This is an open-access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Research Protocols, is properly cited. The complete bibliographic information, a link to the original publication on http://www.researchprotocols.org, as well as this copyright and license information must be included.
Most older Americans do not exercise regularly and many have chronic conditions. Among an increasing number of fitness mobile and Web apps, few are designed for older adults with chronic conditions despite high ownership rates of mobile tools and Internet access in this population. We designed a mobile-enabled Web app, iCanFit, to promote physical activity in this population.
This study aimed to test the usability and acceptability of iCanFit among older adults in a community setting.
A total of 33 older adults (aged 60 to 82 years) were recruited from communities to test iCanFit. Of these 33, 10 participants completed the usability testing in a computer room of a senior community center. A research assistant timed each Web application task and observed user navigation behavior using usability metrics. The other 23 participants used the website on their own devices at home and provided feedback after 2-3 weeks by completing a user-experience survey assessing ease of use, helpfulness, and satisfaction with iCanFit.
Participants completed all 15 tasks on the iCanFit site in an average of 31 (SD 6.9) minutes; some tasks required more time or needed assistance. Participants’ comments were addressed to improve the site’s senior friendliness and ease of use. In the user-experience survey, participants reported high levels of usefulness and satisfaction. More than 56% (13/23) of participants indicated they would continue using the program and recommend it to their families or friends.
Testing usability and acceptability is a very important step in developing age-appropriate and user-friendly Web apps, especially for older adults. Testing usability and acceptability in a community setting can help reveal users’ experiences and feedback in a real-life setting. Our study suggested that older adults had a high degree of acceptance of iCanFit and could use it easily. The efficacy trial of iCanFit is currently underway.
As of January 2014, 87% of American adults have used the Internet. Even among older adults, 88% of those aged 50 to 64 years are online and more than 57% of those older than 64 years are online [
Approximately 13% of the American population is older than 65 years, and by 2030 older adults will account for 20% of the US population [
National surveys suggest that older adults who use the Internet are more likely to seek health information online [
In-line with efforts to promote physical activity among older adults with chronic conditions, especially cancer, we developed a mobile-enabled Web app called iCanFit [
During the iCanFit protocol development, we conducted an iterative heuristic evaluation with experts from behavioral sciences, computer science, human factors and ergonomics, exercise sciences, public health, and gerontology. The goal of the current study was to test usability and acceptability of iCanFit among older adults.
Screenshots of iCanFit.
Goals function in iCanFit.
Entering, tracking, and receiving feedback of physical activity on iCanFit.
Different modes of View Progress on iCanFit.
The study took place in a college town of central Texas from April to June 2014. Following the usability testing methods proposed by Kushniruk et al [
Considering significant variance in computer skills of older adults, testing in a laboratory may not reflect users’ experience in a real-life setting [
Based on these theories and rationale, our testing of iCanFit was designed to include 2 phases. The first phase was usability testing in a computer room at a senior community center. The participants were given the name of the website and asked to explore the site on their own. During this phase, a research assistant (RA) measured and observed the user’s navigation using usability testing metrics (see
We developed a protocol of usability and acceptability testing that guided every step of the process. The protocol included recruitment scripts, a recruitment flyer, usability testing metrics, and a user-experience survey. We conducted intensive training with RAs to ensure the testing protocol was followed with fidelity.
Usability testing metrics (see
The user-experience survey (see
We recruited our participants through active community outreach. Flyers were posted at senior community centers and public libraries and announcements were made during breaks of classes or programs for seniors. Participants who were aged 60 years or older and had used the Internet were invited to participate in our study. A total of 33 participants were recruited for the study; 10 conducted usability testing in a computer room in a senior center and the remaining 23 participants performed acceptability testing by using iCanFit for 2-3 weeks at home followed by a user-experience survey. The study protocol was approved by the Institutional Review Board at the Texas A&M University.
