An Abstract: Transitioning Understanding to Behavioral Change

Abstract

This paper questions current social marketing theoretical focus (for a summary of theories used in social marketing, see Truong 2017; Russell-Bennett and Manikam 2016) where dominant focus is on behavior, rather than behavioral change. How can researchers understand behavior change? Drawing on data from a food waste social marketing intervention, this paper showcases application of a dynamic modeling approach, namely, the Hidden Markov Model (HMM). The Hidden Markov Model (HMM) permits change to be examined empirically. Social marketing outcome evaluations are dominated by research methods that assess the behavior of different groups at different time points (cross sectional series design) or track individual behavior over time points using repeated measures focusing understanding on the behavior at each time point for the individuals participating in the evaluation (e.g., Rundle-Thiele et al. 2015 ; Schuster et al. 2015). These approaches limit understanding to group comparisons of behavior at different time points and a focus on explanation and/or prediction of behavior, which is surprising given social marketing’s core aim is behavioral change. Repeated measure data (pre- and post-intervention) is used to identify different states of behavior and determinants of change from one state to another. The HMM tracked the food waste behavior of 244 Australian consumers, 110 of whom were exposed to the intervention. The Hidden Markov Model was applied to examine behavior states, and then using longitudinal data transition between identified behavioral states was examined (desired change, no change, and undesired change). Finally, HMM identifies factors associated with the changes. We find there are two dynamic states of behavior: non-wasters (less than 10% of fruit and veg wasted) and wasters (more than 10% of fruit and veg wasted). One third of wasters became non-wasters after the campaign, a change that is positively associated with an increase in self-efficacy. Results indicate behavioral change is higher among females and those with no private garden. These results suggest that Hidden Markov Modelling (HMM) can be used to identify behavioral states and determinants of behavioral change. Application of HMM extends social marketing understanding beyond behavior to behavior change. This work also contributes to social marketing understanding challenging the social marketing research community to focus attention on behavioral change.