1 Introduction

In the sport domain, new technologies are increasingly used by coaches and/or athletes (Greenhough et al. 2021). They may contribute to develop individual sport practice (Angosto et al. 2020) and they also may improve the monitoring of athletes through the acquisition of precise data useful to better understand and to increase sport performance (Malone et al. 2017). Among the numerous existing technologies, those concerning virtual reality (VR) are used more and more regularly in the sport domain (Le Noury et al. 2022; Wood et al. 2021), while they have been used for much longer in other training domains, such as first responder training, medical training, military training, workforce training, and education. In such training domains (Xie et al. 2021), VR training may overcome real-world training limitations (e.g., human and temporal costs, safety issues) and has its own specific benefits (e.g., increasing the number of training scenarios, training anywhere, minimal exposure to dangerous situations, appropriate feedback display). For instance, the systematic review of Steen et al. (2024) highlighted that VR training interventions can promote knowledge and skills acquisition in the assessment and treatment of mental health disorders, while regretting that intervention studies with immersive VR applications are lacking.

Recent advances in VR devices, including those of the most immersive devices such as VR Head-Mounted Display (VR-HMD) and Computer Automatic Virtual Environment (CAVE), have enabled the use by coaches and athletes of several VR applications designed to enhance sport performance (Neumann et al. 2018). By simulating sport specific movements and by providing benefits to coaches, VR advancements may change the way coaches focus on developing skills required for improving athletes sport performance (Farley et al. 2020). Coaches may indeed provide training sessions for their athletes, exposing them to fully controlled and reproducible training sessions, by immersing them in virtual environments. VR training modalities may moreover be adapted to the specificities of each sport, to the problems encountered, and to the needs and objectives of coaches. For instance (Düking et al. 2018; Farley et al. 2020; Le Noury et al. 2022; Neumann et al. 2018), VR allows (1) to enable athletes to experience situations in VR that are impossible when training in real life (e.g., repeating exactly the same situation over and over again, under strictly identical conditions from one trial to the next, increase the complexity of the stimuli such as the speed of the opponent), (2) to carry out tasks focusing on specific areas related to sport performance (e.g., perception, decision-making, reaction time), (3) to face a virtual opponent and/or to cooperate with a virtual partner which performance is controlled, (4) to benefit from objective feedback in real time and/or delayed time on the relevance of choices and actions made by athletes, and/or (5) to view techniques performed by experts (or by themselves), while controlling the viewing angle. Of course, training with VR is not intended to replace real-life training. As a complement to regular training, each VR protocol will enable coaches and athletes to work on specific aspects of the athlete’s performance. For instance, VR was used with high-level boxers and athletes in the 4 × 100 m relay to optimize their anticipation skills during the training period preceding the Paris 2024 Olympic Games (Montagne et al. 2024).

A growing number of studies have demonstrated the effectiveness of VR training programs in sport (for reviews, see Faure et al. 2020; Frank 2020; Le Noury et al. 2022; Neumann et al. 2018; Stone et al. 2018), their potential to increase real-world sport performance (for a review, see Michalski et al. 2019), but also their limitations (e.g., Düking et al. 2018; Farley et al. 2020). Fewer studies have focused on the psychological consequences of VR applied to the sport domain such as simulation workload (Harris et al. 2020) and self-efficacy (Frank et al. 2023). Only two studies have investigated acceptance of VR devices intended to enhance sport performance (Greenhough et al. 2021; Mascret et al. 2022) and neither of them has examined their acceptance by coaches from different sports and different levels of competition. However, even in cases where athletes accept VR themselves, acceptance of VR by coaches seems to be an inevitable prerequisite for its ultimate use in athletes’ training sessions.

