Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

A Mobile, Avatar-Based App for Improving Body Perceptions Among Adolescents: A Pilot Test

A Mobile, Avatar-Based App for Improving Body Perceptions Among Adolescents: A Pilot Test Background: One barrier to effectively treating weight issues among adolescents is that they tend to use social comparison instead of objective measures to evaluate their own health status. When adolescents correctly perceive themselves as overweight, they are more likely to adopt healthy lifestyle behaviors. Objective: The purpose of this pilot test was to develop and assess acceptability and usability of an avatar-based, theoretically derived mobile app entitled Monitor Your Avatar (MYA). Methods: The MYA app was engineered for high school adolescents to identify, using avatars, what they thought they looked like, what they wanted to look like, and what they actually looked like based on body measurements. Results: The MYA app was pilot-tested with male and female adolescents aged 15-18 years to assess for acceptability and usability. A total of 42 students created and viewed their avatars. The majority of the adolescents were female (28/42, 67%), age 16 years (16/42, 38%), white (35/42, 83%), non-Hispanic (36/42, 86%), in grade 10 (20/42, 48%), healthy weight for females (23/28, 82%), and obese for males (7/14, 50%). The adolescents had positive reactions to the avatar app and being able to view avatars that represented them. All but one student (41/42, 98%) indicated some level of comfort viewing the avatars and would use the app in the future to see how their bodies change over time. Conclusions: Avatar-based mobile apps, such as the MYA app, provide immediate feedback and allow users to engage with images that are personalized to represent their perceptions and actual body images. This pilot study adds to the increasing but limited research of using games to improve health outcomes among high school adolescents. There is a need to further adapt the MYA app and gather feedback from a larger number of high school adolescents, including those from diverse backgrounds. (JMIR Serious Games 2017;5(1):e4) doi: 10.2196/games.6354 KEYWORDS adolescents; avatars; eHealth; mHealth; perceptions; Web-based; usability testing age, adolescents’ participation in physical activity decreases Introduction [1], while concern about weight and negative body image increases [2]. One in five adolescents in the United States are obese, increasing their risk for related complications such as poor quality of life, asthma, type 2 diabetes, and hypertension. Additionally, as they http://games.jmir.org/2017/1/e4/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e4 | p. 1 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Lyles et al One barrier to treating these adolescent weight issues is that 75% own mobile phones [14], and 40% own iPhones [15], they tend to use social comparison instead of objective measures suggesting that Web- and mobile device-based apps can be to evaluate their own health status [3]. Having inaccurate body powerful technologies for implementing behavior change size perceptions may decrease adolescents’ likelihood of programs in this population. changing their health behaviors [4]. If adolescents correctly The purpose of this pilot test was to develop and assess perceive themselves as overweight, they are more likely to adopt acceptability and usability of an avatar-based, theoretically healthy lifestyle behaviors [5] than will those who do not derived, mobile app entitled Monitor Your Avatar (MYA). recognize themselves as overweight [4]. Positive lifestyle changes can affect adolescents’ body Another barrier to effectively treating weight issues among composition, which in turn can affect their perceptions of their adolescents is misclassifying obesity risk due to incomplete bodies, their body shapes, their satisfaction with their bodies, measures for adolescents. Researchers and health care providers and their emotional well-being as they mature into adulthood. have long used body mass index (BMI), a common, inexpensive measure, to determine healthy weight in adolescents. BMI does Methods not distinguish between fat and muscle, does not identify body Design and Development of the App size perceptions, and cannot detect those at risk for unnecessary and unsafe weight loss behaviors. There is a potential for The MYA app was engineered for male and female high school adolescents to not receive appropriate counseling and referral adolescents to identify, using avatars, what they thought they if their body fat percentage, perceptions, and methods for losing looked like, what they wanted to look like, and what they weight are not assessed along with BMI [6]. actually looked like based on body measurements. The MYA app is interactive and designed for adolescents to change specific Adolescents also contend with using the current methods to body parts of the avatars; they can also access the app assess healthy body sizes. Using BMI may be too vague or recurrently to help monitor their targeted goals. The completion abstract for adolescents to comprehend. Their weight in pounds of the app consisted of three phases: classification, development, on a scale may change insignificantly or not at all even though and prototype testing. The three phases of the project were positive lifestyle changes are substantial. Choosing from a series derived from combining a model-driven approach to developing of body figure silhouettes may not be true representations of software [16] along with an iterative user-centered design the adolescents’ body shapes. These methods of assessing body approach to creating mobile health apps [17]. The classification weight classifications and changes made to the bodies may phase required the collection of accurate datasets, the prevent adolescents from seeing the effects that healthy eating development phase developed the models based on this data, and physical activity can have on their health. and the prototype-testing phases incorporated user testing to As children move into adolescence, they are more likely to validate the models developed in the previous stage. All policies, desire independence in attaining and maintaining healthy bodies. procedures, and ethical concerns for all phases were approved Identification (ie, the extent an individual relates to a model by the Institutional Review Board of the university and high and feels similarity to the model) can increase the likelihood school. All parents provided informed consent and all for a teen to perform a learned behavior [3,7]. The power of adolescents provided assent to participate. identification increases when the models are of the same sex Classification Based on Body Scans [4], race [5], or skill level [6]. Fox and Bailenson [8] conducted three studies and found virtual self-models can be an effective The purpose of the classification phase was to classify the impetus for health behavior change. Virtual self-models such models based on body scans for use when creating the three as avatars (ie, computerized representations of the adolescents’ avatars. bodies) can be connected to goal setting, self-monitoring, direct Procedure for Body Scans reinforcement, and social support processes that drive behavior In order to develop a classification, it was necessary to change. Having adolescents interact with avatars on mobile body-scan male and female adolescents 15-18 years of age. devices, such as tablet computers, engages them in viewing and Regarding recruitment, subject inclusion criteria were as thinking about their bodies in a realistic and positive light. follows: (1) in grades 9-12; (2) able to speak, read, and write Visually rich 3D representations that accurately portray how English; (3) able to stand on a turntable for the 1-minute body bodies appear may have a greater impact on adolescent behavior scan; and (4) comfortable wearing form-fitting clothes such as than can a number on a scale or BMI percentage. In addition, compression shorts, leggings, form-fitting tank top, or adolescents prefer Internet-based health resources because of swimwear. MyBodee by Styku [15] was used to scan the the 24-hour availability and lack of perceived judgment and adolescents. MyBodee [15] is a portable and highly accurate conflict with sensitive topics [9]. Mobile phone technology has body measurement technology. The adolescents wore the ability to track everyday behavior changes in an unobtrusive form-fitting clothes and stood on a turntable. Using a tablet and way, in real time [10,11], and with the potential to provide 3D camera, the researcher measured each adolescent’s full body immediate feedback [12]. shape. The adolescent slowly spun a full 360 degrees on a safe and automated turntable. After 40 seconds, the adolescent’s Researchers and providers can have a greater impact on scan was sent to a secure, private network where the research adolescents’ actions to change their bodies or help maintain team concurrently analyzed the shape and body measurements. healthy bodies by providing interactive and dynamic strategies. The research team also collected additional body measurements, Further, over 93% of teens are active users of the Internet [13], http://games.jmir.org/2017/1/e4/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e4 | p. 2 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Lyles et al including the widths and circumferences of the chest, biceps, Display waist, hips, thigh, and calf using a body tape measure. The scan The models from the Blender software were exported as a single did not measure the widths of these body parts, so collection of JSON file [21] that contained the location and name of all the the circumferences and widths of the body parts was needed to baseline models. Once a representative sample was selected configure ratios to be used during the classification phase. A