Abstract
Objective:
mHealth interventions often favour individual-level effects. This is particularly problematic in contexts where social support and shifts in social norms are critical to sustained behaviour change. Mobile digital games represent a promising health education strategy for youth, including in low-resource settings. We sought to better understand the interpersonal and social interactions that can be elicited by digital games for health.
Design:
We piloted Tumaini, a smartphone game rooted in interactive narrative designed to prevent HIV among young Africans (aged 11–14), in a randomised controlled feasibility study and analysed reports of the household dynamics surrounding gameplay. Following a 16-day intervention period, phone gameplay log files were downloaded, and intervention arm participants (n = 30) completed a gameplay experience survey; eight focus group discussions were held, four with intervention arm participants (n = 27) and four with their parents (n = 22).
Setting:
This study took place in Kisumu, Kenya, in Spring 2017.
Method:
Descriptive statistics were computed from survey responses and log files. Focus group transcripts were labelled, analysed thematically and compared demographically using MAXQDA software.
Results:
Data from log files, surveys and focus groups indicate that the game generated considerable interaction and dialogue with parents, siblings and friends and served as a catalyst for children to act as advocates for healthful decisions about sex, both within the family and beyond. The game showed a high level of acceptability with parents.
Conclusion:
Serious digital games using a smartphone platform can generate considerable household interaction. Games can model and facilitate these exchanges, maximising multi-level effects. An additional app for parents could reinforce these effects.
Introduction
Young Africans continue to account for a high proportion of new HIV infections (United Nations Programme on HIV/AIDS (UNAIDS), 2018); in Kenya, for example, over half of all new infections occur in young people aged 15–24, with young women at particular risk (National AIDS Control Center, 2016).
Digital games represent a promising HIV prevention strategy, especially for youth (Hightow-Weidman et al., 2017). Increasingly accessible smartphone technologies in Africa could make it possible to engage young people – at scale and at low cost – in culturally adapted prevention interventions that require little personnel to implement with consistent quality and have high entertainment and motivational appeal.
Taking increasing advantage of the affordances of the mobile platform, most games for health tend to prioritise individual-level behaviour change (Shegog, 2010). While the broader gaming literature has devoted some attention to the importance of social context in shaping player experience (De Kort and Ijsselsteijn, 2008; Elson et al., 2014a, 2014b), we are aware of no published studies addressing the contextual dynamics of gameplay for games for health. In settings where personhood may be construed in less individualistic terms, the household dynamics of gameplay may be particularly important and may provide an opportunity to extend intervention effects beyond the individual to multiple levels. This may be particularly important in contexts and for behaviours where social support and shifts in social norms are critical to sustained change.
Tumaini
Tumaini is a smartphone-based game designed to increase age and condom use at first sex among young Africans. It seeks to achieve these goals by increasing knowledge about sexual health and HIV; building risk-avoidance and risk-reduction skills and related self-efficacy; challenging HIV stigma and harmful gender norms and attitudes; fostering future orientation, goal setting and planning and promoting dialogue with adult mentors. Tumaini is grounded in theory on narrative and narrative-based applied communication and social behavioural theory and existing evidence-based HIV prevention interventions. Its development drew on extensive research on HIV-themed narratives written by young Africans (Beres et al., 2013; Singleton et al., 2018a, 2018b; Winskell et al., 2011a, 2011b, 2011c, 2018a). The game uses interactive narrative to promote observational learning, cognitive and behavioural rehearsal, problem-solving and immersion. Tumaini was developed in collaboration with a US commercial game developer, Realtime Associates, and specialists in youth HIV prevention in Kenya and the USA. The game has shown promising results on behavioural mediators of sexual debut and condom use in a randomised controlled feasibility study in Kisumu, Western Kenya, in Spring 2017 (Winskell et al., 2018b).
The game consists of three intersecting components: (1) an interactive role-playing narrative where the player makes decisions for six diverse adolescent characters; (2) mini-games that train and reinforce skills and knowledge presented during the narrative and (3) ‘My Story’, where the player reflects on his or her goals and strengths and relates the game narrative to his or her own life. The game includes characters which players can identify with and situations they are likely to face as they grow up. It is designed to help players set goals for the future and develop skills to reach these goals and protect their dreams. It encourages them to talk to their parents and other trusted adults about their problems and to access health services.