Acceptance may be defined as the tendency to be more or less favourable to the use of a technology before or/and after use (Davis 1989; Davis et al. 1992). The process of technology adoption is subdivided into two sub-processes that reflect the temporality of the interaction between an individual and a technology (Alexandre et al. 2018): acceptance before use is an a priori phenomenon, a more or less positive representation of the technology without having used it yet, while acceptance after use is an a posteriori pragmatic evaluation, involving actual prior use of the technology. Acceptance before use and acceptance after use are included in a process influencing final adoption of the technology. In the present study, we focused on acceptance of VR before a first use, sometimes called acceptability (Alexandre et al. 2018). While it may seem surprising to investigate technology acceptance before use, an a priori measurement can on the contrary contribute to a better understanding of the first step leading to the actual deployment of this technology (Venkatesh and Bala 2008). It allows to potentially identify initial blockages and the psychological factors that may cause a rejection of the technology without ever having tried it (Mascret et al. 2022).

Several theoretical models have investigated technology acceptance such as the Technology Acceptance Model (TAM, Davis 1989; Venkatesh and Davis 2000) and the Unified Theory of Acceptance and Use of Technology (UTAUT, Venkatesh et al. 2003). The TAM is the most widely used model (Scherer et al. 2019), based on two main variables in its initial version. Perceived usefulness (i.e., the perceived usefulness of a technology to improve performance) and perceived ease of use (i.e., the degree of ease of use of a technology) are positive predictors of intention to use a technology which, finally, predicts its actual use (Davis 1989; Venkatesh and Bala 2008). In the field of teaching, Mayer and Girwidz (2019) have recently subdivided perceived usefulness into two dimensions (perceived usefulness for teaching and perceived usefulness for students). This distinction seems equally relevant in the field of sport training and the present study distinguishes perceived usefulness for coaching (i.e., the technology may improve the training sessions themselves, for example by improving organization, implementation, group management, and individualization) and perceived usefulness for athletes (i.e., the technology may enhance athletes’ performance, assessed by their results in the field). If perceived usefulness for coaching and perceived usefulness for athletes are effectively related, they are considered two different and complementary constructs. Finally, perceived enjoyment (i.e., the degree of pleasure derived from using a technology) was added in the model as a positive predictor of intention to use this technology (van der Heijden 2004). Initially, attitude towards technology was a variable included in the model (between intention to use and perceived usefulness / perceived ease of use). But for several years now, most studies have agreed that it is no longer necessary to include this variable in the TAM (for a review, see Turner et al. 2010).

In the TAM literature, external variables may also be added to the initial model to increase its explanatory power. When looking at the acceptance of a technology in the workplace, one of the most interesting external variables is job relevance, which is the individual’s perception of the degree of applicability of the technology to his/her work (Venkatesh and Davis 2000). Job relevance is recognized as a positive predictor of perceived usefulness: the higher the perceived applicability of the system to the individual’s work, the more likely the individual will find the technology useful (Venkatesh and Davis 2000). In the present study, coaches were asked about the relevance of VR to the requirements and aims of their coaching job. All the relationships between the TAM variables used in the present study (perceived usefulness for coaching, perceived usefulness for athletes, perceived enjoyment, perceived ease of use, job relevance, intention to use) are presented in Fig. 1.

Fig. 1
figure 1

The hypothesized model. Notes H* + (e.g., H1 +, H2 +…) means that we assume that the first variable positively predicts the second one (e.g., H1 + indicates that we hypothesized perceived usefulness for coaching to positively predict intention to use). Arrows in light grey represent relationships between variables that need to be included in the model to comply with the TAM variable relationship specifications, but they do not represent hypotheses for the present study

