Online hands-on workshops followed by focus groups were conducted. The workshops were conducted between November 2021 and March 2022. Focus groups were designed to be run with 4 to 6 participants, with a total of 15 to 20 participants, based upon previous usability studies using apps and web-based delivery [46-49]. The workshops and focus groups were led by WS and supported by HH (who also took field notes). WS has a PhD in health psychology, was a postdoctoral fellow at the time of the study, and identifies as a man. HH has a master’s degree in rehabilitation science, was a PhD student at the time of the study, and identifies as a woman. WS had prior experience in conducting research focus groups and interviews and HH did not. WS and HH both have prior experience analyzing qualitative data. WS and HH introduced themselves to the participants as research trainees with a focus on PA behavior change.

Eligible participants were adults aged 19 to 64 years who owned a smartphone with either an iOS or Android operating system and were fluent in English. Participants were recruited via online advertisements on the research laboratory’s public X account (X Corp; @bmedlab); public Instagram account (Meta Platforms, Inc; @uvicbmed); public community Facebook (Meta Platforms, Inc) groups local to Victoria, British Columbia; and through snowball sampling. All communication between researchers and participants took place over email. Some participants were known to the researchers because they worked in the same university; however, none of the participants had provided feedback on the app before the study.

Interested participants contacted the researchers via email and were sent an email with the letter of information and consent form. Upon providing consent, participants were asked to provide availability for a 60-minute workshop and focus group. Participants were informed that they would be given a pseudonym during the workshops and focus groups to protect their identity. Participants were later contacted to reconfirm their availability once 4 to 5 participants reported having a mutually available time and were sent a Zoom videoconferencing link upon confirmation (Zoom Video Communications). In addition, participants were asked to download the Pathverse app platform [50] (Pathverse, Inc), which gave them access to the M-PAC app on their smartphone, to be able to actively participate in the workshop portion of the session.

One day before the date of the workshop or focus group, participant emails were used to create profiles on the admin portal of Pathverse, so that participants would have access to the M-PAC app. On the arranged time of the workshop or focus group, participants joined into a Zoom call with 2 researchers. All participants were given pseudonyms and were unable to turn on video. The lead researcher shared their smartphone screen and led the workshop and focus group discussion, while the other researcher managed the video recording software and any questions within the Zoom chat. Upon entering the Zoom call, the lead researcher asked participants to open the Pathverse app and then instructed participants on how to create a profile on Pathverse using their email. Once all the participants had access to the app, the lead researcher began a walkthrough of the app.

Participants were first shown how to access the lessons, along with a brief overview of how to navigate through lesson cards, the different types of cards, and how to mark a lesson as complete, using “Lesson 0: Tutorial” and “Lesson 1: What is Physical Activity” as examples. Afterward, participants were directed to the tracking features on the app. Participants were shown how to sync their step count to the app as well as how to track and log PA, weight, and blood pressure. Participants were then shown the goal-setting feature of the app. The lead researcher created an example goal, demonstrating to participants how to name the goal, set priority and reminders on the goal, and how to mark goals as completed. Finally, participants were shown how to navigate to the home page of the app and to find the in-app navigation tutorial if they got stuck. Participants were then given 5 to 10 minutes to explore the app and its features independently. Overall, the workshop portion of the session took approximately 25 to 30 minutes.

Following the workshop, participants were informed that they would be starting the focus group portion of the session. The lead researcher read the introductory script and then prompted participants to ask any questions before beginning with the list of semistructured questions (Multimedia Appendix 1). The focus group portion of the session took approximately 25 to 30 minutes. The workshop and focus group guide was pilot-tested among the research team. Only the participants and researchers were present during the workshops and focus groups. There were no repeat workshops and focus groups.

The M-PAC app is an evidence-based mobile app designed to support users in changing their PA behavior, developed by researchers specializing in PA behavior. The app is built upon the Pathverse app platform [50]. Pathverse is an mHealth development platform designed for health and wellness professionals and researchers. This platform consists of an admin web-portal interface and a participant Pathverse app (available in both IOS and Android stores). The M-PAC app, specifically, includes educational content in the form of lessons; self-monitoring features, such as step tracking and PA logging; and a goal-setting feature, whereby users can set goals for themselves as well as review completed goals (Figure 1).

Lessons within the app are broken down into individual “cards,” which users navigate by swiping left or right, or via the arrow buttons at the bottom of the screen. Cards present either (1) a brief block of text, (2) a brief block of text with an image placed below or above, (3) a brief block of text with a link to an external web page, or (4) a challenge question. The challenge question card presents the user with a short question and 4 possible answers. Selecting any answer presents the user with a new slide indicating whether their answer was correct or incorrect, as well as a brief explanation for the correct answer.

