Virtual reality among the elderly: a usefulness and acceptance study from Taiwan

The World Health Organization (WHO) projects that the global population will become a super-aged society by 2030 [1]. Currently, 12.5% of people in the world are aged 60 years or older, which is referred to an old society [1], while 24% of Taiwan’s population will be in the older age group by 2030 [2]. Declining mortality rates have been national success stories for the well-being of societies, but they come with consequences of rising morbidity [3, 4]. A study revealed that declines in physical and mental health, the loss of functional capabilities, and a weakening of social ties represent obstacles to active aging among institutionalized older adults [5]. In addition, life satisfaction among older people, with reduced self-care capacity and physical inactivity (PI), was found to be related to several factors, with social, physical, and mental aspects, such as interacting with each other; especially feeling lonely, the degree of self-care capacity, poor overall health with sleep-wake disturbances, a worried feeling, and poor financial resources [6]. The elderly also tend to have a poor quality of life [7] and develop mobility limitations [8] and functional disabilities [4].

Physical activity is seen to play an important role in improving the functional status, psychological status, and well-being, and has social benefits [9]. Several interventions to improve the physical and cognitive functioning among older people were reported [10‐12]. Also, gaming interventions such as virtual environments [13, 14], Pokémon Go [15, 16], and Nintendo Wii™ [17] are being used as supporting programs for older populations. One such emerging technology is virtual reality (VR), a three-dimensional computer-generated world wherein a person can move about and interact, as if one is actually present in the imaginary space [18]. VR currently has several applications in the medical domain [18], and it also generates a secure environment to carry out certain activities [19, 20]. Evaluations of the effectiveness of VR among older populations revealed positive results for VR as a tool to reduce risks of falls [21‐24], and high adherence was also observed for home-based VR settings [25] and cognitive rehabilitation [26]. Moreover, it is seen as being well-tolerated by older populations and can stimulate autobiographical memories during reminiscence therapy [27]. VR interventions have shown positive effects on the social and emotional well-being of older adults [28]. Existing research focused on the effects of VR among older populations, and its uses and benefits. However, there is limited research on the acceptance of VR among older adults and how they perceive it.

Despite the benefits offered by VR-based rehabilitation technologies, attitudes among elderly people toward using VR are still unclear [29]. Our study investigated the acceptance and use of VR among an older age group. The Technology Acceptance Model (TAM) was used to assess different factors affecting the acceptance of VR.

The TAM was introduced by Davis (1986) and was based on the Theory of Reasoned Action (TRA). The TRA postulates that one’s behavior is directly influenced by behavioral intentions which are based on one’s attitudes and subjective norms toward a given behavior [30, 31]. The TAM illustrates a user’s acceptance of a new technology [30, 32]. The model has been extensively used in different research fields to anticipate user behaviors. It provides a theoretical basis to help explain relationships among various external variables, perceived usefulness (PU), perceived ease of use (PEOU), attitudes, intention to use (IU), and actual use [30, 33, 34]. Therefore we used the TAM to assess these factors based on a TAM scale developed by Davis [30]. We also included ‘social norms’ (SNs) [35] and ‘perceived enjoyment’ (PE) from Venkatesh [36, 37].

Davis proposed that PU is ‘the degree to which one presumes that using a specific system would augment his or her job performance’. In our study, PU was defined as ‘the degree to which older people presume that use of VR would augment their daily activities’. Also, according to Davis, PEOU refers to ‘the degree to which one accepts that using a specific system would be free of effort’ [38], and in our study, PEOU was defined as the degree to which older people perceived that using VR was free of effort. According to Davis and other studies, PU and PEOU are important factors in the acceptance of a technology [38, 39]. Thus, in order to verify if the relation between the influence of PEOU and PU in regards to VR acceptance can be applied to this study, we proposed the following hypotheses:

SNs are defined as ‘a person’s perception that most people who are vital to him, think that he should or should not perform the behavior in question’ [40] and in our study, SNs were defined as ‘older adults’ perception that people who are important to them think that they should or should not use VR’. As per previous research, SNs are seen to influence the intention to use (IU) a technology [41, 42]. Thus, we formulated a hypothesis to verify this in the case of VR:

According to Van der Heiden, in hedonic systems, “enjoyment” is an important determinant in the intention to use something [43], and Heerink et al. indicated that enjoyment is also a factor that influences the intention to use something [44]. “Perceived enjoyment” (PE) has been defined as ‘the degree to which using a technology is perceived as fun’ [36, 45]. As VR is also partly hedonic, enjoyment could be a factor that should be included in an analysis of acceptance among older adults. Thus, in our study, PE was defined as the degree to which VR was perceived as fun, and the related hypothesis was:

User experience is a concept covering the physical and psychological reactions or behaviors of the user before, during, and after using a technology or a product [46]. Ning and Kim demonstrated that user’s experience has an effect on usage intentions through perceived utility [47]. In our study, user experience (UE) was the behavior of older participants after using VR. We formulated two hypotheses as:

We composed our research model based on the TAM and a literature review as depicted in Fig. 1. This study attempted to explore the acceptance of the VR among an older population while mitigating problems of aging.

