Introduction

With the advancement of digital service technologies, various services have transitioned into digital experiences, resulting in a more convenient and efficient user experience (Kim and Kang, 2023). Despite the rapid advancement of digital services, not all user groups benefit equally. In particular, older adults often experience difficulties searching for, evaluating, and utilizing digital services due to age-related declines in perceptual, cognitive, and physical abilities (Sriwisathiyakun and Dhamanitayakul, 2022; Lee et al., 2024). These challenges are frequently compounded by heightened anxiety when interacting with digital technologies, which hinders effective use (Andrews et al., 2019). As the aging population grows, the digital divide between older and younger generations is poised to become an increasingly pressing societal issue (Lee et al., 2024). In response to these challenges, prior research has emphasized the importance of leveraging technological innovation to enhance the quality of life and autonomy of older adults (Ausserhofer et al., 2024) while also promoting the development of inclusive and integrated Smart Healthy Age-Friendly Environments (SHAFE) (Dantas et al., 2019).

Within the dual framework of digital service development and SHAFE construction, Immersive Virtual Reality Travel Applications (IVRTA) have emerged as digitally mediated interventions. IVRTA exhibits significant potential to advance gerontological digital inclusion and operationalize SHAFE objectives by integrating multisensory engagement with advanced service architectures. Empirical studies indicate that although IVRTA could significantly enhance social connections and life management for older adults, effectively improving their quality of life and well-being (Fiocco et al., 2021), older adults’ limited familiarity with IVRTA often leads to digital experiences that markedly differ from conventional experiences (Zhang et al., 2016). This discrepancy consequently increases psychological stress and negative emotions during usage, ultimately diminishing the overall user experience of IVRTA (Hauk et al., 2018; Thach et al., 2020). Therefore, further aging-friendly design research around IVRTA is a much-needed problem-solving solution to enhance digital services and SHAFE construction for older adults.

Current research investigations into IVRTA utilization among aging populations remain constrained in scope, predominantly emphasizing tourism marketing and consumer behavior paradigms (Tom Dieck et al., 2019; Hao et al., 2024), with minimal attention directed toward optimizing interface design for older users. Empirical evidence confirms that familiarity-driven interface architectures significantly enhance operational efficiency and experiential quality (Norman, 2013; Pan et al., 2015; Tom Dieck et al., 2019; Lee et al., 2024), especially for older adult users, who have a significant preference for familiar environments and objects (Norman, 2013; Tom Dieck et al., 2019; Lee et al., 2024; Hao et al., 2024). Based on this, this study synthesizes the actual needs of older adults during their IVRTA experience (Tom Dieck et al., 2019) with the advantages of the familiarity design approach (Tom Dieck et al., 2019; Zhang et al., 2019; Lee et al., 2024) to propose a familiarity design approach for IVRTA interfaces from the perspective of older adults. This approach aims to make the interface more intuitive and easy to use by adapting the system presentation and functional interaction design to the cognitive model and experience background of older adults in their daily lives to enhance their sense of familiarity, self-confidence, and motivation in the virtual environment (Zhang et al., 2019; Lee et al., 2024). Specifically, the familiarity design approach focuses on three core dimensions: representation (interface appearance), manipulation (interface interaction), and guidance (information guidance). To simulate the user interface and operation of a familiar product for real-world older adults, the representation and manipulation dimensions of this study focus on the clear distinction between visual and interactive information in the user interface, presented as either a Dual-Layer Three-Dimensional (D3D) user interface or a Single-Layer Three-Dimensional (S3D) user interface. The guidance dimension focuses on using an Anthropomorphic Virtual Agent (AVA) or a Non-Anthropomorphic Virtual Agent (NAVA) to inform older adults about the scenic spots they encounter. The familiarity design approach aims to enhance the user experience of older adults by evoking their previous experiences to fit their cognitive preferences and behavioral patterns.

This study will address two fundamental research questions: whether D3D-AVA integration significantly affects the experiences of older adults when using IVRTA (RQ1) and through what mediation pathways the familiarity design approach influences older adults’ CI (RQ2).

Literature review

In the literature review section, this study begins by defining the familiarity design approach and summarizing the existing literature on utilizing this approach to enhance the design of digital products for seniors. Second, based on the relevant literature on familiarity design techniques and product features, summarize the strategy suitable for IVRTA design and elaborate on the specific application methods and study hypotheses. Finally, the technology acceptance model (TAM) and self-efficacy theory (SET) will be used to investigate how the familiarity design approach influences older adults’ CI.

Familiarity design and aging

Familiarity refers to the extent of an individual’s experience with a specific subject and is essential for improving the usability of products or services (Pan et al., 2015; Tom Dieck et al., 2019). Applying familiarity design could significantly enhance usability, even for users without prior exposure to the product or service (Norman, 2013). Previous research has shown that older adults prefer familiar environments and objects, as their mental models align more closely with well-known items (Pan et al., 2015; Lee et al., 2024). The familiarity of older adults is often associated with older products (Lawry et al., 2019). Hollinworth and Hwang (2011) explored the impact of familiar interactions on older adults’ use of computer applications, highlighting the significant influence of familiarity. Lee et al. (2024) employed a familiarity design approach to develop a self-service food ordering machine interface for older adults, finding that this approach improved order accuracy and reduced task completion time. These findings suggested that familiarity design supports older adults in adapting to and mastering new technologies, improving their overall user experience.

Pan et al. (2015) outlined three familiarity design approaches: symbolic, cultural, and manipulability. The design applications of symbolic familiarity refer to incorporating objects and activities into the target users’ daily lives in the design process. The cultural familiarity design approach emphasizes concepts, artifacts, patterns, traditions, or rituals that align with the user’s cultural background. The manipulable familiarity design approach refers to interaction behaviors reflecting symbolic and cultural familiarity design approaches that aim to mimic real-world interactions. On the other hand, Lee et al. (2024) proposed three familiarity design approaches specifically for products aimed at older adults: representation, manipulation, and organization. Representation involves matching sensory information, such as visual, auditory, and tactile cues, with real-world experiences. Manipulation ensures that user interactions correspond to real-world actions while the organization structures data consistently with prior experiences.