Lessons within the app are scaffolded according to the different constructs within the M-PAC framework. Lessons 1 to 4 are designed to address reflective processes of the M-PAC framework: instrumental attitudes (“What are the health benefits of PA?”), perceived opportunity (“Do I have the resources to do PA?”), perceived capability (“Am I confident I can perform this PA?”), and affective judgments (“Do I like PA?”). Similarly, lessons 5 to 10 are aimed at addressing the regulatory (ie, skills important in facilitating continued PA behavior, such as goal setting and planning) and reflexive processes (eg, habit and exercise identity) of the M-PAC. In addition, lessons contain challenge cards that assess users’ understanding of the content within their current lesson. Refer to Multimedia Appendix 2 [51] for a full list of app lessons, mechanisms of action, BCTs, and behavior change ontologies applied [52,53]. Lessons must be completed in order (ie, lesson 0, lesson 1, lesson 2, etc) to unlock subsequent lessons, but completed lessons can be revisited once unlocked. Given the usability focus of this study, only lessons 1 to 3 (including a lesson 0: tutorial) were presented to participants.

The primary outcome of this phase was qualitative feedback on the usability of the M-PAC app drawn from the focus groups. Specifically, questions within the focus group were drawn from Nielsen's 10 usability heuristics for user interface design [54]. These heuristics were adapted for this work to guide the structured focus group questions, which addressed the following: layout and information, appropriateness of language, backing out or undoing an action, consistency in language and features, errors and mistakes, forgetting acronyms, efficiency of interface and navigation, visual design, ease in diagnosing and resolving errors, and help and documentation. Probing questions for each main question were also planned if group responses were vague or open ended. For example, the first main question was, “How did you find the layout and information of the app?” with probing questions of “If intuitive, what made the layout or information intuitive?” and “If unintuitive, what made the layout or information unintuitive? What would be more helpful?” The specific questions related to these heuristics can be found in Multimedia Appendix 3. Participant responses were recorded using the video recording software Open Broadcast Software (OBS) Studio [55] (OBS Studio Contributors).

Participant demographics of age, gender, and highest attained education were assessed using single-item questions. Furthermore, the brand of smartphone owned, ability to use the features of a smartphone, ability to engage with other technologies, and whether a PA app has been downloaded or used in the past were also assessed using single-item measures.

Focus groups were recorded using OBS Studio and transcribed verbatim [55]. Transcripts were coded and analyzed thematically, according to the 10 usability heuristics [54] adapted for this study, using NVivo (QSR International) [56]. Participants were not asked to provide feedback on the transcripts or findings. Two coders independently coded the transcripts, with conflicts being resolved by a third coder.

A 2-week, single-group pilot study was conducted virtually between August 2022 and February 2023. The researchers were located in Victoria, British Columbia, Canada, and the participants were located anywhere in Canada.

Eligible participants were Canadian adults aged 19 to 64 years who owned a smartphone device with either an iOS or Android operating system, were fluent in English, and reported engaging in <150 minutes of moderate-to-vigorous PA (MVPA) per week. Individuals were excluded if they participated in the phase 1 workshops or focus groups or were involved in the design and development of the M-PAC app. Participants were recruited via online advertisements on the research laboratory’s public X account (@bmedlab), public Instagram account (@uvicbmed), and public community Facebook groups local to Victoria, British Columbia; with paid Facebook advertisements; and snowball sampling. All communication between researchers and participants took place over email. A total of 10 to 15 participants was deemed appropriate based on previous usability testing studies examining apps [57,58].

Interested individuals contacted the research team via email and were sent an email with the letter of information and a consent form. Upon agreeing to participate in the study, participants were sent instructions on how to download the M-PAC app, as well as a demographics questionnaire. Participants also booked a 30-minute baseline meeting with a researcher to walkthrough the M-PAC app over Zoom. During the baseline meeting, participants were given a walkthrough of the M-PAC app, identical to the workshop from phase 1. Upon completing the walkthrough, participants were asked to schedule a 30-minute exit interview in 2 weeks and finish lessons 1 to 3 within those 2 weeks.

During the exit interview, the researcher followed the semistructured interview guide to assess participants’ experiences with the M-PAC app (Multimedia Appendix 3). The interview guide was pilot-tested among the research team. Interviews were conducted by WS over Zoom, audio recorded using OBS Studio, and lasted 20 to 30 minutes. Only the participants and researchers were present, and there were no repeat interviews. After completing the exit interview, participants were told that they would be sent the Mobile App Usability Questionnaire (MAUQ) to their email in 24 hours, and to complete the MAUQ within 3 days of receiving the questionnaire [59]. Upon completion of the MAUQ, participants were given a CAD $15 (US $11.88) honorarium. There were no changes made to the methods upon pilot study commencement.

