EFFECT OF EVALUATIVE CONDITIONING ON ATTITUDES OF SMOKERS TOWARDS SMOKING BEHAVIOUR- THE MODERATING ROLE OF SELF-EFFICACY
This study examined the effect of Evaluative conditioning on attitudes of smokers towards smoking behaviour: With self-efficacy as a moderating variable. Evaluative conditioning was varied into evaluative condition group and control group. Fifty male volunteered undergraduate student of University of Nigeria, Nsukka who were smokers participated in the study. Their age range was between 18 and 30years with mean age of 23.7000. Self-efficacy scale, five dimensions of attitude scales, a Projector, Benson and Hedges, Dorchester cigarette, lighters and five cigarette aversive images were used as materials. Prior to evaluative conditioning manipulation, t-test was used to compare baseline scores on the different attitudes towards smoking between the evaluative condition group and the control group and a non significant difference in attitudes for both conditions was obtained. Multivariate analysis of covariance was used to ascertain whether evaluative condition will affect post scores of different dimensions of attitude using baseline scores as covariates. Result indicates a significant effect of evaluative condition on Attitude towards smoking behaviour: [F(1,43) = 25.3, p<.001], Attitude towards behaviour change: [F(1,43) = 72.8, p<.001], Self-identity towards behaviour change: [F(1,43) = 66.70, p<.001], Perceived control over behaviour: [F(1,43) = 43.7, p<.001] and Intention towards behaviour change: [F(1,43) = 125.2, p<.001]. Moderated logistic regression was further used to test whether evaluative conditioning predicts smoking behaviour and whether self-efficacy moderates the relationship between evaluative conditioning and smoking behaviour. Results indicated that participants in evaluative condition group significantly chose not to smoke more than those in the control condition, – 2 Log Likelihood = 46.39, Goodness of Fit = 4.05 (df = 7, p = .774), ℥2 (3, N = 50) = 22.84, p < .001 and self-efficacy significantly moderated the relationship between evaluative conditioning and smoking behaviour, (Wald statistics = 3.87, df = 1, p = < .05). Lastly, implications, limitations and suggestions were made for further studies.