Connecting Social Psychology and Deep Reinforcement Learning: A Probabilistic Predictor on the Intention to Do Home-Based Physical Activity After Message Exposure

Abstract

Previous research has shown that sending personalized messages consistent with the recipient's psychological profile is essential to activate the change toward a healthy lifestyle. In this paper we present an example of how artificial intelligence can support psychology in this process, illustrating the development of a probabilistic predictor in the form of a Dynamic Bayesian Network (DBN). The predictor regards the change in the intention to do home-based physical activity after message exposure. The data used to construct the predictor are those of a study on the effects of framing in communication to promote physical activity at home during the Covid-19 lockdown. The theoretical reference is that of psychosocial research on the effects of framing, according to which similar communicative contents formulated in different ways can be differently effective depending on the characteristics of the recipient. Study participants completed a first questionnaire aimed at measuring the psychosocial dimensions involved in doing physical activity at home. Next, they read recommendation messages formulated with one of four different frames (gain, non-loss, non-gain, and loss). Finally, they completed a second questionnaire measuring their perception of the messages and again the intention to exercise at home. The collected data were analyzed to elicit a DBN, i.e., a probabilistic structure representing the interrelationships between all the dimensions considered in the study. The adopted procedure was aimed to achieve a good balance between explainability and predictivity. The elicited DBN was found to be consistent with the psychosocial theories assumed as reference and able to predict the effectiveness of the different messages starting from the relevant psychosocial dimensions of the recipients. In the next steps of our project, the DBN will form the basis for the training of a Deep Reinforcement Learning (DRL) system for the synthesis of automatic interaction strategies. In turn, the DRL system will train a Deep Neural Network (DNN) that will guide the online interaction process. The discussion focuses on the advantages of the proposed procedure in terms of interpretability and effectiveness.

Framing and Tailoring Prefactual Messages to Reduce Red Meat Consumption: Predicting Effects Through a Psychology-Based Graphical Causal Model

“…The procedure adopted for the automatic elicitation of a DBN structure from collected data was essentially the same described in Catellani et al (2021) . Only a concise description of the procedure will therefore be given here.…”

“…Once the preliminary screening was completed, for each selected candidate structure, we computed the m ( i ) metric and the DBN scoring the highest value was selected. Further details about the procedure can be found in Catellani et al (2021) .…”

“…So far, some studies have considered the differential effects of gain, non-loss, non-gain, and loss messages on the promotion of physical activity (Catellani et al, 2021) and adherence to environmental policies (Bertolotti and Catellani, 2014). As to the effects of these four types of messages on the reduction of RPMC, Carfora et al (2020bCarfora et al ( , 2021 have showed that they differentially influence attitude and intention toward RPMC, and that this differential influence depends on some psychosocial characteristics of the receivers.…”

“…According to the Regulatory Focus Theory (RFT; Higgins, 1997 ), self-regulation with a prevention focus involves the avoidance of losses and the fulfillment of duties and obligations, while self-regulation with a promotion focus involves the pursuit of gains and the achievement of an ideal desirable state. Previous research has widely shown that the persuasiveness of a message increases when its framing matches the receivers’ regulatory focus (e.g., Yi and Baumgartner, 2009 ) and this is the case also for messages focused on healthy nutrition ( Catellani et al, 2021 ).…”

The psychosocial antecedents of the adherence to the Mediterranean diet

“…Finally, future studies might explore the possibility of applying this theoretical model to the development of communication strategies useful for promoting the adherence to the MeDiet using machine learning. Based on our findings, social psychologists and engineers might build together a model of a dialogue manager capable of fast profiling recipients and selecting the messages that are potentially most persuasive according to recipients' profiles (51,52) .…”