We present feedback provided (1) by participants aged 11–14 and their parents or guardians (hereafter referred to as ‘parents’) in focus group discussions (FGDs) conducted 1–2 weeks after the 16-day feasibility study of Tumaini and (2) by participants aged 11–14 in a survey completed immediately post-intervention. We support these data with analysis of gameplay log files. Our purpose is to better understand the household dynamics of Tumaini gameplay in order to inform next steps in this study, the development of similar studies and greater understanding of the potential for mHealth interventions to extend beyond the individual level.
Methods
We provided the 30 young people randomly assigned to the intervention arm of our study with a low-cost Android phone on which Tumaini was preloaded and headphones to maximise potential for immersion. Each young person provided written assent; parents provided written consent for their children’s and their own participation. They were instructed to play the game for at least 1 hour per day during the 16-day study; parents were asked to remind them to do this and to contact study staff with any problems. Parents were instructed that Tumaini was the children’s game to play as they wanted, their ‘space to learn’. However, they were told that the child should feel free to ask for help and talk to whomever they wanted about the game and that parents should feel comfortable talking with their child about the game or what he or she was learning. If parents did not know the answer to one of the child’s questions, they were instructed to contact the study staff.
The game was designed to accommodate additional player profiles without contaminating study participant data. In other words, siblings or friends could play the game by creating their own game profile. These profiles were recorded in the game log files. At the end of the intervention period, participants identified the profiles as their own or as belonging to their mother/father, younger/older sister/brother, male/female friend or other individual. All phone functions except the game and the alarm clock (for reminders) were blocked.
All intervention arm participants (n = 30) completed a gameplay experience survey immediately following the intervention period. The survey asked with whom they had shared and discussed the game. Intervention arm participants (n = 27) and parents (n = 22) also took part in FGDs (n = 8) shortly after the intervention ended to provide additional data on the game experience. The four adolescent focus groups (labelled A1–A4 below; see Table 1) were stratified by age (11–12 and 13–14) and gender of the study child; the four parent focus groups (labelled P1–P4 below) were stratified by age of the study child alone. The majority of parents were women.
Composition of FGDs.
FGD: focus group discussion.
Table 2 provides demographics of adolescent participants. We believe these to be broadly reflective of the source population in urban and peri-urban Kisumu.
Baseline demographics of FGD participants (prior to allocation to study arm).
FGD: focus group discussion.
‘In your home who has a smart phone? A smart phone is a phone that has a touch screen and access to the internet. Choose all people who have smartphones’. It is not impossible, though we believe unlikely, that one phone was considered to be owned by more than one person.
The discussion guide for adolescent participants included questions about what they would tell their friends about the game and conversations they had with others about the game. The discussion guide for parents included questions about their experience with their child playing the game, how their child played and their interactions related to the game.
The focus groups were conducted in a mixture of English, Dholuo and Kiswahili by moderators fluent in all three languages. Parents and young people were invited to use any combination of languages they felt comfortable with. Focus group transcripts were translated into English and uploaded to the MAXQDA software, where they were labelled with inductive (e.g. ‘family dynamics’) and deductive codes (e.g. ‘how children played’). The data were analysed thematically and compared across demographics.
Results
Log file analyses
Log file data showed that multiple user profiles were created on study phones, and most intervention arm participants shared their phones with others. All but one phone showed profiles under multiple names, with an average of 8.2 profiles per phone. Only four players identified all the profiles as their own, with the rest of the phones having between 1 and 17 additional user profiles, a total of 167 across all phones. The players identified the creators of these profiles as siblings, friends, parents or others. Male peers accounted for the largest proportion of additional profiles (43%, n = 72): 25 were identified as belonging to older brothers, 17 to younger brothers and 30 to male friends. Female peers accounted for 37% of additional profiles (n = 61), including younger sisters (n = 23), female friends (n = 20) and older sisters (n = 18). Of note, five profiles were said to have been created by mothers, aunts or uncles. On average, male study participants’ phones showed more additional profiles (mean = 6.6, SD = 4.8) than those of female participants’ (mean = 4.4, SD = 3.1).
