Cognitive process, anxiety, extraversion, and GSR responsiveness in classical differential GSR conditioning

Traditional predictors of conditioning have included level of anxiety and degree of extraversion. Measures of responsiveness in the peripheral modality to be conditioned have also been demonstrated to predict conditioning performance. In addition to these predictors of conditioning, an alternate correlate was assessed comprised of S's cognitive processing of the conditioning situation. It has been demonstrated that Ss capable of accurately verbalizing stimulus relations exhibit differential conditioning while Ss unable to accurately verbalize do not differentiate significantly. The purpose of this study was to investigate the predictive relations of these co-variates to classical differential GSR conditioning. Forty-one males underwent a differential GSR conditioning procedure. A trace paradigm was employed with an 8 sec. interstimulus interval, shock UCS, and two above threshold, easily discriminable tone CSs. The experiment was comprised of three phases: adaptation, acquisition, and extinction. Unreinforced test trials were included. Content-analysis of a postconditioning interview was used to classify Ss as either accurate or inaccurate verbalizers with regard to their ability to correctly conceptualize the stimulus contingencies. Measures of differential conditioning were obtained from early and late periods of the interstimulus interval, and from a period following the onset of shock. Those Ss classified as accurate verbalizers demonstrated significant GSR differentiation on all acquisition and several extinction conditioning measures. Those Ss whose verbalizations were either inaccurate or whose verbalizations omitted reference to stimulus contingencies did not differentiate significantly. Two measures of differential conditioning, obtained from early and late periods of the interstimulus interval were submitted as dependent variables to a stepwise multiple regression analysis. Predictor variables included: a) classification of Ss as accurate or inaccurate verbalizers; b) level of anxiety; c) degree of extraversion; and d) three measures of GSR activity. One reflected nonspecific activity during a preconditioning resting period, a second reflected nonspecific activity during pre-tone periods throughout acquisition, and a third consisted of orienting responses to early adaptation trials. Both accuracy of verbalization and nonspecific activity during pre-tone periods were significantly correlated with the measures of differential conditioning. Neither of the personality variables, anxiety and extraversion, nor the other measures of GSR activity correlated significantly with conditional differentiation. Application of a stepwise multiple regression procedure demonstrated that these personality variables and measures of GSR activity failed to significantly contribute to prediction of GSR differentiation even after removal of the variance attributable to accuracy of verbalization and nonspecific pre-tone activity.