Cancer Control Research1R03CA126406-01
Smith, Stevens S.
LONGITUDINAL PSYCHOMETRIC INVESTIGATION OF THE WISDM TOBACCO DEPENDENCE MEASURE
DESCRIPTION (provided by applicant): Description: The construct of nicotine dependence (ND) is assumed to underlie tobacco addiction but there is a notable lack of consensus about the nature and optimal measurement of ND. In applied research, ND is often assumed to be a dimensional, trait-like condition with varying degrees of severity although some studies use categorical or discrete diagnostic conceptualizations of ND as found in the DSM-IV diagnosis of Nicotine Dependence. There is increasing evidence that ND is multidimensional. Only recently have psychometrically sound, theoretically-based, multidimensional measures of ND been developed, such as the Wisconsin Inventory of Smoking Dependence Motives (WISDM). The WISDM measures thirteen smoking dependence motives that are hypothesized to influence dependence processes such as inability to quit, nicotine withdrawal, and relapse. The WISDM subscales reflect a variety of theory-based internal factors (e.g., tolerance) and external factors (e.g., social and environmental goads to smoke) that permit investigation of different smoking motives in clinical and experimental contexts. The primary aim of this study is to investigate the nature of ND by assessing measurement and structural invariance across gender and racial/ethnic groups and time as well as investigating the predictive, concurrent, convergent, and discriminant validity of WISDM subscales. This will be accomplished by conducting a series of secondary analyses on datasets from two observational studies and two smoking cessation clinical trials that administered the WISDM to culturally diverse smokers (White, African American, and Hispanic). Latent variable (e.g., multigroup confirmatory factor analyses) and other models will test for invariance of key model parameters (e.g., factor covariances) across gender and racial/ethnic groups as a means of evaluating similarity of multidimensional models across the groups. Validity analyses will include multivariate linear and logistic regression models as well as latent variable modeling that will allow complex tests of individual WISDM dimensions or sets of dimensions in relation to smoking and relapse. Observed measures such as cigarettes smoked per day, latency between first cigarette and regular use of cigarettes, and end-of- treatment cessation success will be related to WISDM measures in these validity analyses. The goal of all of these analyses is to characterize dimensions that represent core features of ND that influence smoking behavior as well as cessation and relapse. Lay Description: This study will investigate how various theoretical dimensions of nicotine dependence, as measured by the 13-factor Wisconsin Inventory of Smoking Dependence Motives, relate to one another and to important smoking- and cessation-related outcomes and whether these relations are consistent across gender and racial/ethnic groups.