The study was conducted among 30 participants of an older age group (six males and 24 females).

Participants were included under the following criteria: 1) aged 60 years and above, 2) visited the Taipei Medical University (TMU) aging center, and 3) were willing to participate in the study components.

A signed consent form was received from each participant after explaining to them the objective of the study, which was carried out from March to May 2018. Mandarin Chinese versions of the questionnaire and consent form were filled out by participants. The questionnaire is given in Additional file 1 (based on the TAM and SNs, UE, and PE from Venkatesh). The questionnaire was used to collect participants’ perceptions about their acceptance of VR. Six sets of questions were clustered under the variables of PU, PEOU, PE, SN, UE, and IU based on a literature review.

We used the Vive htc VR system for the study. These devices contained nine different applications (Appendix) that encouraged physical activity, motivated, and provided entertainment to older people. We ensured that participants accessed the different VR applications at least once during the study period. Participants were asked to use VR for 15 min twice a week for 6 weeks and respond to the questionnaire in Additional file 1 at the end of the study. A 5-point Likert scale was used to score the questions, in which participants were asked to agree or disagree based on a ranking from 5 to 1, where 5 indicated strong agreement, 3 indicated a neutral opinion, and 1 indicated strong disagreement. Information obtained from respondents was centered on their experience and perceptions while using VR. The study was approved by the ethics committee of the Taipei Medical University-Joint Institutional Review Board.

Researchers analyzed the data using SPSS vers. 20 (SPSS, Chicago, IL, USA). Pearson’s product moment correlation was conducted to examine the criterion validity of the questionnaire. Cronbach’s alpha reliability analysis was used to test the internal consistency of the questionnaire items. A linear regression analysis was utilized to identify relationships among the variables: PU, PEOU, UE, SNs, PE, and IU. A mediation analysis was conducted to determine the mediating effect of different variables in the questionnaire.

Table 1 shows characteristics of the 30 older participants. Of these, six (20%) were males, and 24 (80%) were females, with the highest percentage of participants aged 70~75 years.

Figure 2 shows a summary of the responses of the 30 older participants. They were asked about the PU of VR, and 77.8% agreed that VR helped raise their mood, and entertain and keep them motivated to do their daily activities. Participants who believed that VR was easy to use accounted for 64.4, and 80% of participants found it enjoyable. For subjective norms with regards to the support of the use of VR by family, friends, and caretakers, 55.6% agreed, and 77.8% of users expressed a positive experience. Around 71.7% of participants intended to use VR in the future.

Summary statistics and frequency distributions of all the variables are shown in Table 2, with a mean range of 3.27~4.07, and standard deviation of ≤0.661.

The validity coefficient and p values of items in the questionnaire are shown in Table 3.

The critical value of Pearson’s product-moment was 0.361, according to the table of critical values [48], for 30 respondents. As shown in Table 3, for all our variables, the r table product moment (rxy) was > 0.361 for p values of < 0.05.

Cronbach’s α reliability analysis was conducted to assess the internal consistency and reliability of the variables in the study (Table 4).

Seven hypotheses were formulated based on the initial TAM model and a literature review. Each hypothesis was tested for significance based on the regression statistics. Table 5 presents the summary of the regression data for each hypothesis. Larger beta values were observed for larger t values and smaller p values across all hypotheses.

All the hypotheses are detailed as below:

Table 6 describes mediating factors of possible pathways, using a bootstrap analysis method.

Figure 3 describes models of the mediation effect of different variables through pathways A, B, and C′. In models 3C and 3E, paths A and B were significant, and C was not significant. Table 5 indicates that H1 was significant, which was not supported when mediation was used. Thus, SNs provided a strong mediating effect for determining the influence of PU on IU, and PE provided a strong mediating effect for determining the influence of PU on IU. Models 3F and 3G indicated that paths A, B, and C were significant; however, the mediators did not give significant effects for the models.