According to a summary of previous research on familiarity design, familiarity-based interface design techniques have proven successful in various digital product scenarios (see Table 1). Empirical studies on the use of IVRTA by older adults are relatively scarce, nonetheless. Based on the research by Pan et al. (2015) and Lee et al. (2024), this study proposes three design approaches of familiarity—representation, manipulation, and guidance—for developing IVRTA tailored to older adults. These approaches are informed by familiarity and integrate the primary needs of older adults using IVRTA, as identified by Tom Dieck et al. (2019): sense of presence, control, and storytelling. Under the familiarity design approach to representation, this study incorporates concepts from cultural familiarity design (Pan et al., 2015), symbolic familiarity design (Pan et al., 2015), representation (Lee et al., 2024), and presence (Tom Dieck et al., 2019). In the virtual environment, multisensory experiences are facilitated, where ‘representation’ extends beyond visuals to encompass cultural, auditory, and tactile elements, each conveying symbolic meanings.

Table 1 Previous research on familiarity design.
Full size table

The study also integrates operable (Pan et al., 2015), manipulation (Lee et al., 2024), and sense of control (Tom Dieck et al., 2019) into the familiarity design approach of ‘manipulation,’ which is mediated by the power of information input devices during user interactions with the immersive virtual reality applications. The approach of ‘guidance’ combines the concepts of organization (Lee et al., 2024) and storytelling (Tom Dieck et al., 2019), referring to the practical guidance provided to the user within the virtual environment. In an immersive virtual reality application environment with complex information, a structured, narrative-driven guide reduces cognitive load and enhances user immersion.

This study employs IVRTA as a modified design content, applying the ‘representation’ and ‘manipulation’ approaches to the D3D user interface and comparing these with S3D. The ‘guidance’ design approach is outlined in AVA and contrasted with NAVA. This study will evaluate the effectiveness of the familiarity design approach in improving CI among older adults who use IVRTA.

D3D user interface focusing on representation and manipulation

IVRTA features a 3D user interface that enables users to interact with digital content in immersive virtual reality environments (Yue, 2021). The D3D user interface organizes content into a foreground layer for essential information and a background layer for secondary information, facilitating users’ focus on primary content without distraction from less critical elements (Kharoub et al., 2019a, 2019b).

D3D user interface design combines visual representation and functional manipulation of familiar design approaches, crafting an interface that mimics real-world experiences in appearance and interactivity by utilizing multiple interactive layers rather than simply displaying information (Kharoub et al., 2019b). By consistently presenting essential information within the user’s field of view, the D3D user interface reduces cognitive load and enhances operational efficiency. Despite the advantages of the D3D user interface, S3D user interface design applications have gained popularity for providing users with heightened immersion. In applications like Brink Traveler and National Geographic Explore VR (Sihyun and Choi, 2021), the S3D user interface conceals input devices, such as joysticks, during interaction, diverging significantly from interaction methods in traditional digital products, like desktop or mobile applications (Kharoub et al., 2019b). However, due to the interface’s simplified design, S3D interfaces may present difficulties for older adults in recognizing and remembering interface functions.

Prior research has demonstrated that D3D user interface design improves usability and user experience in the user interface (Kharoub et al., 2019a, 2019b). However, previous studies do not precisely understand how older adults utilize the D3D user interface. First, the existing literature is limited and primarily emphasizes ‘manipulation.’ Secondly, prior research on D3D user interface for immersive virtual reality desktops has focused exclusively on interaction design, overlooking the assessment of information presentation (Kharoub et al., 2019b). Third, these studies have primarily targeted younger users, leaving a gap in studies on older adults.

This study addresses these limitations by applying familiar design approaches of ‘representation’ and ‘manipulation’ to create an IVRTA that closely mirrors real-world information presentation. The D3D user interface design prominently displayed the exit button on the interface. In contrast, the S3D user interface design does not display any buttons; instead, the button is located on the joystick. An interactive application was developed to replicate the real-world IVRTA experience. Various informational factors, such as scripting, color scheme, font size, and navigation structure, were meticulously controlled to enable a fair comparison between the D3D user interface and S3D user interface designs. Meanwhile, to answer RQ1, we used an experimental methodology to evaluate the following hypothesis regarding how the D3D user interface affects older adults’ experience:

Hypothesis 1 (H1): Older adults will report higher levels of (a) perceived ease of use (PE), (b) perceived usefulness (PU), (c) self-efficacy (SE), and (d) CI with the D3D user interface compared to the S3D user interface.

Hypothesis 2 (H2): Older adults in the D3D user interface condition will (a) engage with it for a longer usage duration and (b) request assistance less frequently than with the S3D user interface.

AVA focusing on guidance

The 3D medium of IVRTA differs significantly from traditional 2D media (e.g., smart devices and computers) by engaging a comprehensive range of visual, auditory, and tactile senses within an immersive virtual reality environment. Information reception involves selecting presentation methods, sequencing, and other factors influencing the process. This study examined the familiarity design approach for guidance IVRTA information: AVA.

AVA utilizes virtual reality technology to create virtual agents with human-like characteristics (Pimenta et al., 2022). In the AVA design, information is generated and presented by an anthropomorphic service robot (Garner et al., 2016; Carrozzino et al., 2018), which synthesizes and delivers information through a human-like character. Conversely, the NAVA design presents information in a decentralized format, using text boxes and audio outputs. Effective information presentation in immersive virtual reality environments is crucial, as it significantly impacts older adults’ ability to receive and process information (Yuan and Chee, 2005). AVA enhances interactions with older adults in immersive virtual reality environments by applying verbal, non-verbal, and interactive behaviors. The presence of AVA increases pleasure, satisfaction, and acceptance of virtual experiences among older adults (Liu et al., 2023). Thus, AVA has the potential to enhance the appeal of human-computer interaction (Yuan and Chee, 2005).

Older adults may be more familiar with AVA-based guidance, as it can closely resemble real-world scenarios that reinforce user familiarity. AVA functions as digital characters, assistants, or guides, improving user experience, immersion, and engagement (Yuan and Chee, 2005; Numata et al., 2020). AVA also provides emotional support to users through voice and emotive expressions, helping to alleviate feelings of loneliness (Potdevin et al., 2021). However, it is essential to recognize the limitations of incorporating AVA in immersive virtual reality applications. Due to its anthropomorphic nature, AVA requires substantial resources for modeling and programming, leading to increased memory usage. AVA is not widely implemented to minimize application memory demands and reduce the likelihood of system errors.

