An online randomised controlled trial of mental contrasting with implementation intentions as a smoking behaviour change intervention

Abstract

Objective: We assessed the effectiveness of mental contrasting with implementation intentions (MCII), an established self-regulatory strategy, as a brief online smoking behaviour change intervention. We expected that MCII would enhance smoking reduction among the highly cigarette dependent because MCII is most effective for challenging pursuits. Design: Participants interested in reducing or quitting smoking were recruited online via Amazon Mechanical Turk. At Time 1, we assessed cigarette dependence using the Cigarette Dependence Scale (CDS-5), then administered one of two brief self-help interventions: MCII (n = 172) or a government-promoted control strategy (n = 174). Participants were invited to complete an online follow-up survey 4 weeks later (Time 2). Main Outcome Measure: At Time 1 and Time 2, we measured recent cigarette smoking with a retrospective, self-report questionnaire. We used these reports to compute smoking reduction scores, with an intent-to-treat approach. Results: MCII increased smoking reduction compared to the control strategy at high, but not low, levels of cigarette dependence. Conclusion: We found preliminary evidence consistent with MCII, delivered as a brief online intervention, as an effective smoking reduction strategy for highly dependent cigarette smokers. Further research is needed on MCII as a smoking behaviour change intervention.

Keywords:

• Mental contrasting with implementation intentions
• smoking reduction
• self-regulation
• intervention
• cigarette dependence
• self-efficacy

Data availability statement

The data that support the findings of this study are openly available in the Open Science Framework (OSF) at http://doi.org/10.17605/OSF.IO/SPF6A.

Notes

Acknowledgements

We would like to thank Maryam Bin Meshar, Sandra Wittleder, and Timothy J. Valshtein for their assistance and support.

Disclosure statement

We have no funding to report, and no interests to declare.

Notes

1 Randomisation specifications in Qualtrics: In the ‘Survey Flow’, we included a ‘Randomizer’ element set to ‘Randomly present 1 of the following elements’: the MCII or the control intervention. We checked the ‘Evenly present Elements’ option, which utilizes block randomisation to present elements—in this case, the conditions—an equal number of times. Qualtrics does not specify the block size, nor does it provide an option for choosing the block size.

2 The T1 survey also included an attention check following the dependent measures, wherein participants were shown a multiple-choice demographic question with instructions that read, ‘Specifically, we are interested in whether you actually take time to read the instructions. So in order to demonstrate that you have read the instructions, please ignore the ladder below. Instead simply click on the continue symbol. Thank you very much’. Several participants later contacted us, concerned, because they had selected an answer choice before they finished reading the instructions, then found that they were unable to undo their selection (i.e., they could switch their answer choice, but not de-select entirely). For this reason, we did not exclude any participants based on attention check responses.

3 Three participants, after they received the online debriefing form, selected, ‘I do NOT feel that I have been adequately debriefed about the nature of the study’. These participants were later contacted through Turk Prime’s internal messaging system by the first author with the opportunity to ask additional questions. None responded.

4 We also included two items in the T2 survey, after collection of the dependent variables, to ascertain whether participants differed in how often they used or thought about the intervention strategy, depending on which strategy they had learned: ‘If you recall, in the first part of this study, you learned a strategy to help you reduce or quit smoking. Then, after 3 days, you received an email with a reminder of this strategy. How often have you used this strategy since you learned it?’ (0 = I don’t remember learning a strategy, 1 = Never, 4 = About half the days, 7 = Every day); ‘How often have you thought about this strategy since you learned it?’ (1 = Never, 4 = About half the days, 7 = Every day). Using independent samples t-tests, we found no evidence for a difference in frequency of strategy use (MCII: M = 3.65, SD = 1.87, n = 127; control: M = 3.36, SD = 1.92, n = 129; t(254) = −1.22, p = .22, 95% CI [−.76, .18]) or frequency of thinking about the strategy (MCII: M = 3.65, SD = 1.69, n = 127; control: M = 3.48, SD = 1.75, n = 130; t(255) = −.82, p = .41, 95% CI [−.60, .25]).

5 We did exclude 12 randomised participants from the final analyses because they did not meet the current smoker inclusion criterion. Ideally, these participants would have been screened and excluded prior to randomisation. However, given that (1) the smoking status measures were taken prior to randomisation, (2) this was a smoking behaviour change study and these individuals were not smoking at T1, and (3) the number of individuals excluded for this reason was equally distributed across conditions (see Figure 1), we are not concerned about this affecting the validity of our intent-to-treat approach.