JMIR Publications

ABSTRACT

Background:

Often, online computer-tailored interventions are matched in content to an individual’s characteristics, beliefs, and behavioural factors. This is called content-tailoring. Consequently, such interventions lead to better message processing and a higher likelihood of behavior change, such as smoking cessation. However, a meta-analysis of online computer-tailored interventions showed that effect sizes, albeit positive, remain small – this suggests room for improvement. A promising strategy to enhance the effectiveness of online computer-tailored interventions is to tailor the message frame – how a message is communicated – based on the preferred communication style, in addition to content-tailoring. A factor that determines an individual’s communication style preference is the need for autonomy; while some individuals prefer an autonomy-supportive communication style (offering choice and use of suggestive language), others might prefer a directive communication style, which is replete with imperatives and does not provide choice. Tailoring how messages are presented, e.g. based on the need for autonomy, is called message frame-tailoring.

Objective:

The present study aimed to test the effectiveness of message frame-tailoring based on the need for autonomy, in isolation and in combination with content-tailoring within the context of an online computer-tailored smoking cessation intervention. Primary outcome measure was seven-day point-prevalence of smoking abstinence. Secondary outcomes were perceived message relevance, self-determined motivation to quit smoking, and socio-cognitive beliefs.

Methods:

A randomized controlled trial with a 2 (message frame-tailoring vs. no message frame-tailoring) x 2 (content-tailoring vs. no content-tailoring) design was conducted among adult smokers intending to quit (N=273).

Results:

Structural equation modelling in R revealed that the content-tailored condition increased smoking abstinence rates one month after the start of the intervention (b = .57, P = .021). However, neither message frame-tailoring nor its interaction with content-tailoring did significantly predict smoking abstinence. In our model, message frame-tailoring, content-tailoring, as well as their interaction significantly predicted perceived relevance of the smoking cessation messages, which consequently predicted self-determined motivation. In turn, self-determined motivation positively affected attitudes and self-efficacy for smoking cessation, but only self-efficacy consequently predicted smoking abstinence. Participants in the control condition perceived the highest level of message relevance (M= 4.78, SD= 1.27). However, messages that were frame-tailored for high need for autonomy in combination with content-tailored messages, led to significantly higher levels of perceived message relevance (M= 4.83, SD= 1.03) than content-tailored messages only (M= 4.24, SD= 1.05, P = .003).

Conclusions:

Message frame-tailoring based on the need for autonomy seems an effective addition to conventional content-tailoring techniques in online smoking cessation interventions for people with a high need for autonomy – but not in its current form for people with a low need for autonomy. Clinical Trial: Dutch Trial Register (NL6512 / NRT-6700)