Cancer Control Research5R01CA100680-02
A PANEL STUDY OF QUIT SMOKING PROCESSES IN LOW-SES WOMEN
DESCRIPTION (provided by applicant): The main goals of this study are to: (1). Elucidate the smoking outcome constructs inherent to the process of change as postulated in Prochaska and DiClemente (1992) stages of change theory. Specifically, we will assess the relationships and changes over an 18-month period among five smoking outcomes: stage of readiness, motivation, self-efficacy, action, and quitting. (2). Assess how these smoking outcomes are affected by intervening events: changes in pregnancy status and exposure to organized smoking cessation interventions (delivered through health care systems and through television). We will assess direct effects of these events on smoking outcomes, as well as indirect effects through four mediating variables: life stress, social support, social pressure, and health concerns. (3). Assess the nature of the effect of SES conditions on individual smoking outcomes. We will identify which influence paths lead from SES indicators at the individual level and area of residence level to smoking outcomes. We will focus on how SES indicators interact with pregnancy and exposure to smoking cessation interventions in affecting the above mediating variables. We will conduct theory-guided analysis of existing panel data (5 waves over 18 months) from two previous studies that used comparable instruments to evaluate, respectively, a public health clinic smoking cessation program (N=872) and a media-based smoking cessation program (N=1,197) with Iow-SES women. (a) We will review/recode data for comparability and merging. (b) We will complement these data with information about the socio-economic characteristics of the Community Area where subject resided, using available Chicago Community Area and Census publications. (c) We will conduct the multiple analyses needed to answer the above questions, using multilevel structural equation models (in LISREL 8). We will conduct these analyses separately on each of the two sets of data, to take advantage of the information unique to each study and to assess the paths of influence of each of the two types of interventions (health care and media). We will repeat the analyses with the two merged data sets, to assess the individual and area SES effects on smoking cessation processes on a total sample with greater variation on these variables.