A Meta-Analysis Examining the Impact Health Communication Campaigns Have On Behavior Change Among Children

Theoretical Background and research questions/hypothesis:  Currently, communication campaigns are a popular method to disseminate health-related information to mass audiences. Snyder (2005) defines campaigns as an “organized communication activity, directed at a particular audience for a particular period of time to achieve a particular goal” (p. 329). Mass mediated health campaigns are campaigns that distribute their health-related message though means such as print media (e.g. brochures, newspapers), new media (e.g. Internet, social networking websites) or television. While previous meta-analytic work has examined the impact mediated campaigns have on behavior change, these analyses are now dated, having collected data over a decade ago. As countless mediated campaigns have been implemented within the last decade, it is necessary to update and expand on existing meta-analytic work. Given this information, the present study seeks to meta-analyze the effects of mass mediated campaigns tailored to a specific population. The present research seeks to determine the overall impact mass mediated campaigns have on behavior change among children. Objectives: The overall objectives of the present meta-analysis is to (a) determine the average effect mass-mediated health campaigns have on children and (b) determine what factors, if any, moderate these effects.

Methods: Multiple procedures were used to identify relevant studies examining the effects of mediated health campaigns on children’s behaviors. Several databases, including Communication & Mass Media Complete, PsycARTICLES and PsycINFO were searched using combinations of twenty relevant search terms. Additionally, all issues of Health Communication and the Journal of Health Communication were perused to find any additional studies. Thousands of abstracts were searched. Once inclusion criteria were set, 13 relevant studies were coded. The Hunter and Schmidt (1994) model of meta-analysis was utilized to calculate weighted mean effect sizes across studies. To test for moderation, campaigns were coded for (1) study authors, (2) year of publication, (3) campaign topic, (4) adoption type, (5) age of participants, (6) time lag between end of campaign and evaluation and (7) effect size.

Results: Analyses found that campaign exposure had a small, but significant, effect on childrens’ health behaviors. The mean weighted effect size was r = .15, p < 0.05, N = 9,232. An additional variable moderated the relationship between campaign exposure and behavior change. Consistent with previous meta-analytic research (Snyder, Hamilton & Huedo-Medina, 2009), time lag between the end the campaign and follow-up evaluation influenced effect sizes such that evaluations conducted directly after campaigns terminated, had higher effect sizes than those completed later on.  No other moderators were observed. 

Conclusions: The present meta-analysis provides insight into the overall effectiveness of campaigns directed at children and various factors that may moderate the effects of campaigns. 

Implications for research and/or practice: The present investigation provides useful information for health communication practitioners. First, this investigation shows that health communication campaigns are an effective way to drive behavior change among children. Second, tests for moderation provide campaign developers with a richer understanding of the variables that may manipulate effects when creating campaigns and evaluations.