Dynamic user engagement modelling for mhealth programs
This study will be focused on the physical activity module of a workplace health and exercise program in which the participants track their day to day physical activity using step counts. The current system of flagging step entries uses an arbitrary value in order to identify fraudulent step entries. This cutoff is not a personalized value although the physical activity of participants is likely to vary depending on various psychosocial, demographic, weather, and climatic factors. Moreover, accepting and rejecting flagged step entries based on past performance and reasons provided by participants tends to be subjective. Furthermore, once a flagged step entry is rejected or accepted, these findings are not being taken into consideration in order to identify the genuineness of participants. It is expected that the proposed framework will overcome most of these issues.
According to recent statistics from the World Health Organization, 23% of people aged 18 years and over are not sufficiently physically active. At the same time online health, wellbeing and physical activity programs have become more popular. However, these online programs lack structured statistical and machine learning frameworks for enhancing engagement through program personalisation. This thesis fills this gap by developing a framework for monitoring and predicting future online engagement, including personalised achievable goals for daily step counts and systems for personalised participant intervention by administrators. It identifies new opportunities for online physical activity programs for producing better participant outcomes
For achieving successful results among employees at high risk of poor health outcomes remains a significant challenge for interventions. It is hoped that program developers can use this information to create effective interventions particularly for more sedentary employees.
To provide a truly dynamic personalisd environment what is needed are methods for identifying different engagement trajectories.
According to the global World Health Organization 2017 statistics, more than 80% of the world’s adolescents are insufficiently physically active. In response to this problem, physical activity programs have become popular, with step counts commonly used to measure program performance. Analysing step count data and the statistical modeling of this data is therefore important for evaluating individual and program performance. This study reviews the statistical methods that are used to model and evaluate physical activity programs, using step counts.