based on height, weight, and gender input we used Three.js total of 47 adolescents 15-18 years of age were scanned—24 [22], a Web-based, 3D-rendering application programming male and 23 female—using MyBodee [15]. MyBodee [15] was interface, for loading and displaying the appropriate baseline only needed to scan the adolescents so their body scans could model with the manipulating sliders. The manipulation of the be used to classify the models for use in the app. body parts is done independently by an algorithm that morphs the 3D object based on predetermined minimum and maximum Classification 3D surfaces for each body part. The Actual Avatar is generated Before creating the models, the research team classified the by having the users input the body measurements and scanned models using the ratio of height and weight. According algorithmically calculate the morph targets for this input set. to the ratio, we classified both male and female models into six groups. In each group, we selected one or two of the most Analysis representative models and reviewed their measurements. The When the user completed the three avatars, the measurements measurements included height and weight as well as the widths of all body parts for all the avatars were retrieved and the and circumferences of the chest (breast/bust), waist, hips, biceps, WebGL content was saved into an image format. The thigh, and calf. These measurements were used to create baseline measurements of body parts were saved onto a spreadsheet. All models for the app using the MakeHuman software three avatars were compiled onto a screen capture and the final (MakeHuman) [18]. state of the app was saved. A server-side script was written to give administrative access to the research team to easily Development of the Monitor Your Avatar App Avatars download this data for all the adolescents. Overview Prototype Testing of the Monitor Your Avatar App The purpose of the development phase was to use the classified models and build a pipeline to display and manipulate them via Overview a Web app. The Web app called MYA provides an interface The purpose of the third phase was to determine the acceptability (see Figure 1) for the participants to engage with the following and usability of the MYA app in male and female adolescents avatars: Perceived Avatar (what adolescents think they look aged 15-18 years. Acceptability was defined as thoughts on like), Target Avatar (what adolescents want to look like), and viewing the avatars and visual representation, as well as comfort the Actual Avatar (what adolescents actually look like based level of using, intent to use, and satisfaction with the app. on body measurements). Usability was defined as ease of using and actual use the app, need to look for help when using the app, understanding of how Procedure for Prototype Development to use the app, and engagement of the app. The prototype development consisted of three steps: (1) Modeling: generating the baseline models using MakeHuman Procedure for Prototype Testing [18]; (2) Display: rendering the models via a Web interface Participant inclusion criteria included the following: (1) in written in JavaScript and WebGL [19]; and (3) Analysis: saving grades 9-12; (2) able to speak, read, and write English; (3) has the user-generated avatar content to a secure database for not been diagnosed with an eating disorder or depression, as research and analysis. provided by a school nurse; and (4) be comfortable wearing form-fitting clothes such as compression shorts, leggings, Modeling form-fitting tank top, or swimwear for taking measurements. With the provided measurements, the models were created by These adolescents were not part of the body scan study setting up the gender, height, and age in MakeHuman [18]. The completed for the classification phase. The research team scanned model and measurements were used as a reference to introduced the study to 45 adolescents in two high school match the body type and create the baseline model that is physical education classes. One adolescent elected not to be geometrically similar to the scanned model. MakeHuman models part of the study and two others were absent on the day of data are not fragmented into body parts and manipulating one area collection. had no effect on the other. Since we were seeking local control and manipulation based on measurements of specific body parts, During data collection, the adolescents wore form-fitting clothes. we had to use the Blender 3D software (Blender) [20] to The research team measured the height of the adolescents using fragment the model into body parts. A new model whose a stadiometer and their weight and body fat percentage using a individual body parts could be manipulated by varying the Tanita body composition analyzer. We also measured the measurements using sliders was created. With this model, the adolescents’ body parts—biceps, chest/bust, waist, hips, thigh, slider could be used to manipulate each body part independently and calf—using a body tape measure to obtain circumferences. and provide the ability to export the functionality to the Web. The girls also had their bust girth measured, in addition to their chest/bust, for entry into the Actual Avatar feature. After measurement, the adolescents entered an assigned identification number, their gender, measured weight, and http://games.jmir.org/2017/1/e4/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e4 | p. 3 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Lyles et al measured height into the app to populate a baseline avatar using The open-ended reaction questions posed to the adolescents the computer located in the school library. They then designed included the following: “What do you think about being able the avatar to represent how they currently perceive their bodies to view avatars of yourself?” “Do you think the avatars are a to look (Perceived Avatar). The app allowed each adolescent good representation of yourself? Explain why or why not.” to make the body parts (ie, biceps, chest/bust, bust girth, waist, “How comfortable were you creating and viewing your avatars?” hips, thighs, and calves) bigger or smaller using the slider. Their “In the future, would you use these avatars to see how your completed Perceived Avatar then generated on the same screen body parts are changing? Explain why or why not.” “What do so they could design it to represent how they wanted their bodies you like about the avatars?” “What do you not like about the to look within realistic, healthy parameters (Target Avatar). avatars?” “What would you change about the avatars?” This feature was incorporated so the adolescents could work The Software Usability Survey is a short usability 7-point from their current perception rather than having to start over. Likert-type scale survey informed by the product and best On the same screen and to the right of the Perceived and Target practices in software engineering [23,24]. The items used to Avatars, the adolescents generated an Actual Avatar by entering identify usability for the MYA app included the following: (1) their body part measurements into the app. Figure 2 displays simplicity of navigation from one page to another, (2) ease of the three avatars after completion. The avatars could be rotated control to view the avatars and their body parts, (3) ease to 360 degrees. Figure 3 displays the three avatars from the side manipulate each body part and make it look more muscular, (4) view. need to look for help when working with the app, (5) The adolescents were then asked to answer questions about their understanding of what the three avatars are and how to view reactions to creating and viewing the avatars to assess and manipulate them, (6) ability to keep the user engaged and acceptability and to complete the Software Usability Survey. not quit, and (7) overall satisfaction with ease of using the app. Figure 1. User Interaction Screens for MYA App. http://games.jmir.org/2017/1/e4/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e4 | p. 4 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Lyles et al Figure 2. Example of three female avatars (front view). http://games.jmir.org/2017/1/e4/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e4 | p. 5 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Lyles et al Figure 3. Example of three female avatars (side view). results. The frequencies of the Software Usability Survey were Analysis analyzed and calculated into percentages. SPSS version 23 (IBM Corp) was used to conduct the descriptive statistics of the sample. The adolescents’ reactions Results to the MYA app were analyzed using content analysis [25]. The content analysis process included three phases: preparation, The sample characteristics for prototype testing are highlighted organizing, and reporting. The preparation phase included in Table 1, with the majority of the adolescents being female selecting the unit of analysis and classifying the data as a whole. (28/42, 67%), age 16 years (16/42, 38%), white (35/42, 83%), The organizing phase included developing a categorization non-Hispanic (36/42, 86%), in grade 10 (20/42, 48%), healthy matrix and coding the data per the categories. The reporting weight for females (23/28, 82%), and obese for males (7/14, phase included the results of the analyzing process and the 50%). http://games.jmir.org/2017/1/e4/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e4 | p. 6 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Lyles et al Table 1. Sample characteristics of the adolescents (N=42). Sample characteristics n (%) Age in years 15 12 (29) 16 16 (38) 17 11 (26) 18 3 (7) Gender Female 28 (67) Male 14 (33) Race White 35 (83) More than one race 2 (5) Asian 2 (5) Black American Indian 2 (5) American Indian 1 (2) Ethnicity Hispanic 6 (14) Non-Hispanic 36 (86) Grade level 9 2 (5) 10 20 (47) 11 16 (38) 12 4 (10) a b Female BMI -for-age category