Abstract
Approximately one-third of adults in the United States are physically inactive. This is a significant public health concern as physical activity (PA) can influence the risk of cardiovascular disease, diabetes, and certain forms of cancer. To minimize these health risks, effective PA interventions must be developed and disseminated to the vast number of individuals who remain sedentary. Smartphone technology presents an exciting opportunity for delivering PA interventions remotely. Although a number of PA applications are currently available for smartphones, these “apps” are not based on established theories of health behavior change and most do not include evidence-based features (e.g., reinforcement and goal setting). Our aim was to collect formative data to develop a smartphone PA app that is empirically and theoretically-based and incorporates user preferences. We recruited 15 sedentary adults to test three currently available PA smartphone apps and provide qualitative and quantitative feedback. Findings indicate that users have a number of specific preferences with regard to PA app features, including that apps provide automatic tracking of PA (e.g., steps taken and calories burned), track progress toward PA goals, and integrate a music feature. Participants also preferred that PA apps be flexible enough to be used with several types of PA, and have well-documented features and user-friendly interfaces (e.g., a one-click main page). When queried by the researcher, most participants endorsed including goal-setting and problem-solving features. These findings provide a blue print for developing a smartphone PA app that incorporates evidence-based components and user preferences.
Introduction
Despite the benefits of physical activity (PA), 1 –9 approximately one-third of adults in the United States are inactive. 10 Interventions are needed to reach these sedentary individuals. Smartphone technology provides an opportunity for delivering PA interventions remotely. A smartphone PA application (SPAA) could track PA in real time using the built-in accelerometer and provide instant feedback on steps taken or calories burned—an immediate reinforcement of PA. A SPAA could also track progress toward PA goals over time and include other evidence-based features enabling goal setting and problem solving. Current SPAAs are not evidence based, however, and most do not target these key components of successful PA interventions. 11 We conducted formative research aimed at developing theoretically and empirically based SPAAs that incorporate user preferences.
Methods
Participants
Participants were recruited to evaluate three current SPAAs. The study was advertised online. Twenty-six individuals responded and were screened for eligibility. Eligibility criteria included the following: age 18 to 65, smartphone (iPhone) owner, perform <90 min of moderate or 60 min vigorous PA weekly, no medical contraindications to PA, body mass index (BMI) under 35, and not pregnant. Six individuals were ineligible (due to BMI or activity level), 5 were eligible but unable to attend the onsite assessment, and 15 enrolled and completed assessments between November 2010 and February 2011.
Procedure
Participants attended one onsite session during which they provided informed consent and completed a demographic survey. A researcher then assisted participants in downloading three SPAAs: iTreadmill 3.1.0, iFitness Hero 1.0, and Exercise Tracker 1.12. Participants tested each consecutively for 1 week in a prescribed order counterbalanced across participants to control for order effects.
Participants were called at the end of each week to provide quantitative and qualitative feedback. They used Likert scales to rate the ease of use (1=very difficult to 5=very easy), helpfulness (1=not at all helpful to 5=very helpful), and overall satisfaction (1=very unsatisfied to 5=very satisfied) with the SPAA. Qualitative feedback was collected during semi-structured interviews in which participants identified positive and negative aspects of the SPAA, features to change/add, and whether they would use the SPAA in the future. At the final assessment, participants were also asked whether goal-setting or problem-solving features would be useful. Participants received a $10 gift card from iTunes after each assessment.
Descriptive analyses were performed on SPAA ratings. Qualitative data were analyzed to identify preferred SPAA characteristics: executive summaries of each interview were created, integrated into a master document, and coded based on the a priori research question, “What features are considered desirable in a SPAA?”
Results
Fourteen participants completed all assessments, and one completed two of three. Participants ranged in age from 23 to 60 (mean=38.9 years, SD=13.3). Most were married (53.3%) or single (40.0%). Participants identified as white, non-Hispanic (80.0%), Asian (6.7%), black/African-American (6.7%), and Hispanic (6.7%). Nearly half (46.7%) had an annual household income of at least $50,000 and 80% were employed fulltime.
Quantitative ratings are detailed in Table 1. Exercise Tracker appears to have received lower ratings than the other SPAAs—possibly because it does not automatically track PA—though the sample is too small to test for significant differences.
Participant Ratings of Three Physical Activity Applications
PA, physical activity.
Table 2 provides qualitative data on SPAA features considered desirable. Participants also expressed a preference for certain SPAA characteristics. They stated that an optimal SPAA could be used for a range of activities (12/15), including indoor and outdoor activities (4/15). Participants emphasized the importance of ease of use (12/15), with some suggesting a comprehensive or one-click main page from which users could navigate to SPAA features. Participants also said SPAAs should have good documentation of features (6/15) and aesthetically appealing interfaces (4/15). A few (3/15) noted that the smartphone auto-exit feature should not exit the SPAA during exercise sessions. When queried by the researcher at their third assessment, most stated that it would be useful to include features that help users problem-solve common barriers to PA (e.g., bad weather; 11/14) and set weekly goals (13/14). Two participants added that SPAAs should provide guidance on goal setting, such as how many minutes to exercise.
Features Considered Desirable in a Physical Activity Application
BMI, body mass index.
Discussion
Participants identified several characteristics that would enhance SPAAs. They placed a premium on automatic tracking of PA, and emphasized the importance of receiving feedback on PA accomplishments. This is consistent with prior research indicating that proximal reinforcers are key components of successful behavior change interventions. 12 Participants emphasized the need for SPAAs to be flexible and accommodate different forms of PA. Some spontaneously suggested that a goal-setting feature was desirable and most endorsed this when asked. Finally, there was widespread support for including features that help users problem-solve common barriers to PA. Such a feature could ask users to select the barrier(s) interfering with their PA (e.g., rain and fatigue) and list suggestions for overcoming each barrier (e.g., walk in a mall during inclement weather). Features that target goal setting, problem solving, and behavioral reinforcement are consistent with principles of social cognitive theory (SCT) 13 ; as SCT has been used successfully in prior PA interventions, grounding an SPAA in SCT may increase its appeal and efficacy. 14,15 Thus, these findings provide a jumping off point for developing a SPAA that incorporates user preferences and is based in established theories of behavior change. Given that one-third of American adults are physically inactive, 10 and it is estimated that half will own smartphones before 2012, 16 an efficacious SPAA has the potential to exert a significant public health impact by reducing the incidence of diseases associated with inactivity (e.g., cardiovascular disease).
This study is the first to examine user preferences regarding SPAAs. The intent was to collect formative data and qualitative data saturation was easily achieved with 15 participants; thus, the sample size appears appropriate. Nonetheless, given the small sample and the relatively high socioeconomic status of participants, findings may not generalize to the sedentary population. Likewise, all participants already owned smartphones, and may have been more comfortable with smartphone applications than the sedentary population. Additional research is needed to determine whether findings extend to users who have no prior experience with smartphones applications.
Footnotes
Acknowledgments
We thank Kathy Palmer for her assistance with participant recruitment for this project.
Disclosure Statement
No competing financial interests exist.