Full size image

The TAM has been used in many domains (e.g., education, business, healthcare), with different technologies (e.g., mobile apps, smartwatches, robots), different purposes (e.g., learning, health, gaming), and different users (e.g., students, teachers, older adults). It has also been applied to the sport domain to investigate, for instance, acceptance of running on a social exercise platform (Tsai et al. 2021), sports wearable technology (Kim and Chiu 2019), and fitness applications (Angosto et al. 2020). Investigating acceptance of VR was also conducted in numerous studies based on the TAM such as acceptance of VR for business (Manis and Choi 2019), for an aeronautical assembly task (Sagnier et al. 2020), and for preventing older adults’ falls (Mascret et al. 2020). In their SWOT analysis, Düking et al. (2018) highlighted that non-acceptance of VR was considered a threat to the use of this type of device in the sport domain. But surprisingly, studies investigating acceptance of immersive VR (i.e., VR-HMD, CAVE) in the sport domain are scarce (e.g., Gradl et al. 2016; Mahalil et al. 2020), and studies using technology acceptance models (TAM or UTAUT), identifying the relationships between the different variables, are even rarer (Montagne et al. 2024). There are only four of them, focusing on the acceptance of the VR-HMD, which is a device that combines several features that contribute to its current success: effectiveness for sports performance, easily accessible, reasonable price, easily transportable (Mascret et al. 2022). Acceptance of the VR-HMD for enhancing the immersive viewer experience (e.g., when watching a match in his/her living room) has been examined in two studies (Kunz and Santomier 2019; Rynarzewska 2018) but has nothing to do with the use of this device to improve sport performance. Based on the TAM, another study has investigated the acceptance, before use, of VR-HMD intended to enhance sport performance among 1162 athletes of different sport levels (from recreational to international level) in different sports (Mascret et al. 2022), highlighting that the VR-HMD was rather well accepted by athletes before a first use and that no initial psychological blockages appeared. But most athletes felt that people around them, including their coaches, would tend to discourage them from using the VR-HMD, except for international athletes who felt that people around them would neither discourage nor encourage them to use it. However, if the VR-HMD is effective in improving athletes’ performance, if it is well accepted by athletes, but not accepted by their own coaches, there is a strong likelihood that it will not be used at all in the end. Only one study, based on a modified version of the UTAUT, examined acceptance of VR in professional soccer among coaches, support staff, and players (Greenhough et al. 2021). The results showed that performance expectancy (a construct close to perceived usefulness) was the strongest predictor of intention to use VR in professional soccer and that VR may be used in performance analysis, in rehabilitation settings, and to improve tactical awareness, despite of several barriers (monetary cost, coach buy-in, lack of scientific evidence). However, this study did not specifically focus on coaches, who ultimately represent 38 participants out of the 207 participants (19 strength and conditioning coaches, 13 technical coaches, and 6 head coaches). Consequently, no study has currently validated a theoretical model of acceptance specifically focusing on coaches’ acceptance of a VR-HMD intended to enhance sport performance.

Moreover, the Greenhough et al.’s (2021) study was conducted on a single sport (soccer) and on a single level (professional). Since some differences of VR-HMD acceptance were found among athletes from between sports (e.g., perceived usefulness and intention to use was lower in swimming or in cycling, Mascret et al. 2022), it was therefore worthwhile soliciting a sample of coaches from different sports. In the last study, no significant effects of athletes’ sport level were found on perceived usefulness, perceived ease of use, perceived enjoyment, and intention to use the VR-HMD. It was relevant to examine if this pattern was repeated among coaches because elite and high-level coaches are constantly looking for ways to improve their training and their athletes’ performance. Consequently, including coaches from different sports and from different levels of competition may provide a more accurate view of coaches’ acceptance of VR-HMD.

The first aim of the present study was to test the TAM with coaches to investigate their acceptance, before a first use of a VR-HMD intended to increase sport performance. Based on the TAM literature, we hypothesized that perceived usefulness for coaching (H1), perceived enjoyment (H2), perceived ease of use (H3), and perceived usefulness for athletes (H4) would be significant positive predictors of coaches’ intention to use the VR-HMD. We also hypothesized that job relevance would significantly and positively predict perceived usefulness for coaching (H5), perceived usefulness for athletes (H6), and intention to use the VR-HMD (H7). All these hypotheses are presented in Fig. 1.