Survey data
Adolescents’ survey responses indicated that most interacted with family members and peers, sharing the game or discussing it with others, with only eight (27%) saying they had played Tumaini alone (Table 3). The remaining 17 (57%) mixed the two, with 7 playing ‘mostly alone’ (23%) and 10 ‘mostly with others’ (33%). Among those who played socially (73%, n = 22), the most frequent game partners were siblings (brother, n = 9; sister, n = 11), followed by cousins and female friends (each, n = 7). Other specified partners were male friends and parents. In total, 24 players (80%) reported discussing the game with others, mostly with their mothers (n = 14), sisters (n = 14) and brothers (n = 12). Those who spoke to others about the game most commonly reported talking to 1–2 people (n = 9).
Frequency and percentage of participant responses to questions on game-related interactions.
FGDs
Older children, who had stronger English and analytical skills, tended to be more articulate and their focus groups richer. While parents were less informed about the game content and structure, they provided detailed insights from an observer’s perspective into the context of gameplay, particularly family dynamics.
Motivation to play
Parents reported that their children were happy and excited to play: ‘my daughter really loved the game. Any time she was through with house chores and her school work, she would ask me for the phone to play the game. She could play for an hour then extend that time’ (P4). They also enjoyed sharing the game with others. One parent reported, ‘His friends and classmates also liked the game so much that my house was full most of the time’ (P2). Children’s enthusiasm for the game was evident in their own focus groups; for example, when the older boys were asked what they would change about the game, they proposed, in chorus, a dramatic increase in the number of chapters. Several parents spontaneously stated that their children were unhappy when the phone had to be returned. One mother said of her husband, ‘he wished we could leave the phone with the child so that she could continue playing with it for her to change even more than she already did’ (P3).
The game appeared to stand up well to competition from other activities. One parent commented that her daughter is accustomed to playing games on her older sibling’s phone, ‘But she was so preoccupied with this game that she did not even think of the sister’s phone’ (P4). Parents appreciated the fact that the game kept their children occupied so they would not leave the house or would return home to play. One older boy stated, ‘When I was watching TV and I found it boring, I would just go to my bed and look at the game’ (A4).
Emotional context of gameplay
Parents described a mood of excitement, happiness and enjoyment among their children. Several parents described how the game made the whole family happy and helped them. Parents in two FGDs spoke of the ‘joy’ it brought to the house as a whole, and a man spoke of his family’s enjoyment, ‘it was a good experience because I remember all of us enjoyed ourselves, me, my wife and my grandchild’ (P4).
One parent talked of her child as being both excited, but also serious as he was playing: ‘watching him I could tell that there were very serious issues’ (P4). Only one parent reported negative emotions on the part of her child, namely anxiety when she got stuck on one mini-game (P1). Some parents indicated that their children wanted quiet when they were playing: ‘He told me “mother, when learning I want to be in a quiet place not where there are loud children, that is when I understand and I can also be able to teach some other children”’ (P3).
Interaction with siblings and peers
Parents provided additional insight into these data, confirming that some played the game with siblings, cousins or friends, each taking turns; some played while friends or siblings watched and others played alone consistently or oscillated between playing alone or with a group depending on circumstances. Parents’ comments suggest this was contingent on personality, age of siblings or who was in the house at the time. Parents also reported that their children played almost exclusively at home, with only one describing an occasion when the child took the phone out of the house to attend a wedding.
Some parents set parameters for gameplay to ensure that their child balanced it with chores and school work. However, one parent reported – to laughter from other group members – that her daughter and her sisters enjoyed playing so much that they neglected their homework, despite their assurances to the contrary. Another parent had six children aged between 12 and 14 in the house, who all played together. She discussed how her children organised themselves to ensure that each child was able to finish the game: ‘they would sit together and agree that the person who had not finished would be the one to start until everyone got an equal chance’ (P2).
