@Article{info:doi/10.2196/jmir.1998,
author="Vandelanotte, Corneel
and Duncan, Mitch J
and Plotnikoff, Ronald C
and Mummery, W Kerry",
title="Do Participants' Preferences for Mode of Delivery (Text, Video, or Both) Influence the Effectiveness of a Web-Based Physical Activity Intervention?",
journal="J Med Internet Res",
year="2012",
month="Feb",
day="29",
volume="14",
number="1",
pages="e37",
keywords="Physical activity: computer tailoring; mismatch; preferences; delivery method; website-delivered intervention; behavior-change intervention",
abstract="Background: In randomized controlled trials, participants cannot choose their preferred intervention delivery mode and thus might refuse to participate or not engage fully if assigned to a nonpreferred group. This might underestimate the true effectiveness of behavior-change interventions. Objective: To examine whether receiving interventions either matched or mismatched with participants' preferred delivery mode would influence effectiveness of a Web-based physical activity intervention. Methods: Adults (n = 863), recruited via email, were randomly assigned to one of three intervention delivery modes (text based, video based, or combined) and received fully automated, Internet-delivered personal advice about physical activity. Personalized intervention content, based on the theory of planned behavior and stages of change concept, was identical across groups. Online, self-assessed questionnaires measuring physical activity were completed at baseline, 1 week, and 1 month. Physical activity advice acceptability and website usability were assessed at 1 week. Before randomization, participants were asked which delivery mode they preferred, to categorize them as matched or mismatched. Time spent on the website was measured throughout the intervention. We applied intention-to-treat, repeated-measures analyses of covariance to assess group differences. Results: Attrition was high (575/863, 66.6{\%}), though equal between groups (t863 =1.31, P =.19). At 1-month follow-up, 93 participants were categorized as matched and 195 as mismatched. They preferred text mode (493/803, 61.4{\%}) over combined (216/803, 26.9{\%}) and video modes (94/803, 11.7{\%}). After the intervention, 20{\%} (26/132) of matched-group participants and 34{\%} (96/282) in the mismatched group changed their delivery mode preference. Time effects were significant for all physical activity outcomes (total physical activity: F2,801 = 5.07, P = .009; number of activity sessions: F2,801 = 7.52, P < .001; walking: F2,801 = 8.32, P < .001; moderate physical activity: F2,801 = 9.53, P < .001; and vigorous physical activity: F2,801 = 6.04, P = .002), indicating that physical activity increased over time for both matched and mismatched groups. Matched-group participants improved physical activity outcomes slightly more than those in the mismatched group, but interaction effects were not significant. Physical activity advice acceptability (content scale: t368 = .10, P = .92; layout scale: t368 = 1.53, P = .12) and website usability (layout scale: t426 = .05, P = .96; ease of use scale: t426 = .21, P = .83) were generally high and did not differ between the matched and mismatched groups. The only significant difference (t621 = 2.16, P = .03) was in relation to total time spent on the website: the mismatched group spent significantly more time on the website (14.4 minutes) than the matched group (12.1 minutes). Conclusion: Participants' preference regarding delivery mode may not significantly influence intervention outcomes. Consequently, allowing participants to choose their preferred delivery mode may not increase effectiveness of Web-based interventions. ",
issn="1438-8871",
doi="10.2196/jmir.1998",
url="http://www.jmir.org/2012/1/e37/",
url="https://doi.org/10.2196/jmir.1998",
url="http://www.ncbi.nlm.nih.gov/pubmed/22377834"
}