Keywords

1 Introduction

As the general population in Western countries shifts towards an aging demographic, older adults are trying to lead healthy lifestyles [1] while maintaining physical and mental wellness. Maintaining an active lifestyle by engaging in activity equivalent to walking more than three miles per day, in addition to independently carrying out normal daily activities contributes to physical wellness [2]. Additionally, technology assisted physical activities contribute to overcoming cognitive decline thereby facilitating mental wellness [3,4,5]. Information technology has evolved over time to enable the creation of fitness apps [6] and wearable technologies to help adults participate in physical activity (PA) with health benefits [7, 8]. However, cognitive and physical challenges faced by older adults poses barriers and limitations for them to participate in PA [9, 10]. These barriers and limitations can be mitigated through the use of persuasive technologies that encourage physical activity [11]. Research has indicated the advantages of games [12], gamification [13] and persuasive technology [14, 15] to motivate older adults for PA. Therefore, it is important to investigate the reasons for lack of engagement and motivation with PA, and the potential increase in PA motivation through the use of motivational affordances. Motivational affordances are game elements which help foster intrinsic and extrinsic motives [6, 16] especially in mundane activities like PA. This idea follows from gamification, which is the application of strategies from game design (e.g., actions, challenges, and achievements) to non-game activities to help make people’s actions more engaging [17]. Persuasive technologies like gamification have been instrumental in effecting behaviour change in users in diverse fields ranging from education, marketing, crowdsourcing, management information systems to health and wellness domains, to mention a few.

To investigate the influence of gamification elements on older adults’ motivation and engagement in PA, we designed a synchronous, three condition, long-term study over an eight-week duration in which participants were randomized into three groups (gamified, non-gamified and control) (cf. Sect. 3.2). Quantitative analysis indicated significant differences between the groups with higher engagement for gamified and non-gamified groups. These results suggest the relevance of gamification elements as behaviour change elements to motivate and engage older adults in PA.

2 Theory and Background

2.1 Gamification

Persuasive technologies using gamification as a strategy have been used to foster motivation indicated emergent themes such as feedback and monitoring, reward and threat, and goals and planning [13]. While many definitions of gamification exist, its role as a behaviour change agent is more inspiring.

Effective gamification is influencing human behaviour through engaging experiences, using game design principles in non-game applications and services [18]. Gamified app-based interventions to improve diet, physical activity and exercise indicated significant health improvements in comparison to stand-alone apps [19]. Tailoring these apps based on the personality traits of users could increase their efficacy [20]. While incorporating challenges and accountability facilitated usage of persuasive strategies within a gamified design, challenges with privacy (sharing personal information) and reduced self-confidence hindered effective deployment of these gamified apps [21]. A systematic review of empirical studies on gamification in the health and wellbeing domains indicated a positive impact towards health behaviours [22]. While this study indicated that targeted health and well-being domains were: physical health (exercise and fitness), healthcare utilization (dispensing, medication usage and monitoring glucose levels, patient empowerment), and mental health (stress and anxiety reduction, it also indicated that several studies reviewed reported mixed or neutral effects of gamification. Gamified web interventions for rheumatoid arthritis patients indicated increased empowerment, app usage and PA than the control group [23]. Systematic reviews on gamification literature have called for comparing gamified interventions with non-gamified interventions [24,25,26]. This inspired the comparison of the differences between gamified and non-gamified PA interventions [27].

Many commercial systems have combined gamification with PA [6] as a behaviour change strategy to help individuals achieve health and fitness goals. As part of this, fitness trackers motivate workout activities by providing feedback [26]. Gamification or gameful design as a form of persuasive technology offers many hedonic (pleasure, enjoyment, comfort, absence of distress) and eudaimonic (growth, meaning, authenticity, excellence) [28, 29] pathways to implement persuasive strategies to motivate older adults to initiate, maintain, and adhere to healthy behaviour.

2.2 Gamification and Older Adults’ Physical Activity

While research on the general intrinsic and extrinsic motivation of older adults [30] affirmed interest levels towards engaging in PA, more research on player personality characteristics from gamer personality is needed to understand attitudes and motivation of players (users) [31]. Existing research explores age-related needs of older adults [32], barriers and motivations towards PA, and activity levels [30, 33].