Underweight 0 (0) Healthy weight 23 (82) Overweight 4 (14) Obese 1 (4) Male BMI-for-age category Underweight 0 (0) Healthy weight 4 (29) Overweight 3 (21) Obese 7 (50) BMI: body mass index. Underweight: less than the 5th percentile; healthy weight: 5th percentile up to the 85th percentile; overweight: 85th to less than the 95th percentile; obese: equal to or greater than the 95th percentile. To determine acceptability, the adolescents’ responses to the Another adolescent compared the app to the mirror: “It's nice reaction questions (see Table 2) were analyzed using content to see what I actually look like opposed to looking in the mirror” analysis [25]. When asked what the adolescents thought about (obese, 16-year-old male). When asked if the avatars were a being able to view the avatars of themselves, 37 out of 42 good representation of themselves, 38 out 42 (90%) responded adolescents (88%) provided positive feedback. For example, “yes.” One adolescent explained the following: “Yes, because one adolescent commented about the actual and target it shows what my body looks like and what I think it looks like” representations: “I think it's good to see what you actually look (healthy-weight, 16-year-old female). like, then being able to see what your personal target is” (healthy-weight, 15-year-old female). http://games.jmir.org/2017/1/e4/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e4 | p. 7 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Lyles et al Another adolescent indicated how this was something new: There were several reasons the adolescents liked the avatars. “Yes, I've never done something like this. I think it is a good Almost half of the adolescents (18/42, 43%) indicated they liked representation” (overweight, 17-year-old male). One adolescent how the avatars were realistic, while almost a third (15/42, 29%) did not think it was a good representation: “Not really, because liked how they could move the avatars to see them from different I think more measurements should be taken” (obese, 18-year-old angles (see Table 2). Adolescents commented on what they male). liked: When asked if they were comfortable viewing the avatars, 41 That they are suited for the gender that you are and body type. out of 42 (98%) indicated they were comfortable. The research [Healthy-weight, 17-year-old female]. team categorized the level of comfort into three categories: very I like how they're realistic. [Healthy-weight, 15-year-old comfortable, somewhat comfortable, and comfortable based on female]. the explanations the adolescents provided (see Table 2). Almost half of the adolescents (18/42, 43%) were very comfortable How you can play around with them and see what you look like creating and viewing their avatars. When asked if the with certain numbers. [Obese, 16-year-old male]. adolescents would use the avatars in the future to see how their When asked what the adolescents did not like about the avatars, bodies changed, 41 out of 42 (98%) responded “yes” (see Table approximately one-third (13/42, 31%) indicated there were not 2). enough details or customization options with the avatars, while The adolescents explained the reasons they would use the app 13 out of 42 (31%) commented there was nothing to dislike. in the future. Most of the adolescents indicated they would use They do not show definition of muscle. (Obese, 16-year-old the app to track progress and goals (15/42, 36%) and to see the male). changes or differences in their body parts (11/42, 26%). The following are examples of why the adolescents would use the They don't have as much detail, with looks and more avatars in the future: measurements. (Healthy-weight, 16-year-old female). Yes, it would be very helpful because sometimes your The suggestions the adolescents made when asked about what progress can't be seen on the body but is realized in they would change about the avatars included the appearance measurements or on other body/perspectives. of the avatars, such as the clothing, hair, etc (16/18, 38%). One [Healthy-weight, 16-year-old female] participant stated, “I think I would change the avatars to look more like the person making the avatar so they can really feel Yes, because visual representation would help build confidence more into it and help view the different avatars of the person” in achieving the goal. [Overweight, 16-year-old female] (healthy-weight, 16-year-old female). Yes, I would like to see the pictures of my before and after results. [Obese, 16-year-old male] http://games.jmir.org/2017/1/e4/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e4 | p. 8 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Lyles et al Table 2. Responses to the questions identified by the adolescents (N=42). Questions and answers n (%) What do you think about being able to view avatars of yourself? Positive feedback 37 (88) Neutral feedback 4 (10) Negative feedback 1 (2) Do you think the avatars are a good representation of yourself? Yes 38 (90) No 4 (10) How comfortable were you creating and viewing your avatars? Very comfortable 18 (43) Comfortable 15 (36) Somewhat comfortable 8 (19) Neutral 1 (2) In the future, would you use these avatars to see how your body parts are changing? Yes 41 (98) No 1 (2) Reasons the teenagers would use the app in the future To track progress/goals 15 (36) To see the changes/differences in body parts 11 (26) Helpful to see visual representations 8 (19) Accurate representations 4 (10) It was fun/interesting 2 (5) Yes, with no explanation 1 (2) No 1 (2) What do you like about the avatars? Exact/actually look like/realistic 18 (43) Ability to move them/3D 12 (29) Ability to compare avatars 4 (10) It was fun/cool 3 (7) Others’ perspective 2 (5) App is anonymous 1 (2) Target Avatar 1 (2) No response 1 (2) What do you not like about the avatars? Needs more details/customization 13 (31) Nothing 13 (31) Appearance of avatars/clothes 8 (19) Difficulty seeing body changes made 3 (7) Not accurate 2 (5) Uncomfortable to see self 2 (5) Difficulty moving avatars 1 (2) What would you change about the avatars? Appearance of avatars/clothing 16 (38) http://games.jmir.org/2017/1/e4/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e4 | p. 9 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Lyles et al Questions and answers n (%) Nothing 9 (22) Add more details 8 (19) Able to see body changes better 7 (17) Able to move avatars better 1 (2) Add diet to the app 1 (2) The results of the 7-point Software Usability Survey indicated more voluntary physical activity than those who saw an that the adolescents found the app simple to navigate (rating of unchanging virtual self or no virtual representation [8]. 6 or 7; 40/42, 95%), found it easy to control the viewing of the Ridgers and colleagues [27] conducted a systematic review of avatars (rating of 6 or 7; 31/42, 74%), and found it easy to the effectiveness of youth wearable tracker devices to increase manipulate each body part and make it look muscular (rating physical activity levels among children and adolescents. They of 6 or 7; 24/42, 57%). The adolescents had the need to look found these devices have the potential to increase activity levels for help several times (rating of 1 or 2; 42/42, 100%), had a as the adolescents self-monitor their progress and set their goals. better understanding of what the three avatars are and how to The MYA app is similar in that it can be used to provide view them (rating of 6 or 7; 40/42, 95%), were engaged by the feedback of their body shape progress and monitor their goals. app (rating of 6 or 7; 41/42, 98%), and experienced overall Ridgers and colleagues [27] recommended that research be satisfaction with the ease of the app (rating of 6 or 7; 41/42, conducted to establish how youth engage with technology over 98%). longer periods of time. The adolescents indicated future use of the MYA app; however, more research is needed to examine Discussion their engagement with the app over time. Principal Findings Limitations The purpose of the prototype testing was to assess the This study is not without limitations. Due to the sample of acceptability and usability of an avatar-based, theoretically adolescents being relatively small and nondiverse, there was derived mobile app, the MYA app. Prior to the app being ready not enough power to detect group differences. Another limitation for prototype testing, two phases were completed that consisted is that the adolescents only tested the app at one session. To of classification and development. Once completed, the app understand the acceptability and usability over time, it is prototype was tested with a sample of male and female high necessary for the adolescents to interact with the app over a school adolescents. The adolescents created and viewed the longer period of time with several recurrent uses. The app did three avatars: Perceived Avatar, Target Avatar, and Actual not have this capability at the time of testing. Avatar. The adolescents had positive reactions to the avatar app Implications and being able to view avatars that represented them. Almost all of the adolescents indicated some level of comfort viewing This app allows teens to have visual representations of their the avatars and would use the app in the future to see how their bodies—perceived, target, and actual avatars— interactive bodies change over time. individualized programs, access to immediate feedback, and a method to monitor changes to their bodies as a result of positive Adolescents who evaluated a Web-based substance abuse changes in health behaviors. This visual representation is intervention also indicated high ratings for overall usability different from what adolescents currently have access to, such features such as ease of use and future use [26]. The ease of as digital scales and BMI calculators. Adolescents continue to manipulation of the body parts and making the avatars look be overweight and obese while