Testing the bipolarity of the Jungian functions

In this article, I offer an alternative explanation for recent studies that have questioned the assumed bipolarity of the Jungian functions measured by the Myers-Briggs Type Indicator (MBTI) due to the failure to observe negative correlations between the hypothesized opposite functions when a Likert format is used in place of the MBTI's forced-choice format (Cowan, 1989; Girelli & Stake, 1993; Loomis & Singer, 1980). It was suggested that the acquiescence response set might be partially responsible. A new instrument was constructed using agree-disagree items. Negative correlations between the hypothesized opposite functions were not observed. To deal with the acquiescence response set, participants were ranked on the number of statements endorsed, and the scores for the middle third of participants were analyzed again For these participants, correlations of -.67 and -.69 between each pair of theoretically opposite functions were found. I conclude thai this largely refutes criticisms of the bipolarity assumption, but there is still reason to develop a measure not relying on the assumption.