Takashi Numata et al. (2020) empirically demonstrated that the positive emotional expressions of AVA can elicit positive emotions in users, even when users are aware that a computer program controls AVA. Potdevin et al. (2021) investigated the intimacy experienced during interactions with AVA, comparing verbal and text-based chat. Their findings indicated that voice-based communication enhances virtual intimacy between a user and an AVA, capturing the user’s attention. Furthermore, previous research has shown that AVA significantly affects SE and ST in older adults (Baylor, 2009), suggesting that AVA engages older adults and enhances their SE when interacting with IVRTA (Cao et al., 2022).

Despite the limited number of studies on AVA, existing research suggests that AVA provides significant advantages in information presentation and user engagement (Carrozzino et al., 2018). AVA effectively captures user attention and enhances engagement compared to traditional explanatory panels and narrated speech. Meanwhile, AVA guides users through structured content, ensuring they do not miss important information or become disoriented in complex scenarios. (Liszio and Masuch, 2016; Bönsch et al., 2024).

In conclusion, the complexity of information within immersive virtual reality environments poses challenges for older adults, who often face more difficulties using immersive virtual reality applications than traditional media. The reception and processing of information appear to be substantial obstacles for older adults when interacting with immersive virtual reality applications. Research on the impact of information guidance on older adults remains limited.

Based on previous research, this study aimed to examine the effects of different information guidance styles on older adults by establishing AVA and NAVA designs. To ensure consistency in the information received by older adults, identical audio content was used in both situations. Also, to answer RQ1, we tested the following hypotheses through an experimental design to examine the effect of AVA on older adults’ experience:

Hypothesis 3 (H3): Older adults in the AVA design will report higher levels of (a)PE, (b) PU, (c) SE, and (d) CI compared to those in the NAVA design.

Hypothesis 4 (H4): Older adults in the AVA design will (a) demonstrate longer usage duration and (b) have a lower frequency of requesting assistance than those in the NAVA design.

Familiarity design approach influences the underlying mechanisms of CI in older adults’ use of IVRTA

Referring to previous findings on emotion-driven human-robot interaction studies, users generate emotional responses during interaction with interfaces, and these emotions largely influence their CI and decision-making (Alipour et al., 2023), but technological adaptation barriers are the main barriers to the use of digital technologies by older adults (Davis, 1989; Pfitzner-Eden, 2016; Lee et al., 2024). Based on this, the present study should primarily focus on exploring the mechanism of the familiarity design approach on CI, with a particular emphasis on the impact of technology. In the existing literature, the TAM and the SET are the classic theoretical frameworks (Davis, 1989; Chen and Chan, 2014) for exploring the influence of user technology on CI (Bandura and Watts, 1996; Lee et al., 2024). TAM explains the formation of CI mainly from the technology acceptance perspective, whereas SET focuses on users’ confidence in using the technology processes (Davis, 1989; Chen and Chan, 2014; Bandura and Watts, 1996; Lee et al., 2024). Therefore, we synthesize TAM and SET to construct a theoretical integration framework to explore the mechanisms of familiarity design approaches on CI from the perspective of technology impact.

Regarding the selection of mediator variables, previous studies have shown that PE and PU are key factors influencing CI in TAM (Davis, 1989; Lee et al., 2024). However, older adults are more concerned about PE than PU due to their limited cognitive resources (Hauk et al., 2018). Previous studies have also shown that SET emphasizes SE as a key psychological driver in the formation of CI (Pfitzner-Eden, 2016) and that older adults’ SE is key to overcoming technology anxiety (Bandura and Watts, 1996), which directly affects their CI (Lee et al., 2024). In addition, other studies have shown that social influence (e.g., evaluation by others) also affects users’ CI. However, IVRTA is typically an individual-use scenario, and the role of social influence (e.g., evaluation by others) is generally weaker. Therefore, in this study, PE, PU, and SE were used as mediating variables to explore the mediating pathways through which the familiarity design approach affects CIs in older adults using IVRTA, and the following mechanism hypotheses were made:

Mechanism 1 (M1): SE mediates the effect of D3D on CI.

Mechanism 2 (M2): PE mediates the effect of D3D on CI.

Mechanism 3 (M3): PU mediates the effect of D3D on CI.

Mechanism 4 (M4): PE mediates the effect of AVA on CI.

Mechanism 5 (M5): SE mediates the effect of AVA on CI.

Mechanism 6 (M6): PU mediates the effect of AVA on CI.

Method

The experiment employed a 2 (D3D vs. S3D) × 2 (AVA vs. NAVA) between-groups design experiment (see Fig. 1). Older adults in all four designs received the same information to exclude other factors that might interfere with the experimental results.

Fig. 1
figure 1

Screenshots of four different interface styles.

Full size image

Participants

The study employed a random sampling method to select four communities from among the city’s top 15 large residential communities as sample sites. These neighborhoods represented residents with varied financial and educational backgrounds. The research team posted participant recruitment notices within each community and recruited 92 older adult residents aged 65 years and older to participate in the experiment through random screening, with participants being paid for their participation. Initial screening excluded six participants. Of these, 50% (n = 3) were male and 50% (n = 3) were female. The specific exclusion criteria and number of exclusions were as follows: one participant was unable to provide informed consent; three participants had contraindications to IVRTA during use (including a history of frequent migraines and motion sickness); one participant consistently selected the same option throughout the questionnaire; and one participant had a questionnaire with missing answers. Data from the remaining 86 participants were included in the final analysis. Participants ranged in age from 65 to 83 years (Mean = 70.4, SD = 5.2), with 47.67% (n = 41) male and 52.33% (n = 45) female. All participants signed an informed consent form and were informed of their right to withdraw from the study at any time. The institutional review board of the authors’ university approved the research protocol.