The second aim of the present study was to examine the coaches’ level of acceptance of the VR-HMD. Based on the only study conducted with coaches (Greenhough et al. 2021) and on a procedure for comparing scores with the mean of the Likert scale usually used in studies based on the TAM (e.g., Delbes et al. 2022), we hypothesized (H8) that coaches’ scores of perceived usefulness for coaching, perceived usefulness for athletes, perceived enjoyment, perceived ease of use, job relevance, and intention to use would be significantly higher than the mean of the Likert scale, indicating that the VR-HMD was rather well-accepted by coaches. But the previous study focused on professional soccer only and the study conducted with athletes from different sports and different levels of competition (Mascret et al. 2022) showed that athletes felt that people around them, including their coaches, would tend to discourage them from using the VR-HMD, except for international athletes (Mascret et al. 2022). Consequently, hypothesis H8 was formulated with caution and was particularly in need of testing.

The third aim of the present study was to compare coaches’ acceptance according to their level of competition. Since the possibilities for coaches to improve the performance of athletes are becoming more and more scarce as their level of competition increases, we hypothesized (H9) that coaches’ scores of perceived usefulness for coaching, perceived usefulness for athletes, intention to use the VR-HMD, and job relevance would be higher for international and national coaches than for departmental-regional coaches. We also hypothesized (H10) that no significant differences on perceived enjoyment and perceived ease of use would appear because these two variables are directly related to the VR-HMD itself as a device and do not depend on the coaches’ level of competition.

2 Methods

2.1 Participants and procedure

The study sample included 239 French coaches (48 women, 191 men) aged 40.1 ± 11.7 years old (range: 18–78). Participants needed to meet two inclusion criteria to be included in the present study: being of legal age and training competitive athletes over the age of 15 years old. Training competitive athletes was required as the VR-HMD was specifically dedicated to improving sports performance. The usual French classification was used to categorise the self-reported competition levels of trained athletes, completed by the number of years of coaching experience: departmental-regional level (93 participants, 15.1 ± 9.7 years of coaching), national level (89 participants, 15.1 ± 10.2 years of coaching), and international level (57 participants, 16.8 ± 9.0 years of coaching). Supplementary Material 1 lists all the descriptive statistics related to the level of competition. Several sports were represented among the participants. The most represented were soccer (37 participants), gymnastics (32 participants), handball (27 participants), athletics (18 participants), basketball (17 participants), and rugby (15 participants). A total of 126 coaches train athletes between the ages of 15 and 18, 108 train athletes between the ages of 19 and 35, and 5 train athletes over 36. A total of 221 coaches has a recognized coaching diploma, while 20 have no diploma.

The data was collected using Framaforms, an open-source software package that enables online surveys to be conducted anonymously and securely. The questionnaire was distributed online via social networks (e.g., Facebook, LinkedIn, Instagram), sports clubs, and French national sports federations using a nonprobability snowball sampling (Kosinski et al. 2015). The questionnaire was shared on social networks via personal, laboratory and sports institution pages. The experimenters also met coaches to complete the questionnaire in paper or dematerialised (using a QR code) formats. Before completing the questionnaire, participants were asked to read a short text describing the possible uses of a VR-HMD in sports training and its relevance to improving sport performance. This procedure is commonly used in studies investigating the technology acceptance before a first use (e.g., Li et al. 2019; Mascret et al. 2022). Participation was strictly anonymous, and respondents gave informed consent electronically, or on paper where appropriate. The study was approved by the National Ethics Committee for Research in Sports Sciences (CERSTAPS IRB00012476-2021-16-04-103).

2.2 Measures

The different dimensions of acceptance of the VR-HMD by coaches were assessed by three items per dimension: perceived usefulness for coaching (e.g., “I think this VR-HMD would be useful to improve my training sessions”), perceived usefulness for athletes (e.g., “I think that the use of this VR-HMD would be useful to improve the performance of the athletes I train”), perceived ease of use (e.g., “I think the use of this VR-HMD would be simple”), perceived enjoyment (e.g., “Using this VR-HMD in my training sessions would be fun”), and intention to use (e.g., “If I had easy access to this VR-HMD, I would probably use it in my training sessions”). An external variable, job relevance (Alharbi and Drew 2014; Hogan et al. 2020), was assessed using three items (e.g., “In my work as a coach, the usage of a VR-HMD would be appropriate”).

