen (3) focused on the unconscious perception of verbal stimuli. When Ss were presented with stimuli at a level where "Ss can report the presence of the stimuli but cannot identify che stimuli" (i.e., a detection threshold), Fuhrer and Eriksen found no evidence that Ss could perceive meaning in the stimuli presented.

This paper deals with the epistemology of diagnosis. It is suggested that the conventional view of diagnosis, which sees the diagnostician as one who describes and/or explains pathological phenomena, is inadequate. A contrasting model is proposed, one that conceptualizes diagnosis as a prescriptive and problem-solving activity. The difference between these two models is explored with particular reference to the problem of defining “illness.” The relevance of diagnosis to perception and cognition is also discussed.

Summary.-96 kindergarten Ss, randomly assigned to 4 experimental conditions, were tested as to the effect of type of cue, picmre or spelling pattern, and number of cues, single or combination, on the initial and later recognition of 4 consonant-vowel-consonant (C-V-C) words when these words were presented in trigram pairs. The ScheffC method, used to make post hoc multiple comparisons among the means, showed that picture cues neither facilitated nor hindered recognition whereas spelling-pattern cues significantly aided acquisition (p < .01) but hindered retention (9 < .05) when cues were deleted. The number of cues presented had less effect than did type of cue. Among questions that concern reading theorists are those that relate to the type of cue most favorable to beginning reading, the number or combination of cues, and the effect of these cues on remembering words at a later time when cues are no longer present. One approach has been to make available a large number of cues and the child may take from this number those which he finds helpful. Recently, linguistic basal readers have appeared, the proponents of which say that a large variety of cues serve as distracters to the learning of words (Bloomfield & Barnhart, 1961; Fries, Wilson, & Rudolph, 1966; Rasmussen & Goldberg, 1965). The child's attention should be directed only to the word and to the letters of which the word is composed. A spelling pattern approach in which the child says the names of the letters and then the word is one method of doing this. Pictures, especially, have been criticized. The authors of the Merrill LingtliJtic Readers say that pictures have been excluded as a matter of principle in order to focus the pupil's attention upon the reading material itself (Fries, Wilson, & Rudolph, 1766). Rasmussen and Goldberg ( 1965 ) of the SRA Basic Reading Series feel that pictutes may hinder the discovery of important sound-spelling relationships and should never be used to teach word recognition. Also pertinent is whether the type of cue and the number or combination of cues used in the original learning are related to remembering at a later time. A prevalent view is that the more nearly alike the retention session is to the acquisition session, the higher will be the retention score. The reduction of cues during retention may be equated with a situation in which a smaller number of cues is given in both acquisition and retention (Blackman & Holden, Yhis paper is based on the author's doctoral dissertation, written under the direction of C. I. Chase, and submitted to the Graduate School of Indiana University.

This paper reports an investigation of autokinetic response tendency based upon a conceptualization of the autokinetic phenomenon as being rooted in attention deployment. The hypothesis that field-dependence-independence and the autokinetic phenomenon require a similar type of attention deployment was tested in a multiple regression design. Results suggest that the autokinetic response requires the ability to attend selectively to relevant aspects of the external environment while ignoring the spatial context in which they are embedded. Attention to internal stimuli was found to be unrelated to the autokinetic response tendency.

