Properties of Possible Selves and the Social Context: Determinants of Risky Behaviors in Adolescents
thesisposted on 25.10.2015, 00:00 by Chia-Kuei Lee
Adolescent alcohol and tobacco use are significant public health problems. Many known family, parent, peer, and intrapersonal predictors are difficult to change and/or are not very discriminating. Possible selves, highly personalized cognitions about the self in the future, have been consistently shown to predict behavior, but have been underexplored in relation to adolescent alcohol and tobacco use. Because possible selves are modifiable, particularly in the formative stages of development, our objective was to determine the prospective influence of several possible self properties on adolescent alcohol consumption and alcohol problems and to determine whether an expected possible self as a “drinker” predicts tobacco use after controlling for alcohol consumption. A secondary analysis of data from 137 adolescents across the transition from 8th grade to 9th grade was conducted. Properties of the total array of possible selves (number of hoped-for, feared, and expected possible selves and the number of balanced hoped-for/feared possible self pairs) and properties of a single content domain (related to alcohol and the most important possible self) in 8th grade were predictors. Outcomes were 9th grade level of alcohol consumption, degree of alcohol problems, and lifetime tobacco use. Controlling for family structure, family cohesion, parental alcohol problems, friend influence, gender, and baseline behavior (all measured in 8th grade), having an expected possible self as a “drinker” predicted higher alcohol consumption and more alcohol problems in 9th grade. In contrast, having many hoped-for possible selves, having a most important hoped-for or feared possible self related to academics, and having a feared possible self as a “drinker” were protective. We also found that having an expected possible self as a “drinker” in 8th grade predicted tobacco use in 9th grade, even controlling for concurrent (9th grade) alcohol consumption, family structure, family cohesion, parental alcohol problems, friend influence, gender (all measured in 8th grade). Findings suggest that possible selves may be an important prospective predictor of alcohol and tobacco use. Interventions to foster expectations in healthy domains, highlight the negative consequences of drinking, and fostering the importance and personal relevance of academics may reduce adolescent alcohol and tobacco use.