Participants were asked to respond to each item using a Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). Items of perceived usefulness for coaching and perceived usefulness for athletes were adapted for coaches from the study conducted by Mayer and Girwidz (2019) to better distinguish two types of perceived usefulness of the VR-HMD (one type oriented towards improving the activity of coaches itself and one type oriented towards improving athletes’ performance). Internal consistency was measured using McDonald's omegas (Dunn et al. 2014), which showed satisfactory scores (0.796 ≤ Ω ≤ 0.944). Table 1 provides descriptive statistics and McDonald’s omegas for each variable.

Table 1 Descriptive statistics, skewness, kurtosis, and internal consistency
Full size table

Demographic variables were also self-reported: gender, age, level of education, and frequency of VR-HMD use (from never to several times a week on a six-point Likert scale). Frequency of VR-HMD use assessed the participants’ prior use of a VR-HMD, focusing on regular use (e.g., playing video games) and not specifically on improving sport performance. Finally, participants were asked to indicate the sport they coach, the level of competition of the athletes they coach (departmental-regional, national, international), and their seniority as a coach in their sport (in years). The questionnaires (French version) are provided in Supplementary Material 2.

2.3 Data analyses

The JASP (Jeffreys’s Amazing Statistics Program) software (version 0.17.1) with a level of significance defined at p < 0.05 was used to conduct all the data analyses. Gross outliers were detected by using the Mahalanobis distance (χ2(9) = 27.88, p < 0.001) at the multivariate level (In’nami and Koizumi 2013). Values ≥|2| for skewness and ≥|7| for kurtosis indicated variables non-normal in distribution (Curran et al. 1996). A confirmatory factor analysis (CFA) was conducted to evaluate the model fit using several fit indices (Byrne 2010; Hu and Bentler 1999): the χ2/df ratio (with a value ≤ 3), the Comparative Fit Index (CFI ≥ 0.90), the Tucker–Lewis Index (TLI ≥ 0.90), the Root Mean Square Error of Approximation (RMSEA ≤ 0.08), and the Standardized Root Mean Square Residual (SRMR ≤ 0.08). Three main analyses were then conducted.

First, the hypotheses provided in Fig. 1 were tested by using a structural equation modelling (SEM) analysis with the maximum likelihood estimation (Everitt and Hothorn 2011). The model was parameterized by four control variables (gender, age, level of education, frequency of VR-HMD use).

Second, six consecutive one-sample t-tests were conducted to highlight potential differences between the mean of the Likert scale (i.e., 3) and the mean scores of the TAM variables (perceived usefulness for coaching, perceived usefulness for athletes, perceived ease of use, perceived enjoyment, intention to use, and job relevance). Effect sizes were calculated by using Cohen’s d. The same analyses were also conducted with each subsample (departmental-regional, national, and international levels).

Third, six consecutive one-way ANOVAs were used to examine potential differences between coaching levels (departmental-regional, national, and international levels) in the scores of the TAM variables. Post-hoc Tukey tests were used if significant differences appeared. Partial eta-squared (ηp2) was used to calculate effect sizes.

3 Results

3.1 Preliminary results

As each item had to be answered, no missing data were identified. Seven participants were excluded from the statistical analysis: four participants because they were coaching recreational sports which is not considered competitive, one participant because he was coaching blind soccer, which is a sport logically unsuited to the use of a VR-HMD for training, and two participants because they were initially identified as outliers using the Mahalanobis distance. The distribution of the data was considered normal based on the values obtained from the tests of Skewness (− 0.774 ≤ Skewness ≤ − 0.081) and Kurtosis (− 0.776 ≤ Kurtosis ≤ 0.283). Table 1 presented the values of Skewness and Kurtosis for each variable.