Summary.-The cognitive style "tolerance for unrealistic experiences" is often measured by the lower threshold and range of the apparent movement illusion in phi phenomenon. Some of the assumptions underlying the criterion measures are systematically explored. It is demonstrated that there is little difference in the lower threshold whether Ss are told that the apparent movement is an illusion or that it is a "real" motion. Upper threshold was more variable: it was affected by the instructions, and those 5s whose upper thresholds were influenced by the instructions were more field-dependent. Klein, Schlesinger, and Gardner (Klein & Schlesinger, 1951; Gardner, el nl., 1959; Klein, et al., 1962) used phi phenomenon as their criterion task for the cognitive attitude, "tolerance for unrealistic experiences." Rate of alternation of two flashing lights was gradually increased, and E measured the point at which Ss first reported illusory movement as well as the point at which the illusion ceased. Klein and his co-workers always informed their Ss in advance that the two lights are actually stationary, and the impression that one figure is moving back and forth is an illusion. They asserted that the "tolerant" S accepts his experiential impression of motion even when he knows it is at variance with reality, and so reports the illusion of motion earlier (lower initial threshold) and continues reporting it longer (higher upper threshold and broader range). S who is "intolerant of unrealistic experiences" is more impressed by the "real" information and so reports a more restricted movement illusion. The primary aim of the present report was to consider the function of "real information" by evaluating the effect of the instructions given Ss. Initial and upper thresholds were compared when Ss were told the motion is illusory (the Klein instructions) and when told stimuli are moving and non-motion illusory. As Ss' performance on phi phenomenon seems related to their reactions to other ambiguous experiences (cf. Frenkel-Brunswik, 1949), it might correlate with individual differences in imaging, perceptual, and reality testing tasks. Therefore, all Ss were also given tests of imaging and relaced cognitive styles. PROCEDURE Forty-two young women, aged 17 to 21 yr. and all student nurses at Lenox Hill Hospital School of Nursing, served as paid volunteer Ss. In a 2Y2- to 4-hr. session, each S was given a series of tests: the embedded figures test to measure field-articulation (Witkin, 1950), a paper-and-pencil object sorting test to measure equivalence range (Clayton 'We thank George S. Klein, whose ideas formed the basis of this study; also Thomas C. Todd who assisted in the experiments. The experiments were supporred by NIMH Grant No. 5-R01-MH-06733. Our grateful thanks to Miss Joan Jennings, present Director of the Lenox Hill School of Nursing, for her generous cooperation with us.

Stdmmary.-The new McCarthy Scales of Children's Abilities were analyzed in terms of Guilford's structure of intellect model to evaluate the consonance of this model with the structure of McCarthy's scales. The motor tests were analyzed in terms of Guilford's psychomotor matrix. These analyses showed a great deal of consonance between Guilford's systems and McCarthy's scales. A comparison of the abilities assessed by the McCarthy scales, Stanford-Binet, WPPSI, and WISC indicated that the McCarthy and Stanford-Binec measure similar percentages of the various structure of intellect operations, contents, and products but that both batteries assess the operation of evaluation to a lesser degree than the Wechsler scales. The McCarthy Scales of Children's Abilities (MSCA) were developed by Dr. Dorothea McCarthy to assess the mental and motor abilities of children in the 2Y2- to 8%-yr. range. She constructed a wide variety of tasks and then grouped these tasks into six scales: Verbal, Perceptual-Performance, Quantitative, General Cognitive, Memory, and Motor. To determine the placement of each task on one or more scales, McCarthy used rational considerations based on her vast clinical experience, and she considered the results of factor analyses of a portion of the standardization data (McCarthy, 1972, p. 2). The main purpose of this paper is to evaluate the consonance of McCarthy's scale structure with Guilford's ( 1967) structure of intellect model by analyzing each component cognitive test in terms of Guilford's system. Such analyses should demonstrate, for example, the degree to which Verbal tasks contain semantic content, Perceptual-Performance tasks contain figural content, and Memory tasks assess the operation of memory-as defined by Guilford. The analyses should also show the types of abilities measured by the McCarthy scales, permitting comparisons with the abilities capped by other tests for children such as the WISC, WPPSI, and Stanford-Binet. PROCEDURE The McCarthy scales contain 18 tests: 15 which assess cognitive ability, including 2 which also assess motor coordination; and 3 which are noncognitive and measure gross motor coordination. The 15 tests of cognitive ability were categorized in terms of the structure of intellect model based on Guilford's (1967) and Meeker's (1969) discussions of the types of tests which measure different operation-content-product trigraphs (see below for explanation). Of particular value were Meeker's classifications of WPPSI, WISC, and StanfordBiner items according to the $dimensional model. Since a number of tasks in

A pencil-and-paper test for the assessment of individuals' capacities for cognitive innovation was developed. The construct of “cognitive innovation” refers to the process whereby new organization is introduced into the cognitive structure. The test consists of 40 figures which S is asked to interpret in clever and unusual ways. It has satisfactory reliability and appears to be a potentially useful instrument in studies concerned with creativity and reactions to novel stimuli.