In the first phase of usability testing, the participants were invited to a computer room in a senior center. The RAs first introduced themselves and explained the purpose and procedure of the testing. Participants were assured of their privacy and verbal consent was obtained. An RA was paired with a participant. There was enough space between desktop computers to ensure proper testing and observation. A brief survey on participant demographics was administered before the testing. The participant was then given the website name and asked to explore the site on their own. The RA sat behind the participant and gave no instruction to the participant unless the participant could not proceed after repeated efforts. The participant was also encouraged to make comments during the navigation. The RA recorded the participant behavior and comments using observation metrics and took detailed notes. The metrics included 15 tasks to complete on the website, time needed on each task, and if a task was performed without error, with error, or needed assistance. Each participant received a US $20 gift card as compensation for their participation.
In the second phase of acceptability testing, another sample of 23 participants was recruited to test iCanFit on their own devices independently. When participants responded to our flyer by calling or speaking to our RAs at community outreach, they were instructed to visit the iCanFit site and use it for 2 to 3 weeks. An interview was then scheduled at the participant’s convenience to solicit their feedback on the website, including how difficult was it to navigate the site, the usefulness of each function and the site in general, and how satisfied they were with the website. Some interviews were conducted in-person at a community center and some were conducted over the phone. Verbal consent was obtained before the interview and each participant received a US $20 gift card as compensation for their participation.
All data were saved and analyzed in SPSS 16.0 (SPSS Inc, Chicago, IL, USA) and descriptive statistics were used to explore the mean usability and acceptability scores. All text data were extracted from SPSS and entered into ATLAS.ti (Berlin, Germany) for further analysis. We identified the most frequently used phrases or keywords, and delineated a range of responses for each task on iCanFit and their overall experience.
As shown in
Characteristics of participants in usability testing and user-experience survey of iCanFit.
Characteristics | Utility testing sample |
User-experience survey sample |
|
|
|
|
|
|
Mean (SD) | 67.60 (6.3) | 67.6 (6.5) |
|
Range | 60-78 | 60-82 |
|
|
|
|
|
Male | 7 (70) | 4 (17) |
|
Female | 3 (30) | 19 (83 |
|
|
|
|
|
≤High school | 3 (30) | 6 (26) |
|
>High school | 7 (70) | 17 (74) |
|
|
|
|
|
Mean (SD) | 20.30 (10.5) | 20 (8.0) |
|
Range | 3-38 | 6-30 |
|
|
|
|
|
Desk top | 6 (60) | 11 (52) |
|
Laptop | 3 (30) | 8 (35) |
|
Tablet | 1 (10) | 2 (8) |
|
Smartphone | 0 | 1 (4) |
A total of 23 participants completed the user-experience survey; their ages were between 60 to 82 years (mean 68, SD 6.5). Of these 23, 19 (80%) were female and 17 (74%) had more than high school education. Approximately 52% (11/23) participants used a desktop as their primary mode of Internet access, followed by laptops (35%, 8/23), and tablets (8%, 8/23).
The most difficult task appeared to be creating an account. It took participants an average of 9.2 minutes to create a user account. In all, 40% (4/10) could perform this task without error, 40% (9/23) completed the task with error, and 20% (2/10) needed assistance. Only 30% (3/10) watched the instructional video under the Help tab or logged out of the site after they completed all other tasks.
Usability testing results of iCanFit (N=10).