Following the completion of phase 1, several modifications were made to the M-PAC app. Changes to content included the following: (1) modifying the introductory card for the self-efficacy lesson from “Introduction to self-efficacy...” to “Ways to improve confidence (also known as self-efficacy)...”, (2) ensuring all acronyms (eg, MVPA; Canadian Society for Exercise Physiology; frequency, intensity, time, and type [FITT]; and Get Active Questionnaire) were clearly defined, (3) bolding some terms to highlight them (eg, FITT), (4) reducing the amount of text on cards and increasing the number of cards per lesson, and (5) adding a references module and providing links to references from each lesson. Changes to the Pathverse platform included the following: (1) optimizing image loading times, (2) creating an external link placeholder image that reads “Link,” and (3) adding a “Forgot your password?” feature. The updated version of the M-PAC app with the aforementioned modifications was used for the usability pilot study.

The primary outcome of this phase was qualitative feedback on the usability and satisfaction of the M-PAC app, gathered through the exit interview. Probing questions for each main question were also planned if participants’ responses were vague or open ended. For example, for the question, “Can you tell me what you liked best about the M-PAC app?”, probing questions of “Were there any features in particular?” and “What about the content and/or layout?” were planned. The specific questions within this interview guide can be found in Multimedia Appendix 3. Participant responses were recorded using the video recording software OBS Studio [55]. The a priori criteria for moving on to a randomized pilot trial was an overall consensus that the intervention was usable (MAUQ score≥5), and at least 60% (8/13) of participants were satisfied with the intervention, with minor to moderate suggestions for improvement.

Quantitative feedback on the usability of the M-PAC app was collected via the MAUQ [59], specifically the version for stand-alone mHealth apps. This version of the MAUQ is an 18-item self-report questionnaire that assesses an mHealth app’s ease of use, interface and satisfaction, and usefulness. The questionnaire demonstrated construct and criterion validity and reliability, with Cronbach α=0.91 for the total MAUQ, 0.71 for ease of use, 0.93 for interface and satisfaction, and 0.65 for usefulness internal consistency.

Participant demographics of age, gender, and highest attained education were assessed using single-item questions. Furthermore, the brand of smartphone owned, ability to use the features of a smartphone, ability to engage with other technologies, and whether a PA app has been downloaded or used in the past were also assessed using single-item measures. Participants’ average weekly PA was assessed using the modified Godin Leisure-Time PA questionnaire [60]. The estimated weekly MVPA was calculated by multiplying the frequency and duration of moderate- and vigorous-intensity PA minutes and adding them together. The Godin-Leisure-Time PA questionnaire has demonstrated acceptable validity and reliability in a variety of settings, populations, and countries [60,61]. There were no changes made to the assessments or measurements after the trial commenced.

Exit interviews were recorded using OBS Studio and transcribed verbatim. Transcripts were coded and analyzed thematically, according to the 10 usability heuristics [54] adapted for this study, using NVivo [56]. Participants were not asked to provide feedback on the transcripts or findings. Two coders independently coded the transcripts, with conflicts being resolved by a third coder.

Descriptives (ie, mean and SD) for quantitative usability feedback from the MAUQ were calculated for total MAUQ scores, as well as for the ease of use, interface and satisfaction, and usefulness subscales [59].

We conducted 3 focus groups of 5 participants each. Participants were primarily women (13/15, 87%) with a mean age of 30.4 (SD 9.49) years. Most had at least a college or university degree (14/15, 93%) and self-reported having above average abilities to use the functions of their smartphone (11/15, 73%), and just under half had previously downloaded a mobile app to help increase their PA (7/15, 47%; eg, Down Dog, Fitbit, Health, Nike Training Club app, Endomondo, and Sweat app). The most common brand of smartphone was Apple iPhone (9/15, 60%), followed by Samsung (4/15, 27%) and Google Pixel (2/15, 13%; Table 1). None of the participants refused to participate or dropped out of the study.

Fourteen participants took part in phase 2 (usability pilot). One participant did not complete the baseline questionnaire due to unknown reason. Participants were mainly women (11/13, 85%) with a mean age of 41.38 (SD 12.92) years. Most (9/13, 69%) had at least a college or university degree, all (13/13, 100%) self-reported having above average abilities to use the functions of their smartphone, and most (8/13, 62%) had previously downloaded a mobile app to help increase their PA(eg, Fitbit, Optimity, Carrot, Habitshare, MyFitnessPal, C25K, ParticipACTION app, and 7M Workout). The most common brand of smartphone was Apple iPhone (5/13, 38%), followed by Samsung (4/13, 31%), Google Pixel (1/13, 8%), Hauwei (1/13, 8%), LG (1/13, 8%), and Motorola (1/13, 8%; Table 1). The participants reported on average 66.07 (SD 57.92) minutes of MVPA per week. None of the participants refused to participate or dropped out of the study.

A total of 13 participants filled out the MAUQ (Table 3). However, 1 participant did not fill out the MAUQ for unknown reasons. The mean total MAUQ score was 5.06 (SD 1.20), and the mean subscale scores were as follows: “ease of use” 5.36 (SD 1.27), “interface and satisfaction” 5.52 (SD 1.42), and “usefulness” 4.17 (SD 1.31).