physical activity decreases as more muscular resulted in a lower-percentage rating from the they get older. Adolescents who identify with and use these adolescents than the other usability survey items. The responses avatars who represent them can monitor and reach their body regarding the viewing of the avatars support this lower score. shape and health goals. While this app may be user driven, it The adolescents indicated that more details and customization also has the capability for health care providers, parents, and of the avatar bodies were needed and that it was difficult to see adolescents to have a visual communication tool to discuss the changes made to the avatars. These responses may explain adolescents’ body perceptions and goals. the lower ease of manipulation score. In further adaptations, there is a need for the avatar bodies to be more detailed with an Conclusions easier view of the body changes from one avatar to the next. Avatar-based mobile apps, such as the MYA app, provide immediate feedback and allow users to engage with images that The adolescents indicated they would use the avatars in the are personalized to represent their perceptions and actual body MYA app in the future to monitor their body shape changes. images. Adolescents appreciate the use of weight-related This is similar to participants who were assigned a virtual messages that are personalized for them [28]. The high school representation of their physical selves, an unchanging virtual students found the avatars of the MYA app easy to use, were representation, or no virtual representation. Those who witnessed comfortable viewing their avatars, and would use the avatars the virtual representation of their physical selves engaged in in the future to monitor how their body changes. Overall, they http://games.jmir.org/2017/1/e4/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e4 | p. 10 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Lyles et al were satisfied with the MYA app; however, they made among high school adolescents. There is a need to further adapt suggestions on how to make the app more appealing to high the MYA app and gather feedback from a larger number of high school males and females. This pilot study adds to the increasing school adolescents, including those from disadvantaged but limited research of using games to improve health outcomes backgrounds, while testing adherence to the app over time. Acknowledgments We would like to thank the graduate engineering students for help in developing the app prototype for testing. We are grateful for the local high school’s athletic trainer, graduate student trainers, and the high school students willing to be a part of these studies on body perceptions. Conflicts of Interest None declared. References 1. Bruner MW, Chad KE, Beattie-Flath JA, Humbert ML, Verrall TC, Vu L, et al. Examination of physical activity in adolescents over the school year. Pediatr Exerc Sci 2009 Nov;21(4):421-435. [Medline: 20128362] 2. Ingledew D, Sullivan G. Effects of body mass and body image on exercise motives in adolescence. Psychol Sport Exerc 2002;3(4):323-338. [doi: 10.1016/S1469-0292(01)00029-2] 3. Oenema A, Brug J. Exploring the occurrence and nature of comparison of one's own perceived dietary fat intake to that of self-selected others. Appetite 2003 Dec;41(3):259-264. [Medline: 14637324] 4. Bittner FH, Diamond J, Myers R, Gill JM. Perception, intention, and action in adolescent obesity. J Am Board Fam Med 2008;21(6):555-561 [FREE Full text] [doi: 10.3122/jabfm.2008.06.070184] [Medline: 18988723] 5. Maximova K, McGrath JJ, Barnett T, O'Loughlin J, Paradis G, Lambert M. Do you see what I see? Weight status misperception and exposure to obesity among children and adolescents. Int J Obes (Lond) 2008 Jun;32(6):1008-1015. [doi: 10.1038/ijo.2008.15] [Medline: 18317474] 6. Liechty JM. Body image distortion and three types of weight loss behaviors among nonoverweight girls in the United States. J Adolesc Health 2010 Aug;47(2):176-182. [doi: 10.1016/j.jadohealth.2010.01.004] [Medline: 20638010] 7. Davis R, Campbell R, Hildon Z, Hobbs L, Michie S. Theories of behaviour and behaviour change across the social and behavioural sciences: A scoping review. Health Psychol Rev 2015 Sep;9(3):323-344 [FREE Full text] [doi: 10.1080/17437199.2014.941722] [Medline: 25104107] 8. Bandura A. Social cognitive theory of mass communication. In: Bryant J, Zillmann D, editors. Media Effects: Advances in Theory and Research. 2nd edition. Mahwah, NJ: Lawrence Erlbaum Associates; 2008:121-154. 9. Skinner H, Biscope S, Poland B, Goldberg E. How adolescents use technology for health information: Implications for health professionals from focus group studies. J Med Internet Res 2003 Dec 18;5(4):e32 [FREE Full text] [doi: 10.2196/jmir.5.4.e32] [Medline: 14713660] 10. Randall W, Rickard N. Development and trial of a mobile experience sampling method (m-ESM) for personal music listening. Music Percept 2013;31(2):157. [doi: 10.1525/mp.2013.31.2.157] 11. Reid SC, Kauer SD, Khor AS, Hearps SJ, Sanci LA, Kennedy AD, et al. Using a mobile phone application in youth mental health - An evaluation study. Aust Fam Physician 2012 Sep;41(9):711-714. [Medline: 22962650] 12. Harrison V, Proudfoot J, Wee PP, Parker G, Pavlovic DH, Manicavasagar V. Mobile mental health: Review of the emerging field and proof of concept study. J Ment Health 2011 Dec;20(6):509-524. [doi: 10.3109/09638237.2011.608746] [Medline: 21988230] 13. Bandura A, Huston AC. Identification as a process of incidental learning. J Abnorm Soc Psychol 1961 Sep;63(2):311-318. [doi: 10.1037/h0040351] [Medline: 13864604] 14. Fisher JD, Fisher WA, Bryan AD, Misovich SJ. Information-motivation-behavioral skills model-based HIV risk behavior change intervention for inner-city high school youth. Health Psychol 2002 Mar;21(2):177-186. [Medline: 11950108] 15. Tukatech. MyBodee URL: http://www.tukatech.com/styku [accessed 2017-02-24] [WebCite Cache ID 6oWP313RW] 16. Stahl T, Völter M, Bettin J, Haase A, Helsen S. Model-Driven Software Development: Technology, Engineering, Management. Chichester, UK: John Wiley & Sons; 2006. 17. McCurdie T, Taneva S, Casselman M, Yeung M, McDaniel C, Ho W, et al. mHealth consumer apps: The case for user-centered design. Biomed Instrum Technol 2012;Suppl:49-56. [doi: 10.2345/0899-8205-46.s2.49] [Medline: 23039777] 18. Bastioni M, Re S, Misra S. Ideas and methods for modeling 3D human figures: The principal algorithms used by MakeHuman and their implementation in a new approach to parametric modeling. In: Proceedings of the 1st Bangalore Annual Compute Conference. New York, NY: ACM; 2008 Presented at: The 1st Bangalore Annual Compute Conference; January 18-20, 2008; Bangalore, India. 19. Marrin C. Khronos. 2011 Feb 10. WebGL Specification URL: https://www.khronos.org/registry/webgl/specs/1.0.0/ [accessed 2017-02-14] [WebCite Cache ID 6oHfuZojg] http://games.jmir.org/2017/1/e4/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e4 | p. 11 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Lyles et al 20. Blender. URL: https://www.blender.org/ [accessed 2017-02-10] [WebCite Cache ID 6oAqLWOOP] 21. Crockford D. IETF Tools. 2006 Jul. The application/json media type for JavaScript Object Notation (json) URL: https:/ /tools.ietf.org/html/rfc4627 [accessed 2017-02-14] [WebCite Cache ID 6oHg2zld4] 22. Cabello R. GitHub. 2010. JavaScript 3D library URL: https://github.com/mrdoob/three.js/ [accessed 2017-02-14] [WebCite Cache ID 6oHgKBEWz] 23. Brown W, Yen P, Rojas M, Schnall R. Assessment of the Health IT Usability Evaluation Model (Health-ITUEM) for evaluating mobile health (mHealth) technology. J Biomed Inform 2013 Dec;46(6):1080-1087 [FREE Full text] [doi: 10.1016/j.jbi.2013.08.001] [Medline: 23973872] 24. Brooke J. SUS: A 'quick and dirty' usability scale. In: Jordan PW, Thomas B, Weerdmeester BA, McClelland IL, editors. Usability Evaluation in Industry. London, UK: Taylor & Francis; 1996:189-194. 25. Elo S, Kyngäs H. The qualitative content analysis process. J Adv Nurs 2008 Apr;62(1):107-115. [doi: 10.1111/j.1365-2648.2007.04569.x] [Medline: 18352969] 26. Raghupathy S, Forth AL. The HAWK2 program: A computer-based drug prevention intervention for Native American youth. Am J Drug Alcohol Abuse 2012 Sep;38(5):461-467. [doi: 10.3109/00952990.2012.694531] [Medline: 22823599] 27. Ridgers ND, McNarry MA, Mackintosh KA. Feasibility and effectiveness of using wearable activity trackers in youth: A systematic review. JMIR Mhealth Uhealth 2016 Nov 23;4(4):e129 [FREE Full text] [doi: 10.2196/mhealth.6540] [Medline: 27881359] 28. Woolford SJ, Clark SJ, Strecher VJ, Resnicow K. Tailored mobile phone text messages as an adjunct to obesity treatment for adolescents. J Telemed Telecare 2010;16(8):458-461 [FREE Full text] [doi: 10.1258/jtt.2010.100207] [Medline: 20959393] Abbreviations BMI: body mass index MYA: Monitor Your Avatar Edited by G Eysenbach; submitted 14.07.16; peer-reviewed by J Chaplin, R Schnall, PY Yen, W Brown, M Nitsch; comments to author 20.11.16; revised version received 06.01.17; accepted 03.02.17; published 02.03.17 Please cite as: Lyles AA, Amresh A, Huberty J, Todd M, Lee RE JMIR Serious Games 2017;5(1):e4 URL: http://games.jmir.org/2017/1/e4/ doi: 10.2196/games.6354 PMID: 28254737 ©Annmarie A Lyles, Ashish Amresh, Jennifer Huberty, Michael Todd, Rebecca E Lee. Originally published in JMIR Serious Games (http://games.jmir.org), 02.03.2017. 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 Serious Games, is properly cited. The complete bibliographic information, a link to the original publication on http://games.jmir.org, as well as this copyright and license information must be included. http://games.jmir.org/2017/1/e4/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e4 | p. 12 (page number not for citation purposes) XSL FO RenderX http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png JMIR Serious Games JMIR Publications