Persuasive technologies in the form of exergames (games involving physical exercise), web-based interventions and fitness applications involving gameplay, provide a diverse spectrum of available technologies to initiate and maintain PA among older adults. Seniors preferred games that simulated true-to-life scenarios, provided cognitive training and digital games that helped to improve their reflexes [34]. Systematic reviews on exergames and older adults’ PA indicated clustering of studies into rehabilitation, training and wellness themes with focus on theories such as self-determination theory and social cognitive theory [35, 36]. Playing video games also contributes to improving health outcomes in areas of psychological therapy and physical therapy and improving balance among older adults [37, 38]. However, designers must consider age-related functional limitations such as declining vision, hearing, hand-to-eye coordination and cognition issues of older adults [39]. Furthermore, in-person and electronically mediated interventions through persuasive games [40] and inter-personal communications [41] were shown to be effective for influencing and motivating health behaviour change [42]. Additionally, home-based health monitoring practices [43], as researched in Mediframe [44] and in-home rehabilitation technologies [45] discuss challenges in the implementation of these technologies due to age-related infirmities and diseases. Gamification applications like Superbetter (www.superbetter.com) focused on negotiating life’s challenges such as depression, anxiety, pain, finding a job, improving a skill and strengthening a relationship to mention a few, into a gamelike experiential play supported by motivational affordances. Habitica (www.habitica.com) and coach.me (www.coach.me) are also examples of gameful technology that provide the impetus to role playing life in the form of setting up goals, with in-game rewards, punishments and social facilitation.

Many older adults are growing accustomed to evolving technologies and omnipresent video games [23, 46, 47]. New input technologies for these games are becoming more intuitive for an aging demographic of gamers (e.g., movement interfaces such as the Microsoft Kinect [48], the Nintendo Wii controllers [49], large screens present on mobile tablet computers and smartphones). Adaptive engagement, defined as tailoring of older adults’ engagement through customization and personalization of motivational affordances is important to address age-related changes, current health conditions and barriers to do PA [27, 50,51,52].

Research has explored the design of motion-based games with the goal of facilitating occupational therapy and rehabilitation among older adults. Recent examples include motion-based games for stroke rehabilitation where the patient recovered significant motor abilities over a six-week period [53], therapy for older adults with Parkinson’s disease [54], and balance training to reduce the risk of falls among older adults [55, 56]. Motivating older adults in PA with focused gameplay over long periods was seen as an advantage of exergaming [57].

Different approaches have been taken with regard to the promotion of exercise and PA. Non-commercial games like UbiFit Garden provided feedback in the form of flowers based on daily physical activities [11] and Flowie provided feedback on increased number of daily steps taken by the participants [58], targeted the older adult population, and aimed to encourage PA through persuasive technology intervention [59]. While this helps to promote PA, mobile apps can also promote activity using smartphones, focusing on individually tailored feedback and advice [60]. While playful persuasive solutions [61], embodied gaming [62] and augmented gaming [63] can facilitate fun and socially engaging activities, key intrinsic and extrinsic reasons for sustaining motivation in these activities have not been identified. While prior research investigated older adults and PA motivation, little research has been conducted in the intersection of older adults, motivation and gamified PA technology. Therefore, we conducted a synchronous three condition experimental study which included a control group, to investigate the motivational affordances specific to older adults PA technology.

3 Experimental Study

This study was conducted to investigate motivational affordances in PA technology over a longer period. This was undertaken to understand older adults’ preferences and challenges with motivational affordances in a controlled study [50, 51].

3.1 Research Design

In comparison to existing PA technology available commercially the rationale for selecting Spirit50.com as a technology intervention for this study was because it was a gamified fitness activity intervention website for use by older adults. Spirit50 was specifically designed for adults over 50 years of age and incorporated the following gamification elements: goal definition (quest), daily challenges (sub-goals), goal progression meter, points and badges (stars) as motivational affordances (gamification) [27, 51]. Using pedometers was considered to be the non-gamified PA intervention.

