The reliability and validity of ipsative and normative forms of the Hutchins Behavior Inventory

The current trend among theorists in counseling and psychotherapy is toward the development of metatheoretical models that can be used to organize, systematically and comprehensively, existing theories and techniques within the discipline. Some models also provide behavior adaptation guidelines for practitioners who wish to adapt their behavior to client behavior patterns.

Hutchins created the metatheoretical TFA System to accomplish the above goals. He also created the Hutchins Behavior Inventory (HBI) to complement the TFA System. The HBI purportedly measures the thinking, feeling, and acting dimensions of behavior upon which the TFA System is built; it thus enables a practitioner to assess the unique, situationally specific, TFA behavior pattern of a client.

At the time of this study, the only evidence concerning the measurement properties of the HBI was for a form that produces ipsative scores (the HBI-I) . But ipsative scores possess inherent psychometric properties that cause problems when they are subjected to certain types of statistical analyses. Thus, in this study, a normative form of the HBI (the HBI-N) was designed. The measurement properties of the HBI-I and HBI-N were then investigated and compared. Reliability was investigated using test-retest and internal consistency procedures. Construct-related validity was investigated using four procedures: internal consistency analysis of HBI-N scores; factor analysis of the items comprising the scales of the HBI-N; an analysis of a multitrait-multimethod validity matrix containing scores from the HBI-I, HBI-N, Strong Campbell Interest Inventory (SCII), and Myers-Briggs Type Indicator (MBTI); and a factorial validity analysis of scores from the HBI-N, SCII, and MBTI.

Results indicated that the HBI-I possesses a high degree of reliability. Prior evidence of content-related validity suggested that the three constructs measured by the HBI are the thinking, feeling, and acting dimensions of behavior hypothesized by Hutchins. Some of the construct-related validity results obtained in this study supported this conclusion, while the main body of results supported the more limited conclusion that the HBI scales measure different, yet to be more clearly identified, constructs. Based on the evidence in this study, the HBI-I seems appropriate for research and clinical use.