Apparatus

The IVRTA used for the experiments was developed using UNITY software, and the test equipment consisted of a PICO 4 Ultra Enterprise head-mounted display (HMD) from PICO. This HMD consisted of two 2.56 inch Fast-LCD screens, providing a combined resolution of 4320 × 2160, a refresh rate of 90 Hz, and a 105-degree field of view, utilizing the PICP4 VR all-in-one system. Participants navigated the system using a single operating handle.

Manipulation of D3D user interface and S3D user interface design

We employ two 3D user interface designs: the D3D and S3D interfaces. The visual (representation) and the interactive (manipulation) are presented in D3D and S3D formats within the experiment. An additional interface layer is incorporated above the tour interface to display prompt messages and function buttons, following D3D design approaches (Kharoub et al., 2019a, 2019b). The function buttons utilize click-based interaction to reduce user difficulty and cognitive load. In contrast, the D3D features are omitted in the S3D design. The S3D interface retains the conventional interface and interaction methods found in existing IVRTA applications, requiring participants to memorize joystick operations for system interaction.

Manipulation of AVA and NAVA design

AVA in IVRTA serves to guide older adults in completing the virtual experience. Under the AVA design, the AVA will be present in the virtual environment, supplemented by explanatory audio to assist the user in navigating the virtual experience. Given that older adults prefer engaging with female AVA irrespective of gender (Esposito et al., 2021), this study selected a female-image AVA accompanied by female-voice explanatory audio. In the NAVA design, the virtual scene contained no AVA, presenting solely the explanatory audio.

Procedure

Participants were randomly assigned to various experimental conditions, and an HMD device was provided for use. Before the experiment commenced, the researcher provided all participants with a comprehensive overview of the fundamental functioning of the HDM and IVRTA, along with instructions for experimenting using the IVRTA (see Fig. 2). The researcher was not permitted to actively assist the participant during the experiment unless the participant verbally requested help. The purpose of this format was to encourage the participants to independently overcome the difficulties encountered during the use of the IVRTA. At the same time, the researcher used a stopwatch to record the amount of time the older adults spent using the IVRTA. Upon completing the assigned tasks, older adults were requested to complete demographic, digital literacy, and IVRTA experience questionnaires. The questionnaires were administered to gain insights into participants’ backgrounds, technical skills, and feedback regarding their experience with the IVRTA, thereby facilitating the evaluation of user experience and CI across various experimental conditions.

Fig. 2
figure 2

Overview of the experimental procedure.

Full size image

Measures

We assessed participants’ experiences with the IVRTA using self-reported and behavioral data collected during the experiment (see Table 2). While using the IVRTA, participants’ PE was evaluated through self-report measures, with three items assessing PU (Davis, 1989), three assessing SE (Chen et al., 2001), and two assessing CI.

Table 2 List of questions.
Full size table

The study also collected behavioral data to obtain objective results and mitigate potential response biases and social desirability effects (Calisto and Sarkar, 2024). The appeal of the IVRTA was measured by documenting participants’ total usage duration, with the hypothesis that if the IVRTA were challenging to use or visually unappealing, participants would spend less time using it (Cham et al., 2024). The experimental procedure involved the researcher starting a stopwatch when the participant donned the HMD device and stopping it when the participant pressed the “exit program” button. The researcher also recorded whether participants requested assistance to assess SE under different experimental conditions during the experiment. Participants completed the questionnaires to ensure they were unaware of any recording of their responses, preserving the authenticity of their answers.

The questionnaire included multiple items related to the experience of digital devices, with some questions about whether older adults had used IVRTA products. PE and PU were each assessed using three items from Davis (1989). SE was measured according to Bandura and Watts (1996)‘s SET. The ten-item scale (Roque and Boot, 2018) assesses participants’ electronic information literacy. The digital experience was evaluated by the number of smartphone applications used by older adults in the preceding week. All responses were gathered using a 7-point scale to more accurately capture older adults’ attitudes toward IVRTA usage (Zerényi, 2016). Cronbach’s coefficient was computed using SPSS 26.0 software to assess the internal consistency of each multi-item scale. The calculated results showed that Cronbach’s alpha coefficients for PU, PE, SE, CI, and digital proficiency were 0.963, 0.933, 0.941, 0.960, and 0.936, respectively. The coefficients were higher than 0.8, indicating that the question items’ internal consistency was high and could be used for further analysis.

Statistical methods

Statistical analyses of PE, PU, SE, CI, and usage duration data for this study were performed using SPSS 26.0. The data were rigorously validated before conducting Type II ANOVA and mediation analyses. Initial diagnostic testing confirmed data conformity to statistical model requirements: Levene’s test demonstrated homogeneity of variance across PE, PU, SE, CI, and usage duration data (p > 0.05). Normality assumptions were validated via Shapiro-Wilk tests (W > 0.95) and Q-Q plot residuals. Covariate-residual scatterplots exhibited linear relationships without heteroscedasticity, satisfying independence and linearity criteria. Mediation models underwent multicollinearity diagnostics, revealing tolerance values >0.5 and Variance Inflation Factors (VIF) <2.0 across predictors. The data on the frequency of requesting assistance in this study were analyzed descriptively only. Meanwhile, we employed G*Power 3.1 for post hoc power analyses to evaluate the statistical validity of the present sample size. Based on the actual observed effect sizes (e.g., \({n}_{p}^{2}\) = 0.204 for D3D vs. CI, corresponding to Cohen’s f = 0.45), the statistical power reached 99% at α = 0.05 and degrees of freedom (1, 81). For the medium effect size (f = 0.25), the power for the current sample size was 88%, above the 80% benchmark, confirming adequate sensitivity for detecting the hypothesized effects.

Results

Digital proficiency

This research investigated digital proficiency levels among older adults, revealing that a substantial majority, precisely 94.18% (n = 81), reported no prior experience with IVRTA. Findings on the frequency of smartphone applications usage in the preceding week showed minimal engagement (Median = 3.279, SD = 0.916). Objective data indicated that participants exhibited low digital literacy levels; however, subjective assessments showed that older adults reported relatively high levels of perceived engagement (Median = 5.5, SD = 1.003) and perceived proficiency (Median = 6.08, SD = 1.46) in using both IVRTA and smartphones.