Most of the coaches had never used a VR-HMD before (63.6% of the coaches), while others had used one only once (18%) or a few times in their lifetime (13.4%). Very few uses one several times a year (3.8%) or several times a month (1.2%).

3.2 Validation of the tested model

Based on the results of a CFA conducted on the covariance matrix of the items, the hypothesized six-factor model was supported (χ2(120, N = 239) = 180.78, p < 0.001, χ2/df = 1.51, CFI = 0.999, TLI = 0.999, RMSEA = 0.046, SRMR = 0.048). Based on the results of the SEM analysis, the model fit met the expected requirements (χ2(2, N = 239) = 4.06, p = 0.131, χ2/df = 2.03, CFI = 0.998, TLI = 0.985, RMSEA = 0.066, SRMR = 0.011).

The SEM analysis also highlighted that perceived usefulness for coaching, perceived usefulness for athletes, and perceived enjoyment were positive predictors of coaches’ intention to use the VR-HMD (p < 0.001), validating hypotheses H1, H2, and H4. The more the coaches found the VR-HMD useful for improving their training, useful for improving their athletes’ performance, and enjoyable, the more they intended to use it in their training sessions. Contrary to hypothesis H3, perceived ease of use was not found to positively predict intention to use (p = 0.527), indicating that coaches’ intention to use the VR-HMD was not dependent on its perceived ease of use. This result is surprising and needs to be discussed. Moreover, job relevance significantly and positively predicted perceived usefulness for coaching, perceived usefulness for athletes, and intention to use the VR-HMD (p < 0.001), validating hypotheses H5, H6, and H7. The more the coaches found the VR-HMD relevant for their own job, the more they found it useful for coaching and for improving performance, and the more they intended to use it.

Ancillary analyses showed that perceived usefulness for coaching was positively predicted by perceived usefulness for athletes (p < 0.001), perceived enjoyment (p = 0.028), and age (p = 0.013), but not by perceived ease of use (p = 0.888). Perceived usefulness for athletes was not significantly predicted by perceived ease of use (p = 0.086) and perceived enjoyment (p = 0.324). Except for age, all the other control variables were not found to be significant predictors of any previous variables. Consequently, they were deleted for the sake of parsimony (Zigarmi et al. 2018). The final model, presented in Fig. 2, predicted 80.9% of the variance of coaches’ intention to use the VR-HMD.

Fig. 2
figure 2

Validated structural model with standardized path coefficients. The p-values are indicated by * according to the following nomenclature: * p <.05, ***p <.001. Note The dotted arrows correspond to non-significant predictions. Arrows and results in light grey represent relationships between variables that need to be included in the model but do not fit the main hypotheses of the present study

Full size image

3.3 Coaches’ acceptance of the VR-HMD

The results of one-sample t-tests conducted on the whole sample highlighted that perceived usefulness for coaching (M = 3.255, SD = 1.001, p < 0.001), perceived usefulness for athletes (M = 3.484, SD = 1.073, p < 0.001), perceived ease of use (M = 3.376, SD = 0.858, p < 0.001), perceived enjoyment (M = 3.798, SD = 0.880, p < 0.001), intention to use (M = 3.523, SD = 1.156, p < 0.001), and job relevance (M = 3.221, SD = 1.109, p = 0.002) were significantly higher than the mean of the Likert scale. All these results showed that the VR-HMD was well-accepted by coaches and that they intended to use it in their training sessions, validating hypothesis H8. Supplementary Material 3 presents the detailed results of these analyses.