With the aim of appraising the generality and construct validity of the concept of equivalence range, defined as the preferred span of objects or qualities that an individual is willing to consider in one category, a number of cognitive tasks were devised to appraise different facets of styles of categorizing behavior. Several sorting tasks were employed, some original and some adapted from previous research, to sample the breadth of an individual's preferred mode of categorization. These included an object sorting test employing real objects, object sorting tests substituting the names of objects for real objects, tasks in which photographs of people and descriptions of people were sorted, and a drawing sorting task. In addition several perceptual “bandwidth” tasks were especially developed to study Ss' characteristic range of equivalence judgments and categorizations. Also included were tests of criticalness response style and associational ease to appraise the relevance of styles of verbal categorization to other possible facets of equivalence range. Several measures of accuracy in size constancy, and brightness judgments, previously reported by Gardner to relate to sorting behavior, were included in the battery. All tests were administered to a total of 60 female Ss, the measures inter-correlated, factored, and rotated to oblique simple structure. Of the seven factors extracted, three were considered sufficiently large and clear to be interpreted meaningfully. The first and the largest of these was identified as Sorting Equivalence Range. All sorting tests loaded this factor, but none of the other tasks were substantially represented. The other two factors were identified as Conceptual Bandwidth and Associational Ease or Lack of Criticalness. The following conclusions appear warranted. (1) A general stylistic consistency identifiable as Sorting Equivalence Range, represented in a wide variety of sorting tasks and content, appeared as the first and largest factor. Sorting behavior thus showed consistency across content areas, but was not closely related to tests loading on either of the other two factors identified: Conceptual Bandwidth or Lack of Criticalness. Indeed, three rather distinct dimensions were identified. (2) Contrary to published research, performance on size constancy accuracy tasks was found to be largely independent of consistencies in categorizing behavior. No test reported by Gardner (1953) to relate to categorizing behavior loaded the first factor. (3) Group-administered paper-and-pencil object sorting tests developed by Clayton and Jackson (1961), received high loadings on the Sorting Equivalence Range factor, and may in the future be substituted for the more costly individual form. (4) The identification of a factor of Associational Ease or Lack of Criticalness, related to verbal ability, suggests that it may be profitable to explore more fully cognitive styles in intellectual productions. Variables commonly identified as response style or set (Clayton & Jackson, 1961; Jackson & Messick, 1958; Frederiksen & Messick, 1959) bear examination in relation to perceptual and cognitive styles. (5) Further research employing large samples and a broad array of hypothesized measures of equivalence range is needed, both to cross-validate present findings, and to clarify further the nature of dimensions of cognitive styles in categorizing behavior.

The cognitive theory of psychophysics proposed in an earlier paper by Baird claims that the ratio of exponents obtained in the method of magnitude estimation for 2 stimulus attributes is equal to the ratio of information transmitted in the method of absolute judgment for those attributes. This theoretical prediction was tested experimentally for visual length and area. Results partially supported the prediction, inasmuch as the ordinal relations among exponents (length higher than area) was the same as the ordinal relation among information measures. However, in one instance, the exact quantitative relationship was significantly different from the predicted value.