A Mobile, Avatar-Based App for Improving Body Perceptions Among Adolescents: A Pilot Test

Download PDF
12 pages

Loading next page...
 
/lp/jmir-publications/a-mobile-avatar-based-app-for-improving-body-perceptions-among-YKD2UrAHAU

References (23)

Publisher
JMIR Publications
Copyright
Copyright © The Author(s). Licensed under Creative Commons Attribution cc-by 4.0
ISSN
2291-9279
DOI
10.2196/games.6354
Publisher site
See Article on Publisher Site

Abstract

Background: One barrier to effectively treating weight issues among adolescents is that they tend to use social comparison instead of objective measures to evaluate their own health status. When adolescents correctly perceive themselves as overweight, they are more likely to adopt healthy lifestyle behaviors. Objective: The purpose of this pilot test was to develop and assess acceptability and usability of an avatar-based, theoretically derived mobile app entitled Monitor Your Avatar (MYA). Methods: The MYA app was engineered for high school adolescents to identify, using avatars, what they thought they looked like, what they wanted to look like, and what they actually looked like based on body measurements. Results: The MYA app was pilot-tested with male and female adolescents aged 15-18 years to assess for acceptability and usability. A total of 42 students created and viewed their avatars. The majority of the adolescents were female (28/42, 67%), age 16 years (16/42, 38%), white (35/42, 83%), non-Hispanic (36/42, 86%), in grade 10 (20/42, 48%), healthy weight for females (23/28, 82%), and obese for males (7/14, 50%). The adolescents had positive reactions to the avatar app and being able to view avatars that represented them. All but one student (41/42, 98%) indicated some level of comfort viewing the avatars and would use the app in the future to see how their bodies change over time. Conclusions: Avatar-based mobile apps, such as the MYA app, provide immediate feedback and allow users to engage with images that are personalized to represent their perceptions and actual body images. This pilot study adds to the increasing but limited research of using games to improve health outcomes among high school adolescents. There is a need to further adapt the MYA app and gather feedback from a larger number of high school adolescents, including those from diverse backgrounds. (JMIR Serious Games 2017;5(1):e4) doi: 10.2196/games.6354 KEYWORDS adolescents; avatars; eHealth; mHealth; perceptions; Web-based; usability testing age, adolescents’ participation in physical activity decreases Introduction [1], while concern about weight and negative body image increases [2]. One in five adolescents in the United States are obese, increasing their risk for related complications such as poor quality of life, asthma, type 2 diabetes, and hypertension. Additionally, as they http://games.jmir.org/2017/1/e4/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e4 | p. 1 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Lyles et al One barrier to treating these adolescent weight issues is that 75% own mobile phones [14], and 40% own iPhones [15], they tend to use social comparison instead of objective measures suggesting that Web- and mobile device-based apps can be to evaluate their own health status [3]. Having inaccurate body powerful technologies for implementing behavior change size perceptions may decrease adolescents’ likelihood of programs in this population. changing their health behaviors [4]. If adolescents correctly The purpose of this pilot test was to develop and assess perceive themselves as overweight, they are more likely to adopt acceptability and usability of an avatar-based, theoretically healthy lifestyle behaviors [5] than will those who do not derived, mobile app entitled Monitor Your Avatar (MYA). recognize themselves as overweight [4]. Positive lifestyle changes can affect adolescents’ body Another barrier to effectively treating weight issues among composition, which in turn can affect their perceptions of their adolescents is misclassifying obesity risk due to incomplete bodies, their body shapes, their satisfaction with their bodies, measures for adolescents. Researchers and health care providers and their emotional well-being as they mature into adulthood. have long used body mass index (BMI), a common, inexpensive measure, to determine healthy weight in adolescents. BMI does Methods not distinguish between fat and muscle, does not identify body Design and Development of the App size perceptions, and cannot detect those at risk for unnecessary and unsafe weight loss behaviors. There is a potential for The MYA app was engineered for male and female high school adolescents to not receive appropriate counseling and referral adolescents to identify, using avatars, what they thought they if their body fat percentage, perceptions, and methods for losing looked like, what they wanted to look like, and what they weight are not assessed along with BMI [6]. actually looked like based on body measurements. The MYA app is interactive and designed for adolescents to change specific Adolescents also contend with using the current methods to body parts of the avatars; they can also access the app assess healthy body sizes. Using BMI may be too vague or recurrently to help monitor their targeted goals. The completion abstract for adolescents to comprehend. Their weight in pounds of the app consisted of three phases: classification, development, on a scale may change insignificantly or not at all even though and prototype testing. The three phases of the project were positive lifestyle changes are substantial. Choosing from a series derived from combining a model-driven approach to developing of body figure silhouettes may not be true representations of software [16] along with an iterative user-centered design the adolescents’ body shapes. These methods of assessing body approach to creating mobile health apps [17]. The classification weight classifications and changes made to the bodies may phase required the collection of accurate datasets, the prevent adolescents from seeing the effects that healthy eating development phase developed the models based on this data, and physical activity can have on their health. and the prototype-testing phases incorporated user testing to As children move into adolescence, they are more likely to validate the models developed in the previous stage. All policies, desire independence in attaining and maintaining healthy bodies. procedures, and ethical concerns for all phases were approved Identification (ie, the extent an individual relates to a model by the Institutional Review Board of the university and high and feels similarity to the model) can increase the likelihood school. All parents provided informed consent and all for a teen to perform a learned behavior [3,7]. The power of adolescents provided assent to participate. identification increases when the models are of the same sex Classification Based on Body Scans [4], race [5], or skill level [6]. Fox and Bailenson [8] conducted three studies and found virtual self-models can be an effective The purpose of the classification phase was to classify the impetus for health behavior change. Virtual self-models such models based on body scans for use when creating the three as avatars (ie, computerized representations of the adolescents’ avatars. bodies) can be connected to goal setting, self-monitoring, direct Procedure for Body Scans reinforcement, and social support processes that drive behavior In order to develop a classification, it was necessary to change. Having adolescents interact with avatars on mobile body-scan male and female adolescents 15-18 years of age. devices, such as tablet computers, engages them in viewing and Regarding recruitment, subject inclusion criteria were as thinking about their bodies in a realistic and positive light. follows: (1) in grades 9-12; (2) able to speak, read, and write Visually rich 3D representations that accurately portray how English; (3) able to stand on a turntable for the 1-minute body bodies appear may have a greater impact on adolescent behavior scan; and (4) comfortable wearing form-fitting clothes such as than can a number on a scale or BMI percentage. In addition, compression shorts, leggings, form-fitting tank top, or adolescents prefer Internet-based health resources because of swimwear. MyBodee by Styku [15] was used to scan the the 24-hour availability and lack of perceived judgment and adolescents. MyBodee [15] is a portable and highly accurate conflict with sensitive topics [9]. Mobile phone technology has body measurement technology. The adolescents wore the ability to track everyday behavior changes in an unobtrusive form-fitting clothes and stood on a turntable. Using a tablet and way, in real time [10,11], and with the potential to provide 3D camera, the researcher measured each adolescent’s full body immediate feedback [12]. shape. The adolescent slowly spun a full 360 degrees on a safe and automated turntable. After 40 seconds, the adolescent’s Researchers and providers can have a greater impact on scan was sent to a secure, private network where the research adolescents’ actions to change their bodies or help maintain team concurrently analyzed the shape and body measurements. healthy bodies by providing interactive and dynamic strategies. The research team also collected additional body measurements, Further, over 93% of teens are active users of the Internet [13], http://games.jmir.org/2017/1/e4/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e4 | p. 2 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Lyles et al including the widths and circumferences of the chest, biceps, Display waist, hips, thigh, and calf using a body tape measure. The scan The models from the Blender software were exported as a single did not measure the widths of these body parts, so collection of JSON file [21] that contained the location and name of all the the circumferences and widths of the body