Incorporation of Texting Bots into Social-Influence Based Fitness Interventions
Open Access
- Author:
- Jonas, Rebecca M
- Area of Honors:
- Information Sciences and Technology
- Degree:
- Bachelor of Science
- Document Type:
- Thesis
- Thesis Supervisors:
- Benjamin Vincent Hanrahan, Thesis Supervisor
Dr. Edward J Glantz, Honors Advisor - Keywords:
- social influence
mhealth
weight loss
texting - Abstract:
- Obesity and overweight are significant health and mortality problems in America. Social support has been established as a helpful tool in causing people to be more physically active but is not always available to people who could benefit from it. Technology has been used to connect people to social support groups through online health communities and texting-based interventions but there are still limitations in any type of human-to-human social support, particularly for people who are focused on weight maintenance instead of weight loss. In this study, I have evaluated four types of social support when provided by a texting bot instead of a person. Participants had texting conversations, based on prompts and background information, with four Wizard-of-Oz bots that I controlled and that each portrayed one type of social support. After each interaction, participants completed a survey about their perceived reasons and excuses for being inactive. After all four interactions, participants completed a survey about their current fitness habits and participated in a semi-structured interview about their experience in the study. When all participants were evaluated as a group, there were no significant findings. When participants were separated into active or inactive participants, there were significant trends in the data. I found that participants who were physically active preferred to receive emotional support, while participants who were not physically active preferred to receive instrumental support. After interacting with the appraisal bot, physically active participants were seemingly more thoughtful in self-evaluation of their fitness, particularly reasons why they are inactive, than after interacting with the other bots (p < .001).