Summary.-The purpose of this study was to test whether persons of global-passive cognitive style would be more influenced in their affective view of the future by a momentary experience of success and/or failure than active-analyric Ss. 60 university srudents were tested for cognitive style and subjected to one successful, failure, or neutral experience. The simple experience of success or failure in a laboratory task appeared to have, at least temporarily, a diffuse effect on the global-passive Ss' affective view of the future. The purpose of this experiment was to consider the effect of a present experience of success or failure on an individual's affective view of the future. The question about the effect of success and/or failure on a person's affeccive view of the future is complicaced by the factor of individual differences, in the sense that people experience the present (specifically success and failure) in different ways. Research by Witkin and his colleagues (Witkin, Dyk, Faterson, Goodenough, & Karp, 1962) has yielded evidence for a continuum which describes peoples' experience and organization of perceptions of the world. These perceptual modes (accive-analytic/global-passive) were related to broad aspects of psychological functioning and experiencing. Active-analytic Ss presented a more differentiated view of the world, a clearer and more developed conception of their body, a greater sense of separate identity, a capacity for directed activity and less susceptibility to environmental flucniations than global-passive groups. It was hypothesized that global-passive Ss would manifest greater susceptibility to the influence of immediate success and failure feedback than the active-analytic group. This means that the affective view of the future of Ss having an active-analytic disposition should be less susceptible to the momentary experience of success and failure than the view of Ss having global-passive dispositions. METHOD There were two experimental conditions, the experience of success and the experience of failure. There was also a control condition in which Ss received no feedback. Within these groups, assignment of Ss was controlled for perceptual style (active-analytic or global-passive) by means of Witkin's Rod-andframe Test (Witkin, Lewis, Hertzman, Machover, & Wapner, 1954), giving a 2 X 3 experimental design. Experience of success or failure was induced using a technique of Nuttin 'Now at The Devereux Foundation, Devon, Pa. The work described in this paper was carried out at the Psychologisch Instiruut, University of Louvain, Belgium (Director, Prof. J. Nuttin) while the second author was a Nuffield Foundation Social Science Fellow, and with the aid of a grant from the Carnegie Trust for the Universities of Scotland.

In an attempt to account for sex differences in rod-and-frame test (RFT) performance two theoretical notions have been advanced. The most recent theorizing is contained in an extremely thought-provoking paper by Broverman, Klaiber, Kobayashi, and Vogel (1968). They refute what they have called the "cultural reward hypothesis" as a means of accounting for sex differences in a variety of cognitive tasks. For them, sex differences in the RFT are "reflections of the balance between the activating influence of central adrenergic processes and the inhibitory influence of central cholinergic processes, which, in turn, are influenced by gonadal steroid 'sex' hormones" (p. 27). While Broverman, et u1. (1968) have generated a tenable hypothesis to account for sex differences in the RFT, it is difficult to conceive of an adequate test of that hypothesis. Although the second hypothesis is by no means unique to the author, Vaught (1965), in a study specifically designed to account for sex differences in the RFT, showed that Ss regardless of biological sex, whose role identification was in the feminine direction made more errors in the RFT than those Ss whose role identification was in the more masculine direction. It was reasoned that the atuibutes accompanying each cultural role accounted for the differential RFT performance. For example, masculinity has long been associated with mechanical and scientific interests, and with preference for abstract and impersonal activities over aesthetic ones. When the female becomes interested in mechanical tasks and ocher pursuits requiring the ability to deal with spatial relationships, she would not only be identified as relatively masculine but she would also tend to make fewer errors in the RFT. Additional support for this hypothesis had come from a recent study by Barrett and Thornton (1967) who found engineers to be significantly more field independent than a sample of college students. Further support for the so-called "cultural reward hypothesis" has come from correlations between the Embeddedness sub-scale of Evans' (1969) Psychological Differentiation Inventory and RFT performance. Thirty-six male and 39 female college students produced correlations of .18 and .52 respectively. Only the correlation for women reached statistical significance (p < .01). Females who answered items such as "Can you easily separate a figure from its background?" in the yes direction, were more likely to make fewer errors in adjusting the rod to true vertical. It seems likely that the manifest content of the eight items of the Embeddedness sub-scale, reflect prior knowledge and experience with spatial relationships not ordinarily associated with the feminine role identity. Thus, females who have this background have an advantage in the RFT and that advantage shows up in scores in the direction of male performance.

The concepts of field dependence and psychological differentiation have provided an extremely valuable link between theories of personality and psychopathology and laboratory research on perception and cognition. However, confusion between test and construct and failure to control adequately for intelligence have made it difficult to assess the limits of the concept of differentiation. Further, basic understanding of the nature of the individual differences observed has been hampered by insufficient efforts to distinguish between personality differences stemming from differences in ability and those stemming from differences in adaptive choices and strategies. The present paper examines these issues as they are reflected in the current literature and attempts to point toward a more comprehensive approach to research in this area.