Prior research indicated that a minimum effective exercise program for habit formation was six weeks [64,65,66]. Therefore, a randomized experimental study investigating PA over an eight-week duration was designed. This study was a synchronous, three-condition, eight-week study with a total of thirty participants. Participants above 50 years of age were randomized to one of three conditions:

  1. 1.

    Physically active and use a gamified physical activity app (Spirit50.com)

  2. 2.

    Physically active and use a pedometer

  3. 3.

    Physically active (control group)

All three groups had participants who were physically active. The inclusion of a control group was done to make effective comparisons the two groups receiving the intervention.

3.2 Research Question and Hypotheses

The primary question driving this research was:

Can gamification elements be used to foster the intrinsic and extrinsic motivations for physical activity and daily exercise routines among older adults?

Hypotheses.

The hypotheses are rooted in established and validated studies which have investigated the following: the benefits of web based interventions to promote PA by sedentary older adults [67]; improved engagement of older adults with dementia when using artificially intelligent assistive technology [68]; improved motivational benefits of using a telehealth intervention using Nintendo Wii Fit Balance Boards [69]; and changes in attitude towards individual health [70].

Brawley et al. reported that self-monitoring was the behaviour modification strategy that led to maximum participation in PA [3]. Theory-based behaviour change methods for motivating adults’ in participating in PA, providing self-monitoring and socialising modifiers can be done using gamification strategies [13, 71,72,73]. The question is to investigate whether such strategies can be used to foster intrinsic and extrinsic motivation for older adults PA.

These studies provide evidence to support the usage of technology artifacts for PA challenges faced by older adults. Our methodical approach was to leverage this evidence to investigate the effectiveness of gamified technology artifacts as influencers of health behaviour change for improving PA in older adults’. Because of this, my hypotheses are:

  • H0: If older adults use traditional or classic physical activity applications interventions, there is no difference as measured by enjoyment and engagement compared to when they use gamified physical activity applications to influence change in health behaviour.

  • H1: If older adults use traditional or classic physical activity applications interventions, they are less engaged as measured by enjoyment and engagement than when they use gamified physical activity applications to influence change in health behaviour.

  • H2: If older adults use traditional or classic physical activity applications interventions, they are less motivated than when they use gamified physical activity applications to influence change in health behaviour.

3.3 Measures for Dependent Variables

A prior systematic review on the enjoyment of digital games attributed the experience of enjoyment as key to engagement in digital videogames [74, 75]. Boyle et al. (2021) and Mekler et al. (2014) conceptualised the subjective experience of engagement as a moment-to-moment feelings of enjoyment that players experience while gaming [74, 75]. Game enjoyment has been equated with the experience of flow as in the state of immersion or an experiential state of play when interacting with the gaming system [74,75,76,77,78,79,80]. The most frequently used measures of enjoyment was the IMI [75]. Doing an activity for its inherent satisfaction (intrinsic motivation) [81, 82], as opposed to an external outcome translates to paratelic engagement [83,84,85]. Researchers have also used the IMI as a measure of engagement [74, 85]. Therefore, the Intrinsic Motivation Inventory (IMI), an instrument that stemmed from the constructs of the Self-Determination Theory (SDT) [81] was selected as a measure of enjoyment and engagement.

From a motivation to participate in PA, the Psychological Need Satisfaction in Exercise (PNSE) scale has been used by researchers to measure perceived autonomy, perceived autonomy and perceived relatedness [86,87,88]. Preliminary eligibility to participate in the study was established through the PAR-Q [89] document. Baseline assessment of current PA condition of each participant was assessed using the International Physical Activity Questionnaire (IPAQ) [90]. Participants from the three conditions filled in a questionnaire once a week, for eight-weeks, which combined the scales measuring the following dependent variables:

  1. 1.

    Measuring the enjoyment and engagement of the participants over the eight-week period using the IMI Scale [91, 92].

  2. 2.

    Measuring the motivation aspect of the participants over the eight-week period using the PNSE scale [87].

  3. 3.

    User Experience: Measuring exertion using the Rating of Perceived Exertion scale (RPE) [93] after each participant session.

4 Analysis

4.1 Participants

At the onset, there were three groups of participants, 10 participants in each group, and eight weeks of sessions leading to 240 instances of data collection points.