One parent reported that the game (and perception of the study as ‘research’) caused some tension when her son’s friends visited: ‘So they always wanted to be given a chance to play too and he had to explain to them that he had to complete his game first. To the friends it seemed like he was denying them the game’ (P3). Tension was also reported with siblings when they wanted to play too, although the option to create other profiles and play separately seems to have alleviated some of this tension (P1). There were also examples of positive interaction with younger siblings. An older boy said his sister learned how to prepare for menstruation from the game and he discussed it with her, while a younger girl reported a conversation with her sister about menstruation and learned that their mother had helped her sister when she had her first period. Another study child translated the game into Kiswahili for a younger sibling.
Several parents and children reported that older siblings were playing and enjoying the game too. As one parent reported, ‘Later when I asked him what he thought of this game since he is a 17-year-old and the game was designed for 11-14 year-olds. He told me learning never ends and he was getting lessons from the game’ (P4). As the game was tested in a Beta (i.e. not final) version, some bugs remained and some of the mini-games were more challenging than intended. In several cases, this provided a pretext for conversations with older siblings. Several participants asked older siblings (sisters and brothers) for help with mini-games, addressing the steps to correct condom use or the male and female reproductive systems. Parents welcomed and even actively encouraged this involvement of older siblings: ‘I had no fear because the phone made the atmosphere […] conducive, I could see my elder son also create his own profile and play the game, we had no worries because they were helping one another’ (P2).
Effects on communication with siblings and peers
Participants reported having shared lessons from the game with peers and older and younger siblings, either by sharing the game experience directly with them or by discussing what they had learned. One younger girl explained, ‘When I was playing Tumaini, I would go and call my other friends to come and see how I was playing Tumaini so that they could also learn from the game’ (A1).
The lessons they shared included encouraging others not to rush into having a boyfriend or girlfriend and persuading others to wait to have sex. They also passed on the game’s main messages about avoiding bad influences, pursuing goals through education and avoiding sex without condoms in order to protect themselves from sexually transmitted infections (STIs), HIV and pregnancy because ‘if they did it they would not reach their goals’ (A3). Participants felt empowered to share practical skills as well as overall lessons from the game. They reported either having taught others or intending to teach others how to use condoms (A4) and how to resist pressure to have sex (A1).
Beyond sharing newly acquired information and skills, one older female participant advocated for support for those living with HIV, gathering her friends, ‘as a group so that nobody could interfere with us. I told them that they should be kind to those who have HIV’ (A3).
Interaction with parents
Interactions with parents differed based on parents’ availability, interpretation of instructions from the study team and their own and their children’s level of comfort with the subject matter. Some parents actively engaged their children in conversation about the game. In other cases, it was the children who initiated conversations, questioning their parents or choosing to sit next to them while they played, for example, Mostly when she wanted to play the game, she would come and sit close to me. So I would also hear the questions the game asked and how she responded. Once she answered the questions she would tell me that all these things I advised her about should really be done after her education, and even the game said so. (P4)
Some parents reported, in contrast, that their child reached out only when he or she was having challenges, got stuck or needed to request help from the study staff. Some children only talked with their parents about the game rewards or no interaction was initiated by either side; some talked with older siblings or teachers rather than with parents or were redirected to older siblings by their parents. Some parents explained that they worked late and were busy and did not have time to talk with their child, while others interpreted the study staff’s instructions to give the child space to play the game as precluding their involvement and were reluctant to ‘interrupt or maybe interfere’ (P1). In one case, it was the child who interpreted instructions in restrictive terms, insisting on playing the game alone.
Some parents described embarrassment on their part or that of their children. When their children approached them with questions about condoms or the male and female reproductive systems, some referred their child to older siblings (P2), while others helped despite their embarrassment. One parent described her son’s internal conflict about getting help: You could see him make faces like he wanted to ask and I was not ready to answer … But they used to play the game and at times you could see him stop, meaning that he wanted to ask a question but he was not ready because such like discussions have not been going on in the house. (P2)
Some parents described how their children were selective about which game content they wanted to share (avoiding condoms, for example) and with whom. One parent reported that her three children would talk together about the decisions they had made in the game, but fall quiet as soon as she entered the room (P1), while another attributed her lack of challenges to the fact that she had multiple children in the house between the ages of 11 and 14.
