Some conceptual parameters and their relationship to inferential set behavior

Decisions concerning the operational specification of 'conceptual complexity' involve two important choice points. The first is whether the characteristic, conceptual complexity, is of sufficient scope to incorporate both 'interpersonal' and 'impersonal' referents. Explicit in such decisions is the choice's relevance to social perception behavior. The second is whether or not response analysis leading to the inference of 'conceptual complexity' should be unidimensional or multidimensional. Several systems of measures of conceptual behavior presumably concerning conceptual complexity Bieri's 'Cognitive Complexity' (1965), Mayo's 'Cognitive Complexity' (1959), Tuckman's 'Integrative Complexity' (1965) and McGaughran's 'Closed-Open' and 'Public-Private' conceptual dimensions (1954) were employed to test three main hypotheses: (1) Different measures of conceptual complexity within themselves provide a statistically reliable score, regardless of whether the concepts are elicited in response to either 'impersonal' or 'interpersonal' referents presented as stimuli; (2) conceptual behavior is correlated with maturity of 'social perception,' as defined by Jones and Thibaut (1958); and (3) a preponderant commonality exists among the different measures of conceptual behavior, supporting the idea that conceptual complexity is describable as essentially unidimensional. The subjects were 60 randomly selected female college students with a mean age of 21.43 years, S. D. = 6.06 years. They were sufficiently comparable so that the control variagles of Age, Hours of College Credits Completed, Academic Year, Verbal Fluency and Manifest Achievement demonstrated no consistent relationship with performance on either the conceptual complexity, or the social perception variables. Each subject was administered the control, social perception and conceptual complexity measures, the last in randomized order to control for position and practice effects. The statistical technique testing hypotheses 1 and 2 was Pearson's Product Moment Coefficient of Correlation (r). Hypothesis 3 was tested by use of a factor analysis employing the principle axis method. Eight factors were extracted and rotated to simple structure by the Varimax technique. Analysis of the data revealed that the interpersonal and impersonal forms of the Bieri, Mayo and McGaughran measures (for both dimensions) were, for the most part, highly intercorrelated. Thus, Hypothesis 1, that different measures of conceptual complexity within themselves provide reliable scores, regardless of the 'object' or 'person' character of the given referent, was supported. The second hypothesis, that conceptual complexity is correlated with maturity of social perception, was supported selectively. That is, only certain of the measures, (mainly the Mayo and the Tuckman), and certain of the social perception response classes, demonstrated the postulated relationships. Analysis suggested the value of redefining the social perception variables (response classes) presently used in future research, and the use of more heterogenous populations. Finally, the factor analysis revealed the preponderance of factors obtained to be generally technique-specific. No factors were loaded significantly by any substantial proportion of the measures currently used. Hypothesis 3, that conceptual complexity can be adequately represented by describing a unidimensional parameter was less supported by the present data than the alternate conception that conceptual complexity involves a number of discrete constructs. That is, the analysis did not yield a large first factor, or even one on which more than two of the indices were loaded significantly.