The study period will last 8 months, consisting of a one- to two-month intervention, and a 6-month follow-up falling after the end of the pre-committed contracts. Six-months is used as the definition of a “long-term follow-up” in a recent Cochrane review on rewards for smoking cessation, and thus we view this as an appropriate time period for follow-up. Given the difference in contract start date across treatments, endline will either be 6 or 7 months after an individual’s commitment contract ends, depending on their treatment assignment.
Respondents recruited into the study will be randomized into one of four study arms. Respondents in the first treatment arm, the commitment group, will be offered a deposit contract. As part of the contract, respondents will be asked to pay their own money to the study team account, on the condition that they will only receive their money back after a month if they can verify that they have quit smoking with a biochemical test at a Quest diagnostics centre.
Respondents in the second treatment arm, the pre-commitment group, will be asked at baseline if they want to agree to start a deposit contract with the same terms as the first treatment arm, but starting in a month’s time. The third treatment arm is a combination of the first and second arm. Respondents in the third treatment arm will be offered a deposit contract at baseline. Then, any respondents who decline the offer, will be asked, in a replica of the situation in the pre-commitment group, if they want to agree to start a deposit contract with the same terms as the first treatment arm, but starting in a month’s time. The fourth arm will serve as a control group, and respondents in this arm will receive no offer for a deposit contract.
To address the question on whether a lack of information on the effectiveness negatively drives take up, and ultimately hinders smoking cessation success, we will cross randomize all treatments with an information module. The information module will be brief and contain easy-to-digest statistics on the effectiveness of commitment contracts for smoking cessation among individuals who agree to such contracts. The most relevant statistics from which to draw this information come from Halpen et al (2015) which tests the effects of commitment contracts with a large sample.
All study respondents will take part in a baseline at the start of the study and an endline survey 8 months after the start of the study. Further, any respondents who have a deposit contract will be asked to verify their smoking status at the end of their month long commitment period. All surveys, and requests for smoking status, will be administered remotely through Qualtrics, which can be accessed on a computer or a smartphone via the internet. As noted above, being able to access the internet will be an inclusion requirement for the study. All study payments will be processed through Paypal, and all payments into deposit contracts will also be made via Paypal.
All respondents will be asked their smoking status at endline. We will request that any self-reported claim of smoking cessation is verified biochemically with a urine anatabine test at a Quest diagnostics lab. All respondents who self-report that they have quit smoking will be paid for visiting the lab, and arrangements will be made with Quest to ensure that respondents will not be responsible for payment when attending the lab. The study team has already discussed the necessary set up with the Quest diagnostics team and we do not see this being a barrier to study success.