Date of Award

Summer 2003

Document Type


Degree Name

Doctor of Philosophy (PhD)


Psychology and Behavioral Sciences

First Advisor

Walter Buboltz


Psychological reactance (reactance) is a construct that has begun to attract attention in the past few decades. Reactance is the tendency of a person to react in some way to protect personal freedoms from real or perceived threats (Brehm, 1966; Brehm & Brehm, 1981). Current theorizing and research suggest that reactance is a characteristic related to an interaction between the situation and transient variables such as perceptions of self or available alternatives (Brehm, 1976; Cherulnik & Citrin, 1974; Hannah, Hannah, & Wattie, 1976). Current research and theory indicates that psychological reactance is likely characterological in nature and is more of a characteristic of the person than of the situation (Brehm & Brehm, 1981; Buboltz, Woller, & Pepper, 1999; Dowd & Wallbrown, 1993; Dowd, Wallbrown, Sanders, & Yesenosky, 1994; Seemann, Buboltz, Thomas, Beatty, & Jenkins, 2001). This research has primarily involved trait-factor personality constructs, and only recently has personality style been investigated with respect to psychological reactance (Buboltz, Thomas, Williams, Seemann, Soper, & Woller, in press). The current study focused testing predictions from a theoretical model and from an empirical standpoint based on the existing body of knowledge regarding reactant behavior. A population of male, medium-security prison inmates was sampled. A modified version of the theoretical model of personality and psychological reactance proposed by Huck (1998) was tested with six formal hypotheses, and two hypotheses predicted specific MCMI-III personality scale elevations based on level of psychological reactance (high, moderate, or low). In addition, one hypothesis tested for differences in psychological reactance based on race, and another hypothesis tested the prediction that male prison inmates would demonstrate higher levels of psychological reactance than college students. Participants completed the Therapeutic Reactance Scale (TRS; Dowd, et al., 1991), the Millon Clinical Multiaxial Inventory-III (MCMI-III, Millon, et al., 1997), and a demographic data sheet. A total of 438 participants were retained in the current study. Hypotheses 1 and 2 were not supported, indicating that there is no significant difference in obtained TRS total scores between African-American and Caucasian inmates in the current sample, and that there is no significant difference in levels of psychological reactance between inmates in the current sample and an archive sample of college students. A stepwise multiple regression was conducted to test the theoretical predictions of hypotheses 3 through 8. R2 = .311 with an adjusted R 2 of .306; F (3, 434) = 65.233, p < .001. Of the 14 MCMI-III scales entered into the regression, the aggressive ([sadistic], hereafter, aggressive) (β = .290, p < .001), paranoid (β = .277, p < .001), and borderline (β = .106, p < .039) scales emerged as significant predictors, partially supporting hypothesis 5 and fully supporting hypotheses 7 and 8. Hypotheses 3 and 4 were not supported. A MANOVA was conducted to test hypotheses 9 and 10 with level of reactance as the independent variable. A priori comparisons found strong support for hypothesis 9; the passive-aggressive, aggressive, and antisocial personality styles demonstrated a positive relationship with psychological reactance, and the mean MCMI-III scores for these constructs are significantly different given the level of psychological reactance. Hypothesis 10 predicted a negative relationship between the dependent, avoidant, and schizoid personality styles; this hypothesis not only failed to find support, but the opposite results were obtained.