Regression of Myers-Briggs Type Scales

Abstract

From a survey of correlational smdies using the Myers-Briggs Type Indicator, Carlyn (2) concluded that the judgement-perception dimension appears to be related to at least one of the other Myers-Briggs dimensions. Correlations and two regression analyses in the current smdy show that the judgement-perception dimension is related to both the sensing-inmition and the thinking-feeling dimensions. Form G of the Myers-Briggs Type Indicator was administered to 188 students in mechanical engineering technology. Their scores were converted to continuous scores by the procedure discussed by Myers (3). Pearson product-moment correlations were calculated among the four Myers-Briggs scales. Significant correlations include values of -.23 (p < .01) between the extraversion-introversion (M = 102.0, SD = 24.89) and the sensing-intuition (M = 85.78, SD = 22.77) scales, .31 (p < ,001) between the sensing-intuition and the judgementperception (M = 96.69, SD = 27.07) scales, and .31 (p < .001) between the thinkingfeeling (M = 83.73, SD = 22.28) and the judgement-perception scales. These significant correlations agree with data from some of the 15 studies surveyed by Carlyn (2) and some of the three studies reported by Bruhn, Bunce, and Greaser (1). Of these 18 different smdies, two reported significant correlations of -.I3 (201 subjects) and -.38 (61 subjects) (p < .05) for extraversion-introversion with sensing-intuition; 17 reported significant correlations ranging from .24 (201 subjects) to .48 (102 subjects) (P < .05, .01) for sensing-intuition with judgement-perception, and seven reported significant correlations ranging from .10 (591 subjects) to .43 (61 subjects) (P < .05. .01) for thinking-feeling with judgement-perception. Correlations in the current study are clearly stable. Magnitudes are small. Forward stepwise multiple regression was used to predict the judgement-perception scale scores from the sensing-intuition and thinking-feeling scales; the multiple R was .42 (&.US = 19.52, p < .001). A second multiple regression included a sensing-intuition by thinking-feeling interaction (M = 7241.26, SD = 2902.26) of scores on these two scales. Only the interaction was significant and the resulting regression of the judgementperception scale with the interaction gave a multiple R of .41 (K.M = 37.69, P < .001). Clearly a comparison of the multiple Rs shows that the interaction of the sensing-intuition and the thinking-feeling scales predicts scores on the judgement-perception scale almost as well as these same two scales predict collectively. REFERENCES