4.2 Data

Quantitative Analysis.

SPSS was used to analyze the data for normality and subsequent significance testing. Data from the PNSE and IMI scales were tested for normality within each group using the Kolmogorov-Smirnov (K-S) and the Shapiro-Wilk tests.

For the PNSE scale, the dimensions of perceived competence, perceived autonomy and perceived relatedness, D(80), p <  0.05, were significantly non-normal. The Shapiro-Wilk statistic yields exact significance and is more accurate than the K-S test [94]. All dimensions in the three groups, D(80), p < 0.05, were significantly non-normal.

5 Results

5.1 Findings from Quantitative Analysis

The PNSE scale [87] measures perceived need satisfaction and is based on the view that psychological needs are a motivating force in achieving certain goals [95]. This scale was used to measure participants’ motivation for PA and exercise routines on a weekly basis. The IMI was used to evaluate the participants’ subjective experiences, and engagement related to the specific intervention associated with each group. The Rate of Perceived Exertion (RPE) scored participants feeling of exertion, weekly, on a scale of 0 (no exertion at all) to 10 (highest exertion).

Answers from participants for these two scales, collected over an eight-week period were compared between the three groups (Group 1 = gamified, Group 2 = non-gamified, Group 3 = control). This study had 10 participants in each group with a total of 80 responses in each group.

5.2 Overall Tests for Repeated Measures Within Groups

Friedman’s ANOVA was used to test for differences within groups in a repeated measures design where each participant within a group did the exercise routines on a weekly basis and reported their experiences using the PNSE and the IMI scales.

Friedman’s ANOVA (PNSE)

For the PNSE scale, the analyses are shown in Table 1.

Table 1. Output from Friedman’s ANOVA (PNSE)
Full size table

The three dimensions of the PNSE scale: perceived competence, perceived autonomy and perceived relatedness indicated significance (p < 0.05) depending on which type of intervention.

There was a statistically significant difference in Group 1 (χ2(2) = 44.5, p < 0.05), Group 2 (χ2(2) = 66.6, p < 0.05) and Group 3 (χ2(2) = 80.3, p < 0.05). This indicated that the interventions helped to significantly change the dependent variables over the course of eight-weeks. Follow-up tests were carried out and are explained in Sect. 5.4.

Friedman’s ANOVA (IMI)

For the IMI scale, the data analyses are shown in Table 2.

Table 2. Output from Friedman’s ANOVA (IMI)
Full size table

The dimensions of the IMI scale indicated significance (p < 0.05) depending on which type of intervention. There was a statistically significant difference in Group 1 (χ2(2) = 222.0, p < 0.05), Group 2 (χ2(2) = 286.4, p < 0.05) and Group 3 (χ2(2) = 274.6, p < 0.05). This further indicated that the interventions helped to significantly change the dependent variables over the course of eight-weeks. Follow-up tests were carried out and are explained in Sect. 5.4.

5.3 Overall Tests Between Groups

Data were ordinal, non-normal and were binned into the three groups using the grouping variable and tested using the Kruskal-Wallis test for differences.

Based on the Kruskal-Wallis test for the PNSE scale, motivation was significantly affected by the interventions for the dimensions related to perceived competence [H(2) = 28.77, p < 0.5], perceived autonomy [H(2) = 8.76, p < 0.5], and perceived relatedness [H(2) = 17.60, p < 0.5].

Jonckheere-Terpstra test also revealed a significant trend between the groups in the perceived competence (J = 6491, z = −5.33, r = −.34) and the perceived relatedness dimension (J = 8064, z = −2.63, r = −.17). Since the groups were coded as 1 = gamified, 2 = non-gamified and 3 = control, and the negative value of the z statistic indicated a trend of descending medians as the coding variable got bigger, which indicated a rising trend toward the gamified group.

Based on the Kruskal-Wallis test for the IMI scale, significance was indicated in specific dimensions (all effects are reported at p < 0.5). Engagement was significantly affected by the interventions: interest/engagement (H(2) = 12.45), perceived competence (H(2) = 39.65), effort/importance (H(2) = 6.21), pressure/tension (H(2) = 12.56), perceived choice (H(2) = 12.5), value/usefulness (H(2) = 6.43), relatedness (H(2) = 10.42).