3.4 Influence of coaching level on VR-HMD acceptance

First, the results of one-sample t-tests conducted on each subsample (departmental-regional, national, international) showed that (a) perceived usefulness for coaching was significantly higher than the mean of the Likert scale for coaches who trained national (p = 0.001) and international (p = 0.015) athletes, but not for coaches of the departmental-regional level (p = 0.265); (b) perceived usefulness for athletes (PUA), perceived ease of use (PEOU), and perceived enjoyment (PE) were significantly higher than the mean of the scale for coaches who trained departmental-regional (all ps < 0.001), national (pPUA < 0.001; pPEOU = 0.001; pPE < 0.001), and international (pPUA = 0.002; pPEOU < 0.001; pPE < 0.001) athletes; (c) job relevance was significantly higher than the mean of the scale for coaches of national (p = 0.001) and international (p = 0.004) athletes, but not for coaches who trained departmental-regional athletes (p = 0.524); and (d) intention to use was significantly higher than the mean of the scale for all coaches, regardless of the level of the athletes they were coaching (pdep.reg. < 0.001; pnat. < 0.001; pint. = 0.001). Supplementary Material 4 presents the detailed results of these analyses.

Secondly, the one-way ANOVAs showed that the coaching level had a significant effect on job relevance (F(2, 236) = 5.747, p = 0.004, ηp2 = 0.046) but not on perceived usefulness for coaching (p = 0.230), perceived usefulness for athletes (p = 0.610), perceived ease of use (p = 0.333), perceived enjoyment (p = 0.832), and intention to use the VR-HMD (p = 0.676). Post-hoc Tukey tests conducted on the job relevance scores showed that (a) the departmental-regional coaches had significantly lower scores than national (p = 0.023) and international (p = 0.007) coaches; and (b) there was no significant difference between coaches of national and international levels (p = 0.762). National and international coaches considered the VR-HMD more relevant for their job than departmental-regional coaches. Hypothesis H9 was validated for job relevance only, and hypothesis H10 was fully validated. Supplementary Material 3 presents the detailed results of the ANOVAs. Figure 3 presents the results for the full sample and the results according to the sport level.

Fig. 3
figure 3

Comparison with the mean of the scale for each variable of the TAM. Notes The dotted line represents the mean of the Likert scale (i.e., 3): the present study examines with one-sample t tests potential differences between this value and the mean scores of each TAM variable. The error bars represent the 95% Confidence Interval. *p <.05, **p <.01, ***p <.001, n.s. = non-significant

Full size image

4 Discussion

The main purpose of the present study was to examine, before a first use of VR-HMD, coaches’ acceptance of this technology intended to enhance sport performance, to highlight potential initial psychological blockages. In the present study, these psychological blockages, investigated through the self-reported variables of the TAM, are related to the psychological factors that may cause a rejection of the technology without ever having tried it. If these psychological blockages were to appear in the scores of one or more of the TAM variables examined in our study, the concrete consequences could be immediate: There would be a very real risk that the VR-HMD would not be used by coaches, or even tried out, despite its potential benefits for sports performance (Montagne et al. 2024).

First, perceived usefulness for coaching, perceived usefulness for athletes, and perceived enjoyment were found to be positive predictors of coaches’ intention to use the VR-HMD. The more the coaches found the VR-HMD useful for improving their training, useful for improving their athletes’ performance, and enjoyable, the more they intended to use it in their training sessions. These results were in line with those of other studies investigating technology acceptance in general (e.g., Mayer and Girwidz 2019; Venkatesh and Bala 2008), acceptance of immersive VR in general (e.g., Manis and Choi 2019), and acceptance of immersive VR by athletes (Mascret et al. 2022). In the TAM literature, perceived usefulness is the strongest predictor of intention to use a technology (Davis et al. 1992; Mayer and Girwidz 2019). It is also the case in the TAM literature focusing on technology acceptance in the sport domain (Angosto et al. 2020).