parts was needed to baseline models. Once a representative sample was selected configure ratios to be used during the classification phase. A based on height, weight, and gender input we used Three.js total of 47 adolescents 15-18 years of age were scanned—24 [22], a Web-based, 3D-rendering application programming male and 23 female—using MyBodee [15]. MyBodee [15] was interface, for loading and displaying the appropriate baseline only needed to scan the adolescents so their body scans could model with the manipulating sliders. The manipulation of the be used to classify the models for use in the app. body parts is done independently by an algorithm that morphs the 3D object based on predetermined minimum and maximum Classification 3D surfaces for each body part. The Actual Avatar is generated Before creating the models, the research team classified the by having the users input the body measurements and scanned models using the ratio of height and weight. According algorithmically calculate the morph targets for this input set. to the ratio, we classified both male and female models into six groups. In each group, we selected one or two of the most Analysis representative models and reviewed their measurements. The When the user completed the three avatars, the measurements measurements included height and weight as well as the widths of all body parts for all the avatars were retrieved and the and circumferences of the chest (breast/bust), waist, hips, biceps, WebGL content was saved into an image format. The thigh, and calf. These measurements were used to create baseline measurements of body parts were saved onto a spreadsheet. All models for the app using the MakeHuman software three avatars were compiled onto a screen capture and the final (MakeHuman) [18]. state of the app was saved. A server-side script was written to give administrative access to the research team to easily Development of the Monitor Your Avatar App Avatars download this data for all the adolescents. Overview Prototype Testing of the Monitor Your Avatar App The purpose of the development phase was to use the classified models and build a pipeline to display and manipulate them via Overview a Web app. The Web app called MYA provides an interface The purpose of the third phase was to determine the acceptability (see Figure 1) for the participants to engage with the following and usability of the MYA app in male and female adolescents avatars: Perceived Avatar (what adolescents think they look aged 15-18 years. Acceptability was defined as thoughts on like), Target Avatar (what adolescents want to look like), and viewing the avatars and visual representation, as well as comfort the Actual Avatar (what adolescents actually look like based level of using, intent to use, and satisfaction with the app. on body measurements). Usability was defined as ease of using and actual use the app, need to look for help when using the app, understanding of how Procedure for Prototype Development to use the app, and engagement of the app. The prototype development consisted of three steps: (1) Modeling: generating the baseline models using MakeHuman Procedure for Prototype Testing [18]; (2) Display: rendering the models via a Web interface Participant inclusion criteria included the following: (1) in written in JavaScript and WebGL [19]; and (3) Analysis: saving grades 9-12; (2) able to speak, read, and write English; (3) has the user-generated avatar content to a secure database for not been diagnosed with an eating disorder or depression, as research and analysis. provided by a school nurse; and (4) be comfortable wearing form-fitting clothes such as compression shorts, leggings, Modeling form-fitting tank top, or swimwear for taking measurements. With the provided measurements, the models were created by These adolescents were not part of the body scan study setting up the gender, height, and age in MakeHuman [18]. The completed for the classification phase. The research team scanned model and measurements were used as a reference to introduced the study to 45 adolescents in two high school match the body type and create the baseline model that is physical education classes. One adolescent elected not to be geometrically similar to the scanned model. MakeHuman models part of the study and two others were absent on the day of data are not fragmented into body parts and manipulating one area collection. had no effect on the other. Since we were seeking local control and manipulation based on measurements of specific body parts, During data collection, the adolescents wore form-fitting clothes. we had to use the Blender 3D software (Blender) [20] to The research team measured the height of the adolescents using fragment the model into body parts. A new model whose a stadiometer and their weight and body fat percentage using a individual body parts could be manipulated by varying the Tanita body composition analyzer. We also measured the measurements using sliders was created. With this model, the adolescents’ body parts—biceps, chest/bust, waist, hips, thigh, slider could be used to manipulate each body part independently and calf—using a body tape measure to obtain circumferences. and provide the ability to export the functionality to the Web. The girls also had their bust girth measured, in addition to their chest/bust, for entry into the Actual Avatar feature. After measurement, the adolescents entered an assigned identification number, their gender, measured weight, and http://games.jmir.org/2017/1/e4/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e4 | p. 3 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Lyles et al measured height into the app to populate a baseline avatar using The open-ended reaction questions posed to the adolescents the computer located in the school library. They then designed included the following: “What do you think about being able the avatar to represent how they currently perceive their bodies to view avatars of yourself?” “Do you think the avatars are a to look (Perceived Avatar). The app allowed each adolescent good representation of yourself? Explain why or why not.” to make the body parts (ie, biceps, chest/bust, bust girth, waist, “How comfortable were you creating and viewing your avatars?” hips, thighs, and calves) bigger or smaller using the slider. Their “In the future, would you use these avatars to see how your completed Perceived Avatar then generated on the same screen body parts are changing? Explain why or why not.” “What do so they could design it to represent how they wanted their bodies you like about the avatars?” “What do you not like about the to look within realistic, healthy parameters (Target Avatar). avatars?” “What would you change about the avatars?” This feature was incorporated so the adolescents could work The Software Usability Survey is a short usability 7-point from their current perception rather than having to start over. Likert-type scale survey informed by the product and best On the same screen and to the right of the Perceived and Target practices in software engineering [23,24]. The items used to Avatars, the adolescents generated an Actual Avatar by entering identify usability for the MYA app included the following: (1) their body part measurements into the app. Figure 2 displays simplicity of navigation from one page to another, (2) ease of the three avatars after completion. The avatars could be rotated control to view the avatars and their body parts, (3) ease to 360 degrees. Figure 3 displays the three avatars from the side manipulate each body part and make it look more muscular, (4) view. need to look for help when working with the app, (5) The adolescents were then asked to answer questions about their understanding of what the three avatars are and how to view reactions to creating and viewing the avatars to assess and manipulate them, (6) ability to keep the user engaged and acceptability and to complete the Software Usability Survey. not quit, and (7) overall satisfaction with ease of using the app. Figure 1. User Interaction Screens for MYA App. http://games.jmir.org/2017/1/e4/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e4 | p. 4 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Lyles et al Figure 2. Example of three female avatars (front view). http://games.jmir.org/2017/1/e4/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e4 | p. 5 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Lyles et al Figure 3. Example of three female avatars (side view). results. The frequencies of the Software Usability Survey were Analysis analyzed and calculated into percentages. SPSS version 23 (IBM Corp) was used to conduct the descriptive statistics of the sample. The adolescents’ reactions Results to the MYA app were analyzed using content analysis [25]. The content analysis process included three phases: preparation, The sample characteristics for prototype testing are highlighted organizing, and reporting. The preparation phase included in Table 1, with the majority of the adolescents being female selecting the unit of analysis and classifying the data as a whole. (28/42, 67%), age 16 years (16/42, 38%), white (35/42, 83%), The organizing phase included developing a categorization non-Hispanic (36/42, 86%), in grade 10 (20/42, 48%), healthy matrix and coding the data per the categories. The reporting weight for females (23/28, 82%), and obese for males (7/14, phase included the results of the analyzing process and the 50%). http://games.jmir.org/2017/1/e4/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e4 | p. 6 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Lyles et al Table 1. Sample characteristics of the adolescents (N=42). Sample characteristics n (%) Age in years 15 12 (29) 16 16 (38) 17 11 (26) 18 3 (7) Gender Female 28 (67) Male 14 (33) Race White 35 (83) More than one race 2 (5) Asian 2 (5) Black American Indian 2 (5) American Indian 1 (2) Ethnicity Hispanic 6 (14) Non-Hispanic 36 (86) Grade level 9 2 (5) 10 