Jonckheere-Terpstra’s test for the IMI scale revealed a significant trend in the data: since the groups were coded as 1 = gamified, 2 = non-gamified and 3 = control, and the negative value of the z statistic indicated a trend of descending medians as the coding variable got bigger, which indicated a rising trend toward the gamified group. Significant trend in the data was seen in the following dimensions: Interest/Enjoyment: J = 7602, z = −3.42, r = −.22; Perceived Competence: J = 5824, z = −6.46, r = −.41; Effort/Importance: J = 8272, z = −2.28, r = −.14; Perceived Choice: J = 11616, z = 3.45, r = .22; Value/Usefulness: J = 8116, z = −2.60, r = −.16

Kruskal-Wallis Test (RPE).

The comparison for RPE showed significant exertion between the groups H (2) = 24.3, p < .05. The Jonckheere-Terpstra’s test revealed a significant trend in the data: J = 12277, z = 4.618, r = .30. The positive z statistic indicates a rising trend of medians as the coding variable increased, indicating that the participants in the gamified group (Group 1) felt lower exertion compared to the participants from the control group (Group 3).

5.4 Follow-Up Tests

Mann-Whitney test were used to follow up the findings by comparing Group 1(gamified) and 2 (non-gamified), Group 1 and 3 (control), and Group 2 and 3. Bonferroni correction was applied and all effects are reported at 0.0167 (p < 0.05/3) level of significance.

Mann-Whitney Test (PNSE).

Exercise need satisfaction was compared between groups and the results are indicated below:

  • Gamified (Group 1) – Non-gamified (Group 2): Comparing the mean ranks between the groups, Group 1 (gamified) indicated higher perceived competence and perceived relatedness in comparison to Group 2.

  • Tests indicated that perceived competence (U = 2341, r = −0.23) and perceived relatedness (U = 2125, r = −0.29) were significant at this level and ranked higher in Group 1.

  • Gamified (Group 1) – Control (Group 3): Comparing the mean ranks between the groups, Group 1 (gamified) indicated higher perceived competence and perceived relatedness in comparison to Group 3.

  • Tests indicated that perceived competence (U = 1629, r = −0.42) and perceived relatedness (U = 2125, r = −0.26) were significant at this level and ranked higher in Group 1.

  • Non-Gamified (Group 2) – Control (Group 3): When comparing the mean ranks between the groups, Group 2 (non-gamified) indicated higher perceived competence in comparison to Group 3. However, while Group 3 indicated significance in the perceived autonomy dimension (p < 0.0167). Tests indicated that perceived autonomy (U = 2712, r = −0.13) were significant at this level and ranked higher in Group 3.

Mann Whitney Test (IMI).

Experience and engagement in the intervention routines were compared between groups as a follow-up test. Results are indicated as follows:

  • Gamified (Group 1) – Non-Gamified (Group 2): While the mean ranks of perceived competence and perceived relatedness were higher in Group 1 (gamified), the rankings in perceived choice were higher in Group 2 (non-gamified) than in Group 1. All effects reported at p < 0.0167, perceived competence (U = 2312, r = −0.24), perceived choice (U = 2489, r = −0.19) and relatedness (U = 2345, r = −0.23) were significantly different between the groups.

  • Gamified (Group 1) – Control (Group 3): While interest/enjoyment, perceived competence, effort/importance dimensions ranked higher for the non-gamified group, perceived choice was ranked higher in the control group. All effects reported at p < 0.0167, interest/enjoyment (U = 2201, r = −0.27), perceived competence (U = 1481, r = −0.46), effort/importance (U = 2448, r = −0.20), perceived choice (U = 5440, r = −0.27) dimensions indicated a significant difference between groups.

  • Non-Gamified (Group 2) – Control (Group 3): All effects reported at p < 0.0167, interest/enjoyment (U = 2482, r = −0.19), perceived competence (U = 2030, r = −0.31), and pressure/tension (U = 2139, r = −0.29) dimensions showed significant differences between the two groups. The above three dimensions also ranked higher in the non-gamified group compared to the control group.