Task | Time to complete (min) | Perform without error, n (%) | Perform with error, n (%) | Need assistance, n (%) | |
|
Mean (SD) | Range |
|
|
|
Find the website | 2.4 (2.3) | 1-3 | 6 (60) | 2 (20) | 2 (20) |
Create an account | 9.2 (5.2) | 1-17 | 4 (40) | 4 (40) | 2 (20) |
Log into the account | 2.7 (4.1) | 0.5-11 | 8 (80) | 1 (10) | 1 (10) |
Find Healthy Tips and read it | 1.1 (1.6) | 0-5 | 8 (80) | 1 (10) | 1 (10) |
Find Resources and read it | 3.4 (6.3) | 0-20 | 9 (90) | 0 | 1 (10) |
Find the Facebook account through the link on the site | 1.1 (1.3) | 0-3 | 7 (70) | 1 (10) | 2 (20) |
Find the Goal Home | 0.7 (0.6) | 0-2 | 7 (70) | 1 (10) | 2 (20) |
Set a long-term goal | 2.1 (2.3) | 0-7 | 6 (60) | 1 (10) | 3 (30) |
Set a short-term goal | 1.4 (1.3) | 0-4 | 6 (60) | 1 (10) | 3 (30) |
Track the short-term goal by entering physical activity | 3.4 (3.8) | 0-8 | 5 (50) | 2 (20) | 3 (30) |
Enter physical activity without tracking the short-term goal | 2.1 (1.7) | 0-5 | 9 (90) | 1 (10) | 0 |
View physical activity progress through View Progress | 0.9 (1.1) | 0-3 | 9 (90) | 1 (10) | 0 |
Switch view modes in View Progress | 0 |
|
10 (100) | 0 | 0 |
Find help | 1 (1) | 0-2 | 3 (30) | 0 | 0 |
Log out |
|
|
3 (30) | 1 (10) | 6 (60) |
Time to complete the entire site | 31.4 (6.9) | 22-40 |
|
|
|
Through the observations and participants’ comments we obtained from the usability testing, we were able to identify and make the necessary changes to the iCanFit application to improve usability and senior friendliness. For example, comments from some of the participants indicated that some fonts and icons needed to be changed to be more visible for older adults. To address the problems participants had with goal setting, some words were changed to avoid confusion. For instance, after setting a short-term goal, the “Add Activity” button was changed to “Save.”
We also added some hot buttons for frequently used functions. For example, the “Exercised Today?” button was created allowing participants to enter activity before going to Goals and the “Log out” button was placed in a more visible location. During usability testing, we also learned that some pages of iCanFit did not display well on Internet Explorer 7.0 or lower, so we modified our site to make it compatible with more browsers.
Two major complaints from the participants were that they did not know what to do after they went onto the website and that they had trouble creating an account. Therefore, we changed our instructional video and made 3 separate videos, ranging from 30 seconds to 2 minutes in length, and explained (1) what iCanFit is and how to use it, (2) how to create an account, and (3) how to use Goals so that users can easily find the help they need.
Results of user-experience survey (N=23).
Variable | n (%) | Mean (SD) | |
|
|
|
|
|
Flyer at community center | 15 (65) |
|
|
Email listserve | 2 (9) |
|
|
Friend/relative | 5 (22) |
|
|
Other | 1 (4) |
|
|
|
|
|
|
Desktop | 10 (44) |
|
|
Laptop | 7 (30) |
|
|
Tablets | 4 (17) |
|
|
Smartphone | 2 (9) |
|
|
|
|
|
|
<Once/week | 9 (39) |
|
|
Approximately once/week | 9 (39) |
|
|
2-3 times/week | 3 (13) |
|
|
4-5 times/week | 1 (4) |
|
|
Every day | 1 (4) |
|
Total time on iCanFit in past week (minutes; range 0-60), mean (SD) |
|
21.6 (4.0) | |
|
|
|
|
|
Creating account |
|
3.2 (0.4) |
|
Long-term goal setting |
|
3.7 (0.2) |
|
Short-term goal setting |
|
3.8 (0.3) |
|
Short-term goal tracking |
|
3.3 (0.3) |
|
View progress |
|
3.7 (0.2) |
|
Overall difficulty |
|
3.6 (0.3) |
|
|
|
|
|
Instructional video |
|
3.7 (0.3) |
|
Healthy tips |
|
3.5 (0.2) |
|
Resources |
|
3.2 (0.2) |
|
Facebook page |
|
3.1 (0.3) |
|
Overall usefulness |
|
3.4 (0.3) |
|
|
|
|
|
Ever talked to family/friends about iCanFit | 13 (56.5) |
|
|
Would recommend iCanFit to family/friends | 13 (56.5) |
|
a Difficulty score: 1=very difficult, 2=somehow difficult, 3=somehow easy, 4=very easy.
b Usefulness score: 1=useless, 2=a little useless, 3=somehow useful, 4=very useful.