The study applied analysis of covariance to ensure equal variance across groups, followed by a Tukey post hoc test to examine significant differences among the variables, which included age, frequency of IVRTA system usage, physical engagement during IVRTA use, smartphone proficiency, and the number of smartphone applications used in the preceding week. The results indicated no significant differences among these variables, suggesting a lack of statistically significant association between digital proficiency and the other variables.

Hypothesis testing

We used Type II ANCOVA to test the hypotheses. In selecting covariates, suitable candidates included the frequency of immersive virtual reality application usage and PE. However, as most older adults had not previously used an IVRTA, the median usage was 0, resulting in a lack of variability. Objective measures indicate a relatively low number of IVRTA experiences among older adults. At the same time, subjective assessments, such as PE, are disproportionately high, likely due to self-report bias influenced by social expectations. Therefore, another objective indicator, the number of smartphone applications used by older adults in the preceding week, was used as a covariate in this study to ensure that existing patterns of technology use did not confound the effects of any observed independent variables. This covariate was included for two reasons: first, it was based on previous research showing that the frequency of daily smartphone application use effectively indicates an individual’s technological adaptability and is significantly correlated with scores on the Digital Literacy Scale (Venkatesh et al., 2016). The second is that many older adults in China do not use computers extensively and are more familiar with smartphones. However, most objective digital literacy scales tend to use computer-based tests, which are not suitable for use with older adults in China. Therefore, smartphone use and digital competence have a stronger realistic correspondence and representation. This objective measure ensured that the effect of the independent variable was not confounded with participants’ existing technology use patterns, thereby enhancing the study’s rigor (Lee et al., 2024).

The effects of D3D and AVA on older adults’ experience

H1: Self-report results of D3D user interface and S3D user interface design

The analysis revealed a statistically significant main effect of the D3D user interface on PE (see Fig. 3) (F [1,81] = 5.147, p = 0.026, \({n}_{p}^{2}\) = 0.061). The mean PE score was higher for the D3D user interface (n = 44, Mean = 6.14, SD = 0.702) than for the S3D user interface (n = 42, Mean = 4.83, SD = 0.824). A statistically significant main effect of D3D user interface on PU (F [1,81] = 4.633, p = 0.034, \({n}_{p}^{2}\) = 0.055). The mean score for PU was higher for the D3D user interface (n = 44, Mean = 6.00, SD = 0.863) than for the S3D user interface (n = 42, Mean = 5.26, SD = 0.912). A statistically significant main effect of D3D user interface on SE (see Fig. 3) (F [1,81] = 6.642, p = 0.012, \({n}_{p}^{2}\) = 0.078). The mean score for SE was higher for the D3D user interface (n = 44, Mean = 6.27, SD = 0.758) than for the S3D user interface (n = 42, Mean = 4.21, SD = 0). A statistically significant main effect of D3D user interface on CI (see Fig. 3) (F [1,81] = 7.804, p < 0.001, \({n}_{p}^{2}\) = 0.204). The mean score for CI was higher for the D3D user interface (n = 44, Mean = 6.43, SD = 0.625) than for the S3D user interface (n = 42, Mean = 5.55, SD = 0.861).

Fig. 3
figure 3

Comparative results of self-reported main effects.

Full size image

Thus, H1a, H1b、H1c and H1d are supported.

H2: Behavior results of D3D user interface and S3D user interface design

No significant differences were observed in the usage duration older adults took to complete the excursion using the D3D user interface compared to the S3D user interface (see Fig. 4). All older adults participated in the experimental process, including those who requested assistance (n = 35). There was a notable disparity in the frequency of requesting assistance across conditions (see Fig. 5): 27 (77.14%) of the older adults requested assistance in the S3D user interface, while 8 (22.86%) requested help in the D3D user interface.

Fig. 4
figure 4

Comparative results of main and interaction effects of usage duration.

Full size image
Fig. 5
figure 5

Frequency of requests for assistance by older adults in different interface styles.

Full size image

Thus, H2a lacked support, and H2b got support.

H3: Self-report results of AVA and NAVA

We observed significant main effects of AVA on PE and CI (see Fig. 3). There was a significant main effect of AVA on PE (F [1,81] = 3.967, p = 0.05, \({n}_{p}^{2}\) = 0.048): the AVA (n = 43, Mean = 5.98, SD = 0.886) was perceived as higher than the NAVA (n = 43, Mean = 5.002, SD = 0.886). There was a significant main effect of AVA on CI (see Fig. 3) (F [1,81] = 7.804, p = 0.007, \({n}_{p}^{2}\) = 0.09): the AVA (n = 43, Mean = 6.47, SD = 0.592) was perceived as higher than the NAVA (n = 43, Mean = 5.53, SD = 0.855). However, no statistically significant main effects were observed for PU (F [1,81] = 4.633, p = 0.568, \({n}_{p}^{2}\) = 0.004) and SE (F [1,81] = 0.249, p = 0.619, \({n}_{p}^{2}\) = 0.003) (see Fig. 3).

The interaction effects between the virtual agent and user interface design were also tested, but none were significant for any of the dependent variables (see Fig. 6): PE (F [1,81] = 0.11, p = 0.741, \({n}_{p}^{2}\) = 0.001), PU (F [1,81] = 2.154, p = 0.146, \({n}_{p}^{2}\) = 0.027) and SE (F [1,81] = 0.249, p = 0.619, \({n}_{p}^{2}\) = 0.03).

Fig. 6
figure 6

Comparative results of self-reported interaction effects.

Full size image

H4: Behavior results of AVA and NAVA design

AVA had a significant main effect (see Fig. 4) (F [1,81] = 10.989, p = 0.001, \({n}_{p}^{2}\) = 0.122). The usage duration in the AVA (n = 43, Mean = 322.23, SD = 23.70) design is significantly higher than that used in the NAVA design (n = 43, Mean = 292.79, SD = 15.45). In addition, the frequency of requesting assistance was 15 (57.14%) for the AVA design and 20 (42.86%) for the NAVA design (see Fig. 5).

Thus, H4a was supported, and H4b was not.

