Using Online Computer Tailoring to Promote Physical Activity: A Randomized Trial of Text, Video, and Combined Intervention Delivery Modes

Abstract

Website-delivered interventions are increasingly used to deliver physical activity interventions, yet problems with engagement and retention result in reduced effectiveness. Hence, alternative modes of online intervention delivery need to be explored. Therefore, this study aimed to evaluate the acceptability and effectiveness of a computer-tailored physical activity intervention delivered on the Internet in 3 delivery modes: video, text, or both. Australian adults (N = 803), recruited through e-mail, were randomized into the three delivery modes and received personal physical activity advice. Intervention content was identical across groups. Repeated measures analyses of variance were used to compare the three groups regarding acceptability, website usability, and physical activity. Participants in the video group accepted the content of the physical activity advice significantly better (F = 5.59; p < .01), and spent significantly more time on the website (F = 21.19; p < .001) compared with the text and combination groups. Total physical activity improved significantly over time in all groups (F = 3.95; p < .01). Although the combination group increased physical activity the most, few significant differences between groups were observed. Providing video-tailored feedback has advantages over the conventional text-tailored interventions; however, this study revealed few behavioral differences. More studies, examining alternative delivery modes, that can overcome the limitations of the present study, are needed.

Acknowledgments

The Population Research Laboratory at the Central Queensland University, managed by Christine Hanley, was responsible for participant recruitment and data collection using online surveys.

Notes

^a No differences between intervention groups were observed at baseline; the outcomes of this analysis were not reported in the table.

^b Based on binary logistic regression, Cox & Snell R ² = 0.026, Nagelkerke R ² = 0.036. The reported outcomes of this analysis reflect differences between dropouts and completers.

*p < .05. **p < .01. ***p < .001.

V = video; T = text; C = combination.

^a Rated on a 5-point scale ranging from 1 (low acceptability) to 5 (high acceptability).

*p < .05. **p < .01. ***p < .001.

‡Small effect. †Moderate effect. † † Large effect.

*p < .05. **p < .01. ***p < .001.

‡Small effect. †Moderate effect. † † Large effect.

*p < .05. **p < .01. ***p < .001.