20 (47) 11 16 (38) 12 4 (10) a b Female BMI -for-age category Underweight 0 (0) Healthy weight 23 (82) Overweight 4 (14) Obese 1 (4) Male BMI-for-age category Underweight 0 (0) Healthy weight 4 (29) Overweight 3 (21) Obese 7 (50) BMI: body mass index. Underweight: less than the 5th percentile; healthy weight: 5th percentile up to the 85th percentile; overweight: 85th to less than the 95th percentile; obese: equal to or greater than the 95th percentile. To determine acceptability, the adolescents’ responses to the Another adolescent compared the app to the mirror: “It's nice reaction questions (see Table 2) were analyzed using content to see what I actually look like opposed to looking in the mirror” analysis [25]. When asked what the adolescents thought about (obese, 16-year-old male). When asked if the avatars were a being able to view the avatars of themselves, 37 out of 42 good representation of themselves, 38 out 42 (90%) responded adolescents (88%) provided positive feedback. For example, “yes.” One adolescent explained the following: “Yes, because one adolescent commented about the actual and target it shows what my body looks like and what I think it looks like” representations: “I think it's good to see what you actually look (healthy-weight, 16-year-old female). like, then being able to see what your personal target is” (healthy-weight, 15-year-old female). http://games.jmir.org/2017/1/e4/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e4 | p. 7 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Lyles et al Another adolescent indicated how this was something new: There were several reasons the adolescents liked the avatars. “Yes, I've never done something like this. I think it is a good Almost half of the adolescents (18/42, 43%) indicated they liked representation” (overweight, 17-year-old male). One adolescent how the avatars were realistic, while almost a third (15/42, 29%) did not think it was a good representation: “Not really, because liked how they could move the avatars to see them from different I think more measurements should be taken” (obese, 18-year-old angles (see Table 2). Adolescents commented on what they male). liked: When asked if they were comfortable viewing the avatars, 41 That they are suited for the gender that you are and body type. out of 42 (98%) indicated they were comfortable. The research [Healthy-weight, 17-year-old female]. team categorized the level of comfort into three categories: very I like how they're realistic. [Healthy-weight, 15-year-old comfortable, somewhat comfortable, and comfortable based on female]. the explanations the adolescents provided (see Table 2). Almost half of the adolescents (18/42, 43%) were very comfortable How you can play around with them and see what you look like creating and viewing their avatars. When asked if the with certain numbers. [Obese, 16-year-old male]. adolescents would use the avatars in the future to see how their When asked what the adolescents did not like about the avatars, bodies changed, 41 out of 42 (98%) responded “yes” (see Table approximately one-third (13/42, 31%) indicated there were not 2). enough details or customization options with the avatars, while The adolescents explained the reasons they would use the app 13 out of 42 (31%) commented there was nothing to dislike. in the future. Most of the adolescents indicated they would use They do not show definition of muscle. (Obese, 16-year-old the app to track progress and goals (15/42, 36%) and to see the male). changes or differences in their body parts (11/42, 26%). The following are examples of why the adolescents would use the They don't have as much detail, with looks and more avatars in the future: measurements. (Healthy-weight, 16-year-old female). Yes, it would be very helpful because sometimes your The suggestions the adolescents made when asked about what progress can't be seen on the body but is realized in they would change about the avatars included the appearance measurements or on other body/perspectives. of the avatars, such as the clothing, hair, etc (16/18, 38%). One [Healthy-weight, 16-year-old female] participant stated, “I think I would change the avatars to look more like the person making the avatar so they can really feel Yes, because visual representation would help build confidence more into it and help view the different avatars of the person” in achieving the goal. [Overweight, 16-year-old female] (healthy-weight, 16-year-old female). Yes, I would like to see the pictures of my before and after results. [Obese, 16-year-old male] http://games.jmir.org/2017/1/e4/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e4 | p. 8 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Lyles et al Table 2. Responses to the questions identified by the adolescents (N=42). Questions and answers n (%) What do you think about being able to view avatars of yourself? Positive feedback 37 (88) Neutral feedback 4 (10) Negative feedback 1 (2) Do you think the avatars are a good representation of yourself? Yes 38 (90) No 4 (10) How comfortable were you creating and viewing your avatars? Very comfortable 18 (43) Comfortable 15 (36) Somewhat comfortable 8 (19) Neutral 1 (2) In the future, would you use these avatars to see how your body parts are changing? Yes 41 (98) No 1 (2) Reasons the teenagers would use the app in the future To track progress/goals 15 (36) To see the changes/differences in body parts 11 (26) Helpful to see visual representations 8 (19) Accurate representations 4 (10) It was fun/interesting 2 (5) Yes, with no explanation 1 (2) No 1 (2) What do you like about the avatars? Exact/actually look like/realistic 18 (43) Ability to move them/3D 12 (29) Ability to compare avatars 4 (10) It was fun/cool 3 (7) Others’ perspective 2 (5) App is anonymous 1 (2) Target Avatar 1 (2) No response 1 (2) What do you not like about the avatars? Needs more details/customization 13 (31) Nothing 13 (31) Appearance of avatars/clothes 8 (19) Difficulty seeing body changes made 3 (7) Not accurate 2 (5) Uncomfortable to see self 2 (5) Difficulty moving avatars 1 (2) What would you change about the avatars? Appearance of avatars/clothing 16 (38) http://games.jmir.org/2017/1/e4/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e4 | p. 9 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Lyles et al Questions and answers n (%) Nothing 9 (22) Add more details 8 (19) Able to see body changes better 7 (17) Able to move avatars better 1 (2) Add diet to the app 1 (2) The results of the 7-point Software Usability Survey indicated more voluntary physical activity than those who saw an that the adolescents found the app simple to navigate (rating of unchanging virtual self or no virtual representation [8]. 6 or 7; 40/42, 95%), found it easy to control the viewing of the Ridgers and colleagues [27] conducted a systematic review of avatars (rating of 6 or 7; 31/42, 74%), and found it easy to the effectiveness of youth wearable tracker devices to increase manipulate each body part and make it look muscular (rating physical activity levels among children and adolescents. They of 6 or 7; 24/42, 57%). The adolescents had the need to look found these devices have the potential to increase activity levels for help several times (rating of 1 or 2; 42/42, 100%), had a as the adolescents self-monitor their progress and set their goals. better understanding of what the three avatars are and how to The MYA app is similar in that it can be used to provide view them (rating of 6 or 7; 40/42, 95%), were engaged by the feedback of their body shape progress and monitor their goals. app (rating of 6 or 7; 41/42, 98%), and experienced overall Ridgers and colleagues [27] recommended that research be satisfaction with the ease of the app (rating of 6 or 7; 41/42, conducted to establish how youth engage with technology over 98%). longer periods of time. The adolescents indicated future use of the MYA app; however, more research is needed to examine Discussion their engagement with the app over time. Principal Findings Limitations The purpose of the prototype testing was to assess the This study is not without limitations. Due to the sample of acceptability and usability of an avatar-based, theoretically adolescents being relatively small and nondiverse, there was derived mobile app, the MYA app. Prior to the app being ready not enough power to detect group differences. Another limitation for prototype testing, two phases were completed that consisted is that the adolescents only tested the app at one session. To of classification and development. Once completed, the app understand the acceptability and usability over time, it is prototype was tested with a sample of male and female high necessary for the adolescents to interact with the app over a school adolescents. The adolescents created and viewed the longer period of time with several recurrent uses. The app did three avatars: Perceived Avatar, Target Avatar, and Actual not have this capability at the time of testing. Avatar. The adolescents had positive reactions to the avatar app Implications and being able to view avatars that represented them. Almost all of the adolescents indicated some level of comfort viewing This app allows teens to have visual representations of their the avatars and would use the app in the future to see how their bodies—perceived, target, and actual avatars— interactive bodies change over time. individualized programs, access to immediate feedback, and a method to monitor changes to their bodies as a result of positive Adolescents who evaluated a Web-based substance abuse changes in health behaviors. This visual representation is intervention also indicated high ratings for overall usability different from what adolescents currently have access to, such features such as ease of use and future use [26]. The ease of as digital scales and BMI calculators. Adolescents continue to manipulation of the body parts and making the avatars look be overweight and obese while physical activity