A particular finding emerged regarding the interaction effect. Although there was no significant interaction effect in the self-reported results, there was a significant interaction between the D3D user interface and AVA design in terms of usage duration (F [1,81] = 11.634, p = 0.001, \({n}_{p}^{2}\) = 0.128). When the D3D user interface and AVA designs were met, users took longer to use (n = 22, Mean = 314.091, SD = 22.448) (see Fig. 4), and fewer users requested assistance (n = 3, 8.6%) (see Fig. 5).

Familiarity design approach influences the underlying mechanisms of CI in older adults’ use of IVRTA

The results of this study have confirmed that the D3D user interface and AVA design significantly positively affect CI when older adults use IVRTA. We sought to reveal the underlying mechanisms by which the D3D user interface and AVA design affect CI to answer RQ2. Multiple mediation analyses were conducted for this purpose. In this study, the hypothesis of indirect effects was tested using the percentile bootstrap method, a non-parametric sampling technique suitable for analyzing data with small samples. The distribution of indirect effects was obtained through multiple sampling, and to enhance the statistical stability of the findings, we combined the samples and tested the 95% confidence interval.

The model investigates three indirect effects using SE, PE, and PU as mediating variables for the D3D user interface. Firstly, the first indirect effect is the role of SE in moderating the relationship between PE and CI (M1: D3D → SE → CI). The second indirect effect was the effect of PE in connecting the D3D user interface and CI (M2: D3D → PE → CI). The third indirect effect was the effect of PU on connecting the D3D user interface and CI (M3: D3D → PU → CI). Finally, the combined effects of SE (M1), PE (M2), and PU (M3) were analyzed and ranked based on CI.

The results of M1 hypothesis testing demonstrated that the D3D user interface was positively related to CI (β = 0.884, p < 0.05), and CI was associated with SE (β = 2.058, p < 0.05). The indirect effect of SE was 0.595, and the significant result (Boot SE = 0.117; 95% Boot CI = [0.121, 0.591]) suggests an important mediating role of SE between the D3D user interface and CI. Thus, M1 is supported. The results of M2 hypothesis testing demonstrated that the D3D user interface was positively related to CI (β = 0.884, p < 0.05), and CI was associated with PE (β = 1.303, p < 0.05). The indirect effect of PE was 0.350, and the significant result (Boot SE = 0.066; 95% Boot CI = [0.076, 0.336]) suggests an important mediating role of PE between the D3D user interface and CI. Thus, M2 is supported. The results of M3 hypothesis testing demonstrated that the D3D user interface was positively related to PU (β = 0.884, p < 0.05), and CI was associated with PU (β = 0.738, p < 0.05). The indirect effect of PE was 0.073, which was insignificant (Boot SE = 0.046; 95% Boot CI = [−0.028,−0.147]). Thus, M3 is not supported. The results indicate that SE and PE significantly mediate the relationship between the D3D user interface and CI, confirming that SE (M1) and PE (M2) mediate variables via which the D3D user interface positively influences CI.

As a result of the previous tests, no statistically significant main effects were found for SE and PU in the AVA condition. Therefore, M5 (AVA → SE → CI) and M6 (AVA → PU → CI) were not supported. To test M4 (AVA → PE → CI), the role of PE in moderating the relationship between AVA and CI was assessed using PE as the mediating variable for AVA. The results for M4 showed a positive correlation between AVA and CI (β = 0.931, p < 0.05) and between PE and CI (β = 1.024, p < 0.05). A significant mediation effect of PE was observed (effect = −0.350, Boot SE = 0.053; 95% Boot CI = [0.109, 0.318]), indicating that PE partially mediates the relationship between AVA and CI. Thus, M4 is supported.

This study provided results that validated the hypotheses, as summarized in Table 3. The D3D user interface yielded higher PE, PU, SE, and CI scores for older adults using IVRTA than the S3D user interface. H1 received partial support, and H2 was partially supported, indicating that the D3D user interface reduced the frequency of requesting assistance among older adults. However, no significant difference in usage duration was observed between the D3D and S3D user interfaces. Compared to NAVA, the AVA design led to higher PE and CI scores for older adults; however, no significant differences were found in PU and SE. H3 and H4 received partial support, as the AVA design significantly increased system usage duration among older adults, though it did not notably reduce the frequency of requesting assistance. We found that SE and PU moderated CI in the D3D user interface within RQ2, while PE moderated CI in the AVA design.

Table 3 Hypotheses and results.
Full size table

Discussion

D3D user interfaces: Representation and manipulation

This study incorporates familiarity design approaches—representation (user interface) and manipulation (interface interaction)—into the D3D user interface design to create an IVRTA that simulates realistic scenarios. The D3D user interface supports information presentation and interaction through joystick clicks across multiple interfaces. In contrast, the S3D user interface employs a unified design, with interaction focused primarily on joystick buttons, similar to the IVRTA designs currently on the market.

The findings indicate that older adults perceive IVRTA, developed with a familiarity design, as more user-friendly than prevalent market options. These results are significant, as PE is essential for the long-term adoption of immersive virtual reality technology among older adults (Bandura and Watts, 1996; Lee et al., 2024).

Behavioral outcomes aligned with self-reported outcomes: Older adults using the D3D user interface requested assistance (n = 8) less often than those in the S3D user interface (n = 27). This discrepancy suggests that older adults face more significant challenges with the S3D user interface, corroborating previous findings (Kharoub et al., 2019a, 2019b) that indicate superior performance with the D3D user interface. The experimental results using immersive interactive media in this study provide additional validation for the effectiveness of D3D user interfaces. The D3D user interface optimizes the IVRTA for information representation and interaction dimensions, promoting a friendly transformation of immersive virtual reality technology for older adults. This design approach provides a theoretical foundation and design assistance for the future development of age-appropriate immersive virtual reality user interfaces for older adult users, and it can be considered a tangible application of the SHAFE concept in immersive virtual reality applications.

