A Theory of Personalised Nudging: Integrating Heterogeneity and Behavioural Science into Political Decision-Making

Mills, Stuart (2020) A Theory of Personalised Nudging: Integrating Heterogeneity and Behavioural Science into Political Decision-Making. Doctoral thesis (PhD), Manchester Metropolitan University.

Preview

Available under License Creative Commons Attribution Non-commercial No Derivatives. Download (4MB) | Preview

Abstract

While successful, behavioural nudges have often been one-size-fits-all, inducing different behaviours from different people despite both people being nudged in the same way. This is called the problem of heterogeneity, and one proposed solution is to personalise behavioural nudges. One area where personalised nudges may be of pertinent interest is the online political advertising space. In recent years, concerns regarding the use of social media sites as part of highly targeted political campaigns have grown. For any personalised nudging programme, this is area of social significance. This thesis investigates two strategies for personalising nudges using an experimental approach. Following an RCT experimental design (n = 962), the effect of impersonal nudges embedded into hypothetical political advertisements are first examined. The first part of this study finds limited evidence that impersonal nudges can influence decision making. In the second part, two strategies for personalising nudges are used to investigate if personalisation renders nudging more effective in this domain. These strategies involve personalising the type of nudge shown to a participant (so-called delivery personalisation) and personalising the outcome which a participant is nudged towards (so-called choice personalisation). Across all personalisation strategies (choice, delivery, and both combined), this thesis finds personalised nudges are statistically significantly more effective at influencing political decision-making than impersonal nudges and not nudging at all. Furthermore, data from the personalisation stage suggests further refinement of this experiment is possible, and so the effects of personalisation may be even greater than observed here.