When asked to rate how difficult it was to use each function, most participants reported no difficulty or little difficulty in completing the major functions of the site with a score range of 3.2 to 3.8 (1=very difficult, 4=very easy). The overall difficulty for the major function of Goals was rated 3.6 (SD 0.3). When asked to rate usefulness of each function, most participants gave a rating of 3.1 to 3.7 (1=useless, 4=very useful). Participants gave the iCanFit website an average overall usefulness rating of 3.4 (SD 0.3). More than half (57%, 13/23) of the participants had talked to their family or friends about the iCanFit program and the same number of participants would recommend iCanFit to their family or friends.
Approximately 30% (7/23) of participants reported preferring to track physical activity in Goals through the “Exercised Today?” hot button and only 9% (2/23) liked to track activity through Track Goals; 22% (5/23) indicated having no preference because both were easy to use (data not shown).
Our qualitative data from the user-experience survey revealed that with different prior online experiences and varying statuses of current physical activity, participants had different experiences with iCanFit. For some participants who exercised regularly, they felt that the program added little to their current life: “I am doing exercise regularly, and I have a pedometer.” A couple of participants who did not use the computer often made remarks such as: “I prefer hardcopies of goals that can be stuck to the refrigerator so I can check it easily.” Most participants, however, were very positive about their experience with iCanFit, making comments such as: “Super great program, keep it up,” “It increased my activity because I was trying to get 100% of my goals,” and “It’s great to see how many times I have exercised; it gave me a kick to get up and accomplish something every day.”
Participants also offered suggestions on how to use mobile tools for older adults. For example, a 71-year-old male user commented, “It has to be something automatic or very easy to use. I like the dropdown menu when entering activities so I don’t need to type.” A 65-year-old female user added, “It would be nice if it has function to remind me to exercise, since we don’t remember things well at this age. And I like the graphs to see my progress.” They also shared thoughts on how to motivate older adults to exercise regularly. For instance (shared by a 68-year old female user), “If we can get people to start a program that combines a fitness class and how to use this site (iCanFit), you can motivate many sedentary people.” A 62-year-old male user suggested, “For those living alone, it is more about helping them find friends and getting them involved.”
Older adults from the community were recruited to test the usability and acceptability of a Web application designed to promote physical activities for older cancer survivors. Usability and acceptability was tested in settings familiar to the participants and on devices they often used to maximize their real-life experience. The relative user efficiency data, such as time to complete the tasks and errors made in first-time use, were within an acceptable range and reflected the anticipated usability gap between expert and novice users [
As the use of mobile tools continues to increase, especially among older adults, mobile technology is being used increasingly as an efficient tool for health promotion [
In addition, when developing mobile programs for older adults, it is important to involve end users from early stage of design and conduct on-going usability testing [
Several limitations of the study should be noted. First, iCanFit was originally designed to promote physical activity among older cancer survivors [
Despite these limitations, to the best of our knowledge, our study was one of the first to report testing of usability and acceptability of a Web app to promote physical activity among older adults. Our findings underscore the importance of using validated metrics and mixed methods to test multidimensional usability and acceptability of an application. An efficacy trial of the iCanFit Web application and development of iCanFit mobile app are both currently underway. After the trial and further refinement, it will be scaled up to assist a large population of older adults with chronic conditions. We are aware that there are many mobile and Web apps that serve similar purposes as iCanFit and users always have many options in terms of mHealth tools. iCanFit was not meant to replace existing physical activity applications; instead, we believe it is a beneficial supplement to the existing ones. Because few mobile or Web apps involved usability testing in older adults [
iCanFit usability testing metrics.
iCanFit user-experience interview guide.
Metabolic equivalent of task
research assistant
This study was supported by a pilot grant from Scott & White Healthcare Senior Health Program and the Mobile Application Program Award from National Library of Medicine South Central Region Network. It also received support from CTxCARES, which is funded by the Centers for Disease Control and Prevention (Cooperative Agreement Number 1U48 DP001924). The findings and conclusions in this manuscript are those of the authors and do not necessarily represent the official positions of the funding agencies. We appreciate support from Angie Alaniz from Center for Community Health Development, Ronnie Gibson in Brazos Valley Area Aging Agency, and Marcie Rodgers in College Station Senior Center, and graduate students including Jairus Pulczinski, Aaron Hammond, Cheolhwan Yoon, Funmi Osofundiya, and William Johnson for their assistance in data collection. We also thank all participants in our study.
None declared.