Gender differences in risk-taking behavior: from family of origin to personality variables

Abstract

Gender differences in risk-taking behavior are reported frequenth in the more current risk-taking literature. The purpose of this study was to create a link between the powercontrol theory's explanation of gender differences in risk-taking (Grasmick, Hagan, Blackwell, & Ameklev, 1996; Hagan, Simpson, & Gillis, 1985, 1987, 1988) in the sociological literature with the theory that Sorrentino, Hewitt, and Raso-Knott (1992) have espoused, i.e., that an individual differences variable, uncertainty orientation, is strongly correlated with gender differences in risk-taking. Because the first theory is based mostly on delinquent risky behavior and the second theory is based on risk-taking in chance situations (a risky choice task), measures of both were administered to 118 male and 146 female undergraduate university students enrolled at Texas Tech University. Other measures administered included the following: Amett's Risky Behavior Scale (RBS; 1994), which includes items related to delinquency (mentioned above), reckless driving, substance abuse, and risky sexual behavior; the Amett Inventory of Sensation Seeking (AISS; Amett, 1994); the short form of the Personal Attributes Questionnaire (PAQ; Spence & Helmreich, 1978); the Rosenberg Self-Esteem Scale (Rosenberg, 1979); and both the projective measure for assessing uncertainty orientation and the adaptation of the Byme and Lam berth (1971) acquiescence-free measure of authoritarianism to assess certainty orientation (Sorrentino, Hewitt et al., 1992). In addition, the same information collected by Grasmick et al. (1996) to test power-control theory also was collected. These questions related to parental occupational patriarchy; parental attitudinal patriarchy; intensity of childhood parental control; and risk preferences. These measures would serve to link this study to existing literature related to gender differences in risk-taking behavior.

Results found both consistencies and inconsistencies with previous research. Congruencies with previous studies were: males preferred higher levels of risk than females on the risky choice task and the RBS, and higher levels of sensation seeking on the AISS. Factor analysis of the RBS yielded four discrete types of risky behavior (risky sexual behavior, reckless driving, delinquency, substance abuse). Males reported engaging in significantly more reckless driving and delinquent behavior. Gender differences were not statistically significant for risky sexual behavior or substance abuse. Family of origin predicted risky preferences for women. The specific relationship between lower patemal control and an increase in women's risky preferences was replicated. Furthermore, a replication of the correlation between variables related to an egalitarian family of origin and increase in women's delinquent behavior was obtained. Overall, the power-control theory did well to predict delinquency in both genders. Correlations between uncertainty orientation and some types of risky behaviors (delinquent behavior, substance abuse, and risky preferences) were found for women, but not for men. Data from the present study yielded no significant relationship between need for uncertainty and risk choice on the probability pairings task, though gender alone influenced behavior in this game as males preferred higher levels of risk than females. Additional findings and implications for future research are discussed.