Effects on parent–child communication
Across three of their four FGDs, parents described how the game supported their efforts to talk with their child about HIV and related themes: We were happy because in the family it is very difficult to sit down with a child and teach her about HIV. So when we got this gadget [referring to the phone] I could say it has made this discussion easy. (P4)
The game helped children ask their parents questions they did not previously feel comfortable asking. One mother said of her son, ‘the game has made us freer. And the discussions about drugs and sex, he has those discussions with me’ (P3). One mother’s conversation with her daughter about the risks of transactional sex was prompted by the mother asking her daughter about game content she had overheard (P3). In addition to increasing discussions relating to sexual risks, parents shared instances where the game had started or enhanced conversations about future planning and goal setting. Several described conversations the game provoked not just with the study child but with other children in the family, fostering a freer and more open dialogue.
In half of their FGDs, parents noted that the game had taught their child that he or she should be communicating with them. One mother described her child telling her that she was unable to judge what behaviours were appropriate because ‘there is no dialogue between us’: The game made her open up and we discussed where she has difficulties and the things that she wanted me to do for her. […] there were questions asking if she has a problem who would she share with? She realised that she should share with the parent. (P2)
Another mother attributed this realisation to the fact that ‘they see that we have passed through all these stages’ and that, as women, they share these experiences: ‘Right now we talk about everything because she knows I am a female like her. So we are very free’. (P3)
For some families, the game supported, reinforced or validated existing conversations. One mother had participated in the Families Matter! Program, a local group-based intervention to strengthen parents’ skills to talk with their children about sexuality, and attributed her child’s level of comfort to her prior conversations.
Parents also reported that their messaging and that of the game were mutually reinforcing. When asked a question, some parents referred their child back to the game, instructing him or her to listen carefully to the message given there (P2). One mother explained that ‘[the game] helped me so much because I have always talked to my child about HIV, and after playing she told me “mother so you have been telling me the truth”’ (P4).
Parents and children also described specific conversations that arose as a result of the game. In several cases, the conversation emerged out of a scenario from the game: after overhearing the game audio, the parent reinforced its messages. A younger girl reported, She heard what Daniel was saying when he was being told by his close friends to smoke and then he said that I would rather have no friend than those friends then my mother told me that are you listening to what he is saying then I said yes, then she told me that what he has said has a lot of weight that I can use when friends are having bad influence on me or forcing me to do things that i don’t like. (A1)
One mother described intentionally ‘eavesdropping’ on the game, which provided an entry point for conversation with her previously ‘stubborn’ daughter. The mother described several conversations about HIV, transactional sex, menstruation and her future career goals, concluding, ‘You know at this stage men may also be interested in this young girl, and if such a thing happens right now I know she would tell me’ (P3). Similar conversations were reported between daughters and fathers. A younger female participant described a conversation about condoms: One day I was sitting in our house my father was at home and he saw me when I was playing the part where (fumbling) arranging the condoms then he said that he wanted to see, I gave him then he said that this thing is good because it is teaching you when you are ready you can use a condom. (A1)
Another father recalled his 13- to 14-year-old daughter asking, ‘Father so it is true that when out there if a boy calls you to go to where he is, you can refuse?’ (P4).
Parents reported that participants had sought out not only themselves and older siblings but also teachers or media sources to discuss or validate the information presented in the game. Another participant emphasised the game’s value in its ability to ‘[teach] the child at a level that we parents can’t get to’ (P4).
Discussion
Parents’ and adolescents’ accounts of the enjoyment they derived from playing, their desire to continue interacting with the game and the game’s own log file findings suggest a high level of motivational appeal. We had initially feared that the game might operate as an individual-level behavioural intervention which did not attract interactions within families and peer groups. These fears proved to be unfounded. Although children played the game in diverse ways, alone and in groups, it was clear that for most children it generated considerable interaction and dialogue with parents, siblings and friends.
