Idiodynamic associative sets in a clinical population



Journal Title

Journal ISSN

Volume Title



The present investigation had four purposes: 1.To explore the applicability of the concept of idiodynamic associative set as it applies to a typical clinical word association task; 2.To test the rather widespread assumption that there are reliable differences between groups of individuals, according to diagnosed psychopathology, in word association productions; 3.To investigate the effect that passage of time has upon word association responsivity; and 4.To test five more specific hypotheses derived from the literature. The experimental sample in this study consisted of two subsamples of 200 psychiatric patients each, tested at least ton years apart. In each subsample, patients were serially selected to include 50 individuals in each of four major diagnostic groups: Hysterical Personality, Personality Disorder, Psychoneurotic, and Paranoid Schizophrenic. The instrument utilised for all 400 Ss was a word association list considered to be fairly representative of such lists often found, in clinical settings. Sample measures Included both demographic data and 28 word association scoring variables, including both structural variables and 'complex indicators.' This study examined the following hypotheses: 1. The concept of idiodynamlc associative sets is a useful one for distinguishing individuals of widely different psychopathologies. 2. Categories of word association responses can be used to distinguish members of different psychodiagnostic groupings. 3. There are significant differences in word association scoring categories across a time span of approximately ten years. 4. More specifically, it is hypothesized that: A. Paranoid Schizophrenics, since they presumably represent a more disturbed group, score higher than other psychiatric groups on such disturbance measures as Blanks, Clang Associations, Distant Responses, etc. B. Reaction Time relates negatively to Commonality. C. There is a decrease in the frequency of Superordinate Responses over time. D. Men give more Contrast Responses than women. E. Women give more Definitions than men. Factor analysis of the data obtained from the 166 males in this study yielded seven distinct factors. Factors III and VI were general test ability factors; Factor V appeared to be a residual factor of little interest; Factors I, IV, and VII provided evidence for operation of the idiodynamlc associative sets termed dimension-referent, concept-referent, and perceptual-referent in previous studies. Factor II was construed as representing a possible 'pathology of response' set involving heavy usage of unusual stimulus-response combinations. It was also found that one category, Functional Responses, could be broken into two meaningful types, loading on separate factors; Type I was a more nearly abstract response connection, while Type II seemed to involve a more concrete functional imagery. It was judged that many of the differences between diagnostic groups, demonstrated by Student's t-test analysis, could be attributed as much to differences in the sexual composition of the diagnostic groups as to actual difference in diagnosed psychopathology. It was also found that, despite an attempt to restrict the range of variation in age and education of the 400 Ss In this analysis by reducing this number to only 186 Ss, level of education interacted markedly with diagnosis, suggesting that educational differences may have played a major role In the assignment of diagnostic labels for this population. Comparison of early and late time samples demonstrated an Increase in level of education and of Verbal IQ over a ten year period. Early Ss exhibited less variability in the range of the responses, and were characterized primarily by the employment of Definitions and a relatively concrete type of functional response. Late Ss showed more variety of responsivity; their efforts were characterized by Predicates, Contrasts, Logical Coordinates and a more abstract type of functional response. Late Ss also had relatively high Conmonality scores, which was interpreted as being related to the higher level of education. Little support was found for the idea that increases in intellectual level might parallel a developmental sequence of word association responses. As predicted, Paranoid Schizophrenics differed from the other groups on such measures as Unrelated Response, Mishearing the Stimulus Word, and Distant Response. With all Ss, Reaction Time was found to relate negatively to Commonality, Only the Psychoneurotic group evinced dimunition of Superordinate Responses over time. Ken gave significantly more Contrast Responses, as predicted, but, contrary to hypothesis, men also produced slightly more Definitions than women. It was concluded that the concept of associative sets Is a useful one for clinical investigation.



Association tests, Psychotherapy patients