Experimental Design Details
For this study, we have designed a Randomized Controlled Trial (RCT) to be carried out in the field in collaboration with Friskis & Svettis (F&S) Uppsala in Sweden. F&S is a not-for-profit sports club, with classes and gyms. F&S Uppsala has facilities at three locations in Uppsala, and approximately 28,000 members.
From a pre-study in September 2015, we have learnt that all respondents – current F&S Uppsala members as well as guests and other non-members – would like to exercise more frequently than they did the past month. (On average, the respondents would like to exercise 0.46 times more per week.) Hence, it is relevant to invite also current members, not only new members.
Invitations to participate in the study will be included in a newsletter, which is emailed to all members, including those who are currently inactive. Additional promotion may be done with posters at the three F&S facilities in Uppsala and or an ad or banner at F&S Uppsala’s homepage.
Those who agree to participate will first be asked to fill out a baseline questionnaire, including a set of questions on current exercise habits and desired/target exercise habits/goals. The baseline questionnaire also includes questions on age, gender, educational attainment, origin, household income, their purpose with exercise, reasons for not exercising as desired, trust in own knowledge about how one should exercise to take care of own health, how one should eat with the same purpose, length & weight, an index of self-control, and time preferences (impatience and consistency).
The participants will be randomly divided into three groups: i) control, ii) goal iii) incentives. Those assigned to the control group will keep exercising as usual. The only further intervention for those will be a follow-up questionnaire administrated to all participants after two to four months.
Those randomized to the goal and incentives treatments will be offered a tailored commitment contract. By tailored is meant that the frequencies of exercising covers their desired/target exercise habits/goals as expressed in the baseline questionnaire. The idea behind the tailored commitment contracts is the theoretical finding by Bhattacharya et al (2015), that increased exercise may not be welfare improving.
The difference between the goal and the incentives treatments is that subjects will be able to attach economic incentives to the commitment contract in the incentives treatment. The economic incentives will be completely self-funded. Half of those offered to sign a contract with economic incentives will get a nudge about the effectiveness of stakes. Those randomized to an offer to sign a commitment contract with economic incentives, regardless of nudge or not, may choose any amount between SEK 10 and SEK 1,000. The amount is to be deposited via Swish. If the commitment contract is fulfilled, the money is repaid; else it is donated to a preselected charity or other organization. Each participant choosing to sign a deposit commitment contract also chooses an organization that will get the money in case they fail to comply with their contract. The sequence is that those in the incentives group first choose a contract, if any, and then are asked if they would like to attach economic incentives to it.
Across all treatment arms, default frequency of exercise in the offered commitment contract will be randomized to either the respondent’s own target frequency or a higher frequency. Contract length is between 1 and 4 months. Across all treatment arms, those signing a contract will be randomized to no default, or a default contract length of four months (preselected, but possible to alter).
The control group gets no offer to sign a commitment contract, and obviously no economic incentives. They only respond to the initial questions (after agreeing to participate), then we follow-up on their exercise habits at the end of the intervention, and ask them to fill out a follow-up questionnaire at the end of the study.
After filling out the follow-up questionnaire, all participants are given the choice between a TRISS scratch card (retail price 30 SEK) or a donation of 30 SEK to a charity of their choice (same list of charities to choose from as in the incentives treatment) to thank them for participating.
Both the baseline and the follow-up questionnaire include questions on weight (and length, so we can calculate BMI), sleep and other stress related variables, as well as questions on smoking.