The game was perceived as a fun learning activity, and children were eager to share what they learned with others. It served as a catalyst for children to act as advocates for healthful decisions about sex, including promoting condom use and delayed sexual debut, both within the family and beyond. This indicates potential for the game, even if delivered at the individual level, to have multi-level effects.
Both younger and older siblings were engaged by the game and enjoyed playing and learning, indicating that, while the game was intended for 11- to 14-year-olds, its appeal and potential relevance is not limited to that age range. The trial was scheduled earlier than originally expected to avoid the run-up to the 2017 Kenyan general election when there were fears of potential unrest. As a result, the game used in the trial was a Beta version which had not been as exhaustively tested as planned and some mini-games proved more challenging than intended. This had the unintended positive effect of encouraging participants to reach out to parents and older siblings for help, leading to interaction and dialogue. Coincidentally, this focused on more sensitive areas: steps to condom use and male and female reproductive systems. While the benefits of this added incentive to reach out to others are notable, it is important to ensure that the difficulty of the game components does not undermine players’ feeling of competence or interrupt the flow of game play, which could act as a deterrent to play (Crutzen et al., 2016).
The game also showed a high level of acceptability with parents, who noted not only the effects on their children’s behaviours but also, in many cases, credited the game with creating a ‘freer’ atmosphere and leading to more conversations, including sensitive topics, initiated by both parents themselves and their children. Some parents felt restricted by the study instructions, interpreting them to mean that they should not engage with their child about the game at all, even to help when a child was stuck. In future studies, these instructions will need to be clarified so they do not act as a barrier to family-level effects of the game, not least in light of the importance of pre-risk parent-child communication for adolescents’ decision-making about sex and sexual risk-taking (Miller et al., 1998; Widman et al., 2016).
Shegog (2010) notes that ‘teaching a child health behaviour skills through a serious health game may be of little consequence if the enablers of behaviour (e.g. parents) do not receive some form of intervention to support and reinforce the behaviour’ (p. 222). In this study, some of the parents who did engage with their children expressed discomfort in doing so or chose instead to refer their children to older siblings. In the light of this, it may be useful to provide a complementary intervention to parents to support communication with their children.
Interventions targeting parent–child communication about sex have demonstrated positive effects both on communication and on young people’s decision-making with relation to sex (Miller et al., 2009; Sutton et al., 2014). Some evidence-based HIV prevention interventions for youth, such as Let Us Protect Our Future, include homework assignments designed to promote and ease parent–child communication (Jemmott et al., 2014). While many game-based interventions are designed as individual-level interventions, D’Cruz et al. (2015) investigated the acceptability of an intergenerational gaming (IGG) intervention to promote parent–child communication about sex. A Tumaini-style app could provide parents with interactive opportunities for cognitive and behavioural rehearsal before they are called on to discuss sex-related issues with their own children. In a follow-up FGD, a year after the initial intervention, a group of parents expressed enthusiastic support for this idea, requesting an app that included information, advice and practice exercises to help them talk to their children.
Limitations
This study is not without limitations. The level of interaction between family members and peers may at least, in part, stem from the novelty of the intervention and platform, which may abate during a longer exposure time or be less prominent with young people with greater access to smartphones. The context of a randomised controlled study may have influenced household dynamics of gameplay, with some parents, for example, interpreting the study instructions as precluding interaction with their children about the game.
Conclusion
Study findings suggest that mobile games, even on small-screen smartphones, can generate lively household engagement, interaction and discussion. The study illustrates the need for further research using a variety of methods, including qualitative and ethnographic methods, to better understand the context of gameplay of games for health in diverse settings. Insights provided by such studies could inform the design and potential of other similar interventions appropriately adapted to other contexts.
Footnotes
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship and/or publication of this article: Research reported in this publication was supported by the National Institute of Mental Health of the US National Institutes of Health under Award Number 5R34MH106368 (PI: KW). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. This research was also supported by the Emory Center for AIDS Research (P30 AI050409) and the Andrew W. Mellon Foundation. The sponsors played no role in review and approval of the manuscript for publication.