AVA design: Guidance

We found that information guidance is a key factor in familiarity design. Self-reported results from older adults showed that AVA significantly improved users’ PE and CI but not SE or PU. This finding warrants further exploration. Due to physiological deterioration caused by aging, older adults often face challenges in spatial cognition and verbal memory (Finkel et al., 2007). The design of the AVA, which incorporates voice prompts, facial expressions, and a friendly demeanor, can help enhance the coherence and clarity of content delivery (Carrozzino et al., 2018) while also capturing users’ attention and encouraging them to interact (Guadagno et al., 2007). However, PU, as users’ subjective perception of system benefits and usefulness, mainly relies on the deep assessment of functional value. AVA guidance primarily focuses on simplifying and optimizing the operation process, which makes it difficult to trigger users’ re-perception of functionality directly and thus has a limited impact on PU. In addition, the effects of AVA are more through creating a sense of intimacy and companionship than functional cognitive enhancement (Potdevin et al., 2021), and AVA directly enhances CI by increasing older adults’ emotional engagement (e.g., sense of pleasure, sense of companionship) (Liu et al., 2023). In contrast, the enhancement of SE typically relies on continuous operational practice and the accumulation of repeated successful experiences, which may be one reason why the bootstrapping effect of AVA is not significant in enhancing SE.

Behavioral data further supported the above view that although there was no difference between AVA and NAVA regarding the number of requests for help, the usage duration was significantly longer for older adults in the AVA condition, suggesting a higher level of engagement and persistence in the experience. More critically, the interaction effect analysis revealed that a significant interaction effect was present when AVA was used in conjunction with the D3D user interface, resulting in a further extended usage duration. This result highlights the role of AVA in influencing the use behavior of older adults. AVA makes it easier for older adults to use IVRTA by optimizing the information guidance method. It promotes the acceptance of immersive virtual reality technology among older adults, providing a reference for creating a friendly information environment, as SHAFE advocates.

Interaction between interface design and guidance

In the interaction analysis of interface design and guidance, self-reported data showed no significant effect of PE, PU, SE, and CI for older adults. However, the results of the behavioral data showed an increase in the duration of IVRTA usage. Analyzed from the perspective of multimodal feedback theory (Zou et al., 2025), when the D3D user interface design and AVA conditions work together, the user’s visual and auditory channels receive simultaneous signal inputs, forming a multichannel information integration. In particular, the D3D user interface enhances information acquisition efficiency and task performance by reinforcing spatial structure and visual elements. In contrast, AVA enhances the users’ emotional stability and concentration continuity through natural language cues and emotional support. The two may complement each other at the behavioral level, thus improving operational performance by increasing the usage duration. However, such behavioral synergies may require longer or deeper experience to be reflected in subjective perceptions and may not be immediately reflected in older adults’ cognitive evaluations. Therefore, this study speculates that user interface design and guidance interaction may be mainly reflected in users’ behavioral performance rather than cognitive evaluation. Meanwhile, PE, PU, SE, and CI failed to show significant changes because they mainly reflect users’ subjective cognitive judgment of the system. This finding suggests that although the synergistic optimization of user interface design and guidance could effectively enhance the behavioral performance of older adults, future research should further explore its delayed effects and mediating mechanisms on cognitive and affective assessment variables.

Underlying mechanism

The findings of this study reveal the underlying mechanisms by which information guidance and user interface design influence the CI of older adults when using IVRTA. Previous research on the role of guidance and user interface design in immersive virtual reality applications used by older adults has been limited and sometimes conflicting. This study demonstrates that the AVA design and the D3D user interface significantly boost older adults’ CI using IVRTA, primarily by enhancing PU and increasing CI. In the sequential multiple mediation model, SE and PE emerged as key mediators between the D3D user interface and CI. The D3D user interface has improved older adults’ perception of IVRTA’s usability, which helps enhance their proficiency and confidence when using IVRTA, thereby increasing their CI when using the system. PE also mediates the relationship between the AVA design and CI, suggesting that as older adults find the IVRTA more straightforward, their intention to continue using it grows. In the case of the D3D user interface, SE impacts CI directly and indirectly, affirming the importance of familiarity design in supporting SE. Thus, SE is critical in fostering older adults’ engagement with and continued use of the D3D user interface.

According to the TAM, PU is facilitated by PE (Davis, 1989), with both working together to enhance CI and subsequently influencing user behavior. This study uncovers an interesting distinction: PU directly impacts CI only in the D3D user interface, with no direct or indirect effect on CI in the AVA design. In contrast, PE influences CI directly and indirectly in the D3D user interface and AVA design. This result suggests that familiarity design approaches may substantially impact older adults’ PE compared to their PU. In summary, while PE and PU are crucial, familiarity design approaches appear to be incredibly influential in improving ease of use, a pivotal factor in older adults’ CI when using IVRTA.

Implications for research and practice

Implications for research

This study makes a substantial theoretical contribution to IVRTA and familiarity design research. It is the first to integrate familiarity design with the unique characteristics of immersive virtual reality applications, establishing a familiarity design approach—representation, manipulation, and guidance—specifically for IVRTA targeting older adults. The contributions of this research can be summarised in five main areas: First, this study expands the application range of familiarity design approaches, demonstrating their effective integration into IVRTA through representation, manipulation, and guidance, which significantly enhances the CI for older adults and could improve their user experience when using IVRTA. Specifically, familiarity design approaches enhanced ease of use and interest among older adults. Secondly, unlike prior studies primarily focused on user behavior (Beck et al., 2019; Leung et al., 2023; Baker et al., 2023), this study empirically examines the mechanisms underlying older adults’ CI when using IVRTA. Previous research has explored the intention to use immersive virtual reality applications and IVRTA (Wang et al., 2024; Sinha et al., 2024) but has often overlooked the underlying mechanisms of sustained use. This study provides insight into the mechanisms that foster CI in older adults when using IVRTA, laying the groundwork for future research. Third, this study advances research on IVRTA design for older adults, as IVRTA offers multiple benefits to this demographic (Yu et al., 2020; Fiocco et al., 2021; Kim and Kang, 2023). While existing studies have investigated the effects of immersive virtual tours on older adults (Yu et al., 2020; Fiocco et al., 2021), few studies have specifically addressed IVRTA, which is designed with older adults in mind. This study fills a gap in the current research. Fourth, through a randomized controlled experiment, this study tested the efficacy of familiarity design approaches in IVRTA for older adults. By collecting both self-report and behavioral data, this study mitigated potential biases from social desirability in self-reporting (Calisto and Sarkar, 2024). Including behavioral measurements enhances the study’s objectivity and reliability, providing a valuable reference for future research initiatives. Finally, this study enriches the intelligent aging design theory system of SHAFE. The familiarity design approach extends SHAFE’s ‘intelligent aging’ from the physical environment to the level of the virtual reality interface. This strategy constructs a set of repeatable and verifiable design strategies for immersive virtual reality user interfaces.