decreases as more muscular resulted in a lower-percentage rating from the they get older. Adolescents who identify with and use these adolescents than the other usability survey items. The responses avatars who represent them can monitor and reach their body regarding the viewing of the avatars support this lower score. shape and health goals. While this app may be user driven, it The adolescents indicated that more details and customization also has the capability for health care providers, parents, and of the avatar bodies were needed and that it was difficult to see adolescents to have a visual communication tool to discuss the changes made to the avatars. These responses may explain adolescents’ body perceptions and goals. the lower ease of manipulation score. In further adaptations, there is a need for the avatar bodies to be more detailed with an Conclusions easier view of the body changes from one avatar to the next. Avatar-based mobile apps, such as the MYA app, provide immediate feedback and allow users to engage with images that The adolescents indicated they would use the avatars in the are personalized to represent their perceptions and actual body MYA app in the future to monitor their body shape changes. images. Adolescents appreciate the use of weight-related This is similar to participants who were assigned a virtual messages that are personalized for them [28]. The high school representation of their physical selves, an unchanging virtual students found the avatars of the MYA app easy to use, were representation, or no virtual representation. Those who witnessed comfortable viewing their avatars, and would use the avatars the virtual representation of their physical selves engaged in in the future to monitor how their body changes. Overall, they http://games.jmir.org/2017/1/e4/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e4 | p. 10 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Lyles et al were satisfied with the MYA app; however, they made among high school adolescents. There is a need to further adapt suggestions on how to make the app more appealing to high the MYA app and gather feedback from a larger number of high school males and females. This pilot study adds to the increasing school adolescents, including those from disadvantaged but limited research of using games to improve health outcomes backgrounds, while testing adherence to the app over time. Acknowledgments We would like to thank the graduate engineering students for help in developing the app prototype for testing. We are grateful for the local high school’s athletic trainer, graduate student trainers, and the high school students willing to be a part of these studies on body perceptions. Conflicts of Interest None declared. References 1. Bruner MW, Chad KE, Beattie-Flath JA, Humbert ML, Verrall TC, Vu L, et al. Examination of physical activity in adolescents over the school year. Pediatr Exerc Sci 2009 Nov;21(4):421-435. [Medline: 20128362] 2. Ingledew D, Sullivan G. Effects of body mass and body image on exercise motives in adolescence. Psychol Sport Exerc 2002;3(4):323-338. [doi: 10.1016/S1469-0292(01)00029-2] 3. Oenema A, Brug J. Exploring the occurrence and nature of comparison of one's own perceived dietary fat intake to that of self-selected others. Appetite 2003 Dec;41(3):259-264. [Medline: 14637324] 4. Bittner FH, Diamond J, Myers R, Gill JM. Perception, intention, and action in adolescent obesity. J Am Board Fam Med 2008;21(6):555-561 [FREE Full text] [doi: 10.3122/jabfm.2008.06.070184] [Medline: 18988723] 5. Maximova K, McGrath JJ, Barnett T, O'Loughlin J, Paradis G, Lambert M. Do you see what I see? Weight status misperception and exposure to obesity among children and adolescents. Int J Obes (Lond) 2008 Jun;32(6):1008-1015. [doi: 10.1038/ijo.2008.15] [Medline: 18317474] 6. Liechty JM. Body image distortion and three types of weight loss behaviors among nonoverweight girls in the United States. J Adolesc Health 2010 Aug;47(2):176-182. [doi: 10.1016/j.jadohealth.2010.01.004] [Medline: 20638010] 7. Davis R, Campbell R, Hildon Z, Hobbs L, Michie S. Theories of behaviour and behaviour change across the social and behavioural sciences: A scoping review. Health Psychol Rev 2015 Sep;9(3):323-344 [FREE Full text] [doi: 10.1080/17437199.2014.941722] [Medline: 25104107] 8. Bandura A. Social cognitive theory of mass communication. In: Bryant J, Zillmann D, editors. Media Effects: Advances in Theory and Research. 2nd edition. Mahwah, NJ: Lawrence Erlbaum Associates; 2008:121-154. 9. Skinner H, Biscope S, Poland B, Goldberg E. How adolescents use technology for health information: Implications for health professionals from focus group studies. J Med Internet Res 2003 Dec 18;5(4):e32 [FREE Full text] [doi: 10.2196/jmir.5.4.e32] [Medline: 14713660] 10. Randall W, Rickard N. Development and trial of a mobile experience sampling method (m-ESM) for personal music listening. Music Percept 2013;31(2):157. [doi: 10.1525/mp.2013.31.2.157] 11. Reid SC, Kauer SD, Khor AS, Hearps SJ, Sanci LA, Kennedy AD, et al. Using a mobile phone application in youth mental health - An evaluation study. Aust Fam Physician 2012 Sep;41(9):711-714. [Medline: 22962650] 12. Harrison V, Proudfoot J, Wee PP, Parker G, Pavlovic DH, Manicavasagar V. Mobile mental health: Review of the emerging field and proof of concept study. J Ment Health 2011 Dec;20(6):509-524. [doi: 10.3109/09638237.2011.608746] [Medline: 21988230] 13. Bandura A, Huston AC. Identification as a process of incidental learning. J Abnorm Soc Psychol 1961 Sep;63(2):311-318. [doi: 10.1037/h0040351] [Medline: 13864604] 14. Fisher JD, Fisher WA, Bryan AD, Misovich SJ. Information-motivation-behavioral skills model-based HIV risk behavior change intervention for inner-city high school youth. Health Psychol 2002 Mar;21(2):177-186. [Medline: 11950108] 15. Tukatech. MyBodee URL: http://www.tukatech.com/styku [accessed 2017-02-24] [WebCite Cache ID 6oWP313RW] 16. Stahl T, Völter M, Bettin J, Haase A, Helsen S. Model-Driven Software Development: Technology, Engineering, Management. Chichester, UK: John Wiley & Sons; 2006. 17. McCurdie T, Taneva S, Casselman M, Yeung M, McDaniel C, Ho W, et al. mHealth consumer apps: The case for user-centered design. Biomed Instrum Technol 2012;Suppl:49-56. [doi: 10.2345/0899-8205-46.s2.49] [Medline: 23039777] 18. Bastioni M, Re S, Misra S. Ideas and methods for modeling 3D human figures: The principal algorithms used by MakeHuman and their implementation in a new approach to parametric modeling. In: Proceedings of the 1st Bangalore Annual Compute Conference. New York, NY: ACM; 2008 Presented at: The 1st Bangalore Annual Compute Conference; January 18-20, 2008; Bangalore, India. 19. Marrin C. Khronos. 2011 Feb 10. WebGL Specification URL: https://www.khronos.org/registry/webgl/specs/1.0.0/ [accessed 2017-02-14] [WebCite Cache ID 6oHfuZojg] http://games.jmir.org/2017/1/e4/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e4 | p. 11 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Lyles et al 20. Blender. URL: https://www.blender.org/ [accessed 2017-02-10] [WebCite Cache ID 6oAqLWOOP] 21. Crockford D. IETF Tools. 2006 Jul. The application/json media type for JavaScript Object Notation (json) URL: https:/ /tools.ietf.org/html/rfc4627 [accessed 2017-02-14] [WebCite Cache ID 6oHg2zld4] 22. Cabello R. GitHub. 2010. JavaScript 3D library URL: https://github.com/mrdoob/three.js/ [accessed 2017-02-14] [WebCite Cache ID 6oHgKBEWz] 23. Brown W, Yen P, Rojas M, Schnall R. Assessment of the Health IT Usability Evaluation Model (Health-ITUEM) for evaluating mobile health (mHealth) technology. J Biomed Inform 2013 Dec;46(6):1080-1087 [FREE Full text] [doi: 10.1016/j.jbi.2013.08.001] [Medline: 23973872] 24. Brooke J. SUS: A 'quick and dirty' usability scale. In: Jordan PW, Thomas B, Weerdmeester BA, McClelland IL, editors. Usability Evaluation in Industry. London, UK: Taylor & Francis; 1996:189-194. 25. Elo S, Kyngäs H. The qualitative content analysis process. J Adv Nurs 2008 Apr;62(1):107-115. [doi: 10.1111/j.1365-2648.2007.04569.x] [Medline: 18352969] 26. Raghupathy S, Forth AL. The HAWK2 program: A computer-based drug prevention intervention for Native American youth. Am J Drug Alcohol Abuse 2012 Sep;38(5):461-467. [doi: 10.3109/00952990.2012.694531] [Medline: 22823599] 27. Ridgers ND, McNarry MA, Mackintosh KA. Feasibility and effectiveness of using wearable activity trackers in youth: A systematic review. JMIR Mhealth Uhealth 2016 Nov 23;4(4):e129 [FREE Full text] [doi: 10.2196/mhealth.6540] [Medline: 27881359] 28. Woolford SJ, Clark SJ, Strecher VJ, Resnicow K. Tailored mobile phone text messages as an adjunct to obesity treatment for adolescents. J Telemed Telecare 2010;16(8):458-461 [FREE Full text] [doi: 10.1258/jtt.2010.100207] [Medline: 20959393] Abbreviations BMI: body mass index MYA: Monitor Your Avatar Edited by G Eysenbach; submitted 14.07.16; peer-reviewed by J Chaplin, R Schnall, PY Yen, W Brown, M Nitsch; comments to author 20.11.16; revised version received 06.01.17; accepted 03.02.17; published 02.03.17 Please cite as: Lyles AA, Amresh A, Huberty J, Todd M, Lee RE JMIR Serious Games 2017;5(1):e4 URL: http://games.jmir.org/2017/1/e4/ doi: 10.2196/games.6354 PMID: 28254737 ©Annmarie A Lyles, Ashish Amresh, Jennifer Huberty, Michael Todd, Rebecca E Lee. Originally published in JMIR Serious Games (http://games.jmir.org), 02.03.2017. 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 Serious Games, is properly cited. The complete bibliographic information, a link to the original publication on http://games.jmir.org, as well as this copyright and license information must be included. http://games.jmir.org/2017/1/e4/ JMIR Serious Games 2017 | vol. 5 | iss. 1 | e4 | p. 12 (page number not for citation purposes) XSL FO RenderX

Journal

JMIR Serious Games – JMIR Publications

Published: Mar 2, 2017

Keywords: adolescents; avatars; eHealth; mHealth; perceptions; Web-based; usability testing

There are no references for this article.