6 Discussion

We conducted a synchronous, eight-week study with three groups: gamified, non-gamified and a control group to investigate whether gamification elements or motivational affordances could foster older adults PA.

To understand older adults’ enjoyment and experiential aspects of using technology for PA, it was critical to examine the relevance of technology in the context of PA motivation, setting up goals, feeling of accomplishments, fears and barriers, and rewards, and tracking. By investigating the influence of gamification elements in PA technology, this paper extends prior work of using web-based interventions to promote PA by sedentary older adults (55+) [67], supporting improved behavioural changes and effective changes in PA of older adults (50+) [96] due to computer-tailored interventions and justifying the need for improved web-based interventions for older adults (50+) for better sustainability [97].

The quantitative analyses indicated that adding gamification elements such as goal-based fitness quest scaffolded short-term goals focussed on working towards long-term goals and increased older adults’ engagement in PA. Adding the choice of difficulty levels and differentiated exercise intensity levels in (Group 1) fostered perceived autonomy, and perceived competence, two of the three major constructs of the SDT [81]. These results supported prior game based PA intervention studies where rules, challenges, competition, rewards and feedback mechanisms served as motivational strategies for PA [98, 99]. Perceived autonomy was significant in the non-gamified group (pedometers) with increased walk distance and duration than the participants in the control group. These findings extend the results of prior studies on PA and older adults using gaming technologies [48, 100]. Addition of motivational affordances (gamification elements) in Group 1 provided older adults with the choice of keeping track of their progression, achievements and providing them with an improved sense of control of their exercise selections and establishing a validation of their PA efforts on a weekly basis. These results also extend the results of prior studies on improving PA in older adults through mail-based interventions [101, 102] in the context of gamified PA interventions. Furthermore, this shows us that technology facilitation of PA can be achieved through the usage of motivational affordances through the gamification construct and quantification of PA.

These results reject the null hypothesis (H0) and supports the hypothesis that gamified PA applications would increase participant engagement and motivation in PA activity (H1, H2). This further shows that gamification elements or motivational affordances can help with improving engagement because these provide opportunity for improving one’s skill sets (competence), choice of challenges, intensity levels and type of exercises (autonomy). Furthermore, addition of these elements serves as a basis for customization and personalization of PA technology. This allows for tailoring fitness apps for older adults PA mindful of their physical limitations, barriers to exercise and current health conditions. Therefore, this paper extends the concept of adaptive engagement [27, 50] where tailoring of the user’s experiences can be designed to foster unique enjoyment through customized PA technology.

7 Conclusion and Limitations

While motivational affordances or gamifications elements have been used in many areas, research studies are often limited to pre-test and post-test intervention designs which were carried out over a short timeframe. However, effectiveness of the inclusion of gamification elements into PA technology for older adults can be determined only over a long-term engagement with such technology. To bridge this gap, we conducted a synchronous, three condition, eight-week study which included three groups: gamified, non-gamified and a control group. Participants (50+ years old) were randomized into the three groups and answered the IMI [91, 92], PNSE scale [86,87,88] and long form questionnaires. Results indicated that engagement was higher in the gamified and the non-gamified group compared to the control group. Furthermore, engagement was higher in the gamified group compared to the non-gamified group implying that gamifications elements can contribute to the enjoyment of PA over a longer period of time. Therefore, gamification elements or motivational affordances can help to foster competence and autonomy among older adults mindful of their physical limitations, barriers to exercise and current health conditions. While gamified technology for PA can help with motivation, engagement and enjoyment, participants in the gamified group may be biased due to novel features in the gamified PA application. Maintaining enrolment in the eight-week program or fatigue was also a major concern during the experimental study. Three participants from the control group decided to drop out from the study which required enrolment of new participants to this condition. Additionally, personalities and temperament of older adults also change with age-related impairments and current health conditions. Finally, older adults’ perception of the games and gamification in the context of their understanding of gamification can be different in comparison to younger adults already adept with such PA technology. Therefore, onboarding of older adults with the stages and complexities of gamification technology could help alleviate their fears of technology acceptance.