|| Checking for direct PDF access through Ovid
Policymakers and some scholars have expressed concerns regarding potential links between youth viewing smoking in movies and smoking in real life. Some advocacy groups have expressed the view that causal links between movie smoking and youth smoking definitively exist. However, research on actual smoking behaviors is relatively narrow and correlational, and it tends to produce very small effect sizes. The current article reviews the methodology of movie-smoking studies and analyzes their results meta-analytically. A total of 12 independent data sets were analyzed, which involved youth smoking behaviors as outcome (total participants = 70,788). Meta-analytic results suggested that, due to the large sample size of most studies, results often achieved “statistical significance” despite producing trivial effect sizes. Further, even these trivial effects may be due to widespread methodological limitations including demand characteristics and potential researcher expectancy effects. Recommendations are made for improving this research field.