Implications for practice

This study provides valuable design approaches for manufacturing firms and developers creating IVRTA for older adults. The findings underscore the significance of familiar design approaches in fostering sustained engagement. Prior research has demonstrated that “representational” and “manipulation” familiarity design approaches significantly improve older adults’ interactions with digital products (Lee et al., 2024). This study further validates these familiarity design approaches in IVRTA, showing that they reduce the frequency of assistance requests from older adults. For future IVRTA designs, it is recommended to focus on clear information presentation and intuitive interactive manipulation, incorporating elements that resonate with older adults rather than catering solely to the preferences of younger users (Lawry et al., 2019; Lee et al., 2024; Zhang et al., 2016). For IVRTA, which involves extensive information and complex operations, presenting information and interactions through a D3D user interface is advisable (Kharoub et al., 2019b). Specifically, the content presentation should ensure that users can access the surrounding content, interactive buttons, and other functional information with only a slight turn of the head without destroying the immersive viewing experience. At the same time, the color scheme and layout of the interface should be consistent with the type of information and functional use to enhance recognition and intuitive operation, thereby optimizing the overall experience for older adults. In addition, considering the physiological characteristics of older adults with weakened eyesight, the interface should be designed with larger font sizes and buttons to fit their reading and operating habits (Chen et al., 2024).

As an information guide, AVA should also be a priority in IVRTA design. AVA design can enhance familiarity for older adults by incorporating visual cues that mirror real-world objects, thus boosting PE and their CI when using IVRTA. Although AVA design is less commonly used in current IVRTA, AVA design can potentially elevate the CI for older adults by offering familiar and practical guidance. In terms of design, the appearance and sound parameters of the AVA should be oriented towards enhancing accessibility, for example, by using the image of a young woman as a reference to improve the accessibility and acceptance of the virtual agent (Esposito et al., 2021). The speech output should be adjustable in both ‘slow’ and ‘high’ volumes to better meet the physiological needs of older adults regarding hearing and processing speed. Future IVRTA design and development should prioritize AVA to ensure a more intuitive and supportive experience for older adults.

Implications for policy

The results of this study validate the effectiveness of the familiarity design approach in creating IVRTA suitable for use by older adults, emphasizing the need for policymakers to address the digital divide issues older adults face and promote inclusive technology development. First, the government should establish special funds to support the research and development of IVRTA specifically designed for older adults. Through incentives such as tax breaks, technology companies should be encouraged to prioritize adopting principles based on familiarity design (such as using D3D user interfaces and AVA) in their products. Second, regulators should collaborate with industry experts to develop accessibility standards for virtual reality applications tailored to older adults. The standards should require intuitive interface designs (e.g., layered information displays) and anthropomorphic guidance systems to minimize cognitive load and improve usability. Third, policymakers should fund a national digital literacy training program in IVRTA. Training modules must focus on teaching practical skills in D3D user interfaces and AVA to help older adults use IVRTA independently. Fourth, academia, industry, and non-profit organizations are encouraged to collaborate in designing age-friendly IVRTA solutions. Such collaboration could effectively integrate public resources and expertise to address the unique needs of an aging society.

Limitations and future research perspectives

This study has certain limitations. First, although the power analyses support the study’s ability to detect moderate and above effects, the small sample size may still affect the interpretation of weak effects. In addition, the uneven sample size between groups may have introduced bias. Larger samples (greater than or equal to 30 per group) and pre-registration designs are recommended for future studies to improve reproducibility. Second, although this study confirmed that the AVA significantly enhanced CI among older adults, it has not yet explored which bootstrap design had the greater impact on their experience. Subsequent studies could adopt multimodal guidance, focusing on the crucial roles of voice and visual guidance in emotional arousal and technology acceptance, and explore in depth the path of influence of multimodal design on older users in IVRTA from the perspective of multisensory integration. Thirdly, the sample size of this study consisted of 86 participants, selected through random sampling, with the majority from urban residential communities. Although it represents the characteristics of highly active older users in cities to some extent, its representativeness in the overall population remains limited. Future research should expand the sample size and focus on the diversity and heterogeneity of sample sources to enhance the study’s external validity and generalization value. Fourth, although ‘the number of smartphone applications used by older adults in the preceding week’ has some explanatory power as a covariate, its coverage of digital literacy is limited, especially the lack of targeted measurements of adaptability to emerging technologies (e.g., virtual reality). Future research should develop a multidimensional digital literacy scale for older Chinese adults, incorporating IVRTA data (e.g., task completion time, error rate) to control the impact of technology familiarity more accurately. Fifth, this study mainly started from the perspective of technology influence, integrating TAM and SET to explore whether the familiarity design approach significantly affects older adults’ CI when using IVRTA through PE, PU, and SE. Although a more comprehensive theoretical framework has been established, the multidimensional influences on the process of CI formation have not yet been fully explored. Future studies can incorporate additional psychosocial variables (e.g., social support, technology anxiety, value of use recognition, emotional arousal) to construct a more comprehensive, multi-path structural model, thereby enhancing the explanatory power and applicability of the model.

Conclusion

This study highlights the effectiveness of the familiarity design in enhancing the experience and CI of older adults when using IVRTA. Prioritizing a D3D user interface in representation and manipulation design is advisable, as it notably influences older adults’ PE, PU, and CI when using IVRTA. In information guidance design, prioritizing the use of AVA design is advisable. Providing suitable guidance for older adults in unfamiliar virtual environments is crucial, as it facilitates their effective and efficient use of IVRTA while enhancing their motivation to engage with it. The IVRTA design should prioritize older adults’ cognitive abilities and usage experiences rather than necessitating their adaptation to the technology. Applying familiarity design to the design of IVRTA will help older adults overcome digital barriers, enabling them to enjoy the rich experiences that IVRTA provides. This will improve the quality of their later life and contribute to the overall development of the IVRTA industry and the establishment of SHAFE.