Human Computer Interactions Adapted to the Personality of Users: Application of Motivation to Physical Activity

The jury will be composed of:

• Élise Lavoué, Professor, Université Jean-Moulin-Lyon-III, Examinatrice,
• Christophe KOLSKI, Professor, Université Polytechnique Hauts-de-France, Rapporteur et Examinateur,
• Elena MUGELLINI, Professor, Haute école d’ingénierie et d’architecture de Fribourg, Rapporteuse et Examinatrice,
• Xavier SANCHEZ, Professor, Université d’Orléans et Université Paris-Saclay, Examinateur.

Résumé de la thèse

Physical activity is a recognized means of preventing non-communicable diseases. It has a positive impact on health and well-being if carried out regularly, with a minimum level of activity. Lack of physical activity has long been recognized and documented as a public health problem, and is of major societal importance, accentuated in recent years by the Covid-19 crisis. It is therefore necessary to motivate people to take part in physical activities that are accessible and easy to implement, such as walking. Many walking interventions with tracking devices, such as pedometers or smartphones, have led to an increase in physical activity, but only in the short term, and the positive effects of these interventions quickly fade. Beyond the novelty effect of a device or mobile application, it is the lack of personalisation and adaptation to the user that seem to diminish the longer-term effects of walking interventions.

In this thesis, we propose a new model of adaptive human-machine interactions for mobile walking motivation. This model is inspired by two psychology theories in particular: the self-determination theory, which describes three fundamental needs to be satisfied in order to achieve motivation and well-being, and the regulatory focus theory, which describes individuals’ action strategies depending on the context, their goal and how to achieve it. Our model also takes into account users’ activity, so as to offer them motivational interactions adapted to their profile.

We implemented this model in an Android mobile application and evaluated its effectiveness in terms of walking performance and motivation through three longitudinal field studies and a design study. The first field study implemented the dimension of adapting a daily walking goal to the user’s performance. Results showed an increase in short-term walking, followed by a decrease in application use and user performance. The second field study added the dimension of adapting motivational messages. The motivational messages were adapted to the users’ regulatory focus profile. The results were encouraging, particularly with regard to the reduction in user amotivation, but we again noted a decline in application use over the longer term. The third field study was designed to evaluate the adaptation dimension of the gamification of our model, replacing the motivational messages. Our gamification takes into account both the self-determination theory and the regulatory focus theory to propose an interaction adapted to the user’s profile.

We look back at the strengths and weaknesses of our contributions, and offer some perspectives on the joint use of these two theories to motivate walking on smartphones.