Impacts of Public Announcements of Goals and Outcomes on Goal Completion

Physical activity is known to help reduce obesity and the problems that accompany it. Walking has many benefits that make it an excellent way for obese people to increase their physical activity. It is hard, however, for people to stick with walking programs. Setting specific and challenging walking targets that are combined with timely feedback about discrepancies between desired and actual performance lead to higher performance. There is conflicting theory and evidence about the effects of publicly announcing those targets on commitment to the targets and on performance.

Our research question is whether public announcements reduce the challenge level of commitments people set, and whether the reputational consequence of public announcements is severe enough to increase performance. The investigators propose a controlled trial that will integrate Facebook with the objective monitoring of walking via the use of pedometers to test the effect of public announcements on commitments and step counts.

All participants will wear a pedometer and upload via the internet for 14 to 15 weeks, depending on their date of randomization. The first 2 to 3 weeks the investigators will determine the participant's baseline step counts. The following 12 weeks, the investigators will recommend a daily step count target based on the median number of steps the participants walked the previous week. Each week, participants will set a commitment by stating the number of days in the following week that he/she will meet the computer-set step count target.

Commitments will be entered into an interface that will link with Facebook. Participants will select who of their Facebook friends will receive their commitments and/or results via Facebook posts. The interface will distribute those posts in a way that fits in with existing online social practices.

There will be 3 experimental conditions. 1.) No public announcements 2.) Public announcement of commitment and 3.) Public announcement of commitment and results. In addition, groups 2 and 3 will provide us with emails of 3 friends or family members who will act as the participant's support team. The investigators will send announcements directly to support team members.

The investigators plan to recruit 165 participants primarily by obtaining lists of potentially eligible University of Michigan Health System patients from the clinical data repository, and sending targeted letters to these patients. Because this research - and the physical activity programs it will inform - aims to reach a large number of people efficiently, enrollment will take place online with a click-through consent document. The investigators have extensive experience with delivering online walking programs in similar populations.

The investigators hope to learn best practices for using public announcements and online social networks to encourage people to exercise more or to otherwise promote wellness.