Contrary to our hypothesis, perceived ease of use did not positively predict coaches’ intention to use the VR-HMD in their training sessions. In other terms, coaches’ intention to use the VR-HMD was not dependent on its perceived ease of use. Although the absence of a significant relationship between perceived ease of use and intention to use is not the most common finding in the TAM literature, two main explanations usually emerge when this result appears. The first is that the technology studied is already widespread and well-established in a certain context (e.g., cell phone of university students). Because VR is not yet widespread and well-established in sport training, this first explanation does not seem acceptable in the present study. The second explanation seems more plausible. Venkatesh et al. (2003) have shown that perceived usefulness is the strongest predictor of intention to use a technology, and in some cases, it is so strong that it can overwhelm other variables in the model, notably perceived ease of use. Consequently, perceived ease of use does not seem to be an essential criterion for sport coaches, that mainly evaluates the intention to use VR in the light of its perceived usefulness in improving their athletes’ performance and their own coaching (Farley et al. 2020; Montagne et al. 2024).

Two complementary explanations for this non-expected result can also be discussed. First, the present study did not examine whether the coaches’ prior experiences with technology in general might have influenced their perceptions of ease of use, when it could have potentially helped to better understand this result. But coaches’ frequency of VR-HMD use (from never to several times a week) was integrated in the present study as a control variable. But this control variable was not found to be a significant predictor of any TAM variables and, consequently, was deleted for the sake of parsimony (Zigarmi et al. 2018). Secondly, Sprenger and Schwaninger (2023) have shown that the task needs to be intrinsically connected to the technology to find a positive influence of perceived ease of use on intention to use the technology. In the present study, the text read before completing the questionnaire presented VR in a global way, and therefore did not present sport-specific VR tasks. Consequently, the connection between the task and the technology (Sprenger and Schwaninger 2023) was not made explicit enough for coaches, through VR applications that are adapted to the specificities, requirements, and constraints of each sport, also explaining why perceived ease of use did not predict coaches’ intention to use the VR-HMD in the present study.

Finally, job relevance was investigated as an external variable which may predict perceived usefulness for coaching, perceived usefulness for athletes, and intention to use the VR-HMD. The results highlighted that the more the coaches found the VR-HMD relevant for their own job, the more they found it useful for coaching and for improving performance, and the more they intended to use it. This result was not surprising because job relevance is a strong predictor of perceived usefulness when the technology enhances work effectiveness in an accurate, understandable, and effective manner (Hong et al. 2021). Moreover, the more an individual considers the suitability of a technology for his/her work, the more he/she accepts it and has the intention to use it (Ketikidis et al. 2012), which reinforces the importance of showing end-users (i.e., the coaches in the present study) the added value of a technology in relation to their work, even before they use it, to increase the likelihood that they will use it in the end.

The results also clearly showed that the VR-HMD was rather well accepted by coaches before a first use. On average, they found it useful for coaching, useful for their athletes, easy to use, enjoyable, relevant for their job, and they had the intention to use it in their training sessions. These results were in line with those of Greenhough et al. (2021) conducted with professional coaches in soccer, but the present study extends these results to coaches from different sports and different levels of competition. The present study also addresses a concern identified by Mascret et al. (2022), whose results had shown that athletes mainly thought that people around them would not encourage them to use the VR-HMD, even though the athletes themselves intended to use it. The results of the present study seem to show that this fear expressed by the athletes was unfounded and that coaches accept the VR-HMD intended to enhance sport performance. However, perceived usefulness for coaching and job relevance were not found to be significantly different from the mean of the scale, identifying that the VR-HMD was neither useful and relevant, nor useless and irrelevant, for departmental-regional coaches. Since it is more difficult to achieve incremental improvements at the highest performance levels, elite and high-level coaches are constantly looking for innovative ways to improve their training and their athletes’ performance, such as VR-HMD, which is not automatically necessary for coaches training athletes at a departmental-regional level.

The present study has its limitations, but these do offer promising prospects for future studies. First, coaches’ responses may have been subject to social desirability (Phillips and Clancy 1972) because their responses were self-reported. However, this potential effect remains limited, since the responses were anonymous, the questionnaire did not deal with a highly socially connoted theme, and the scores of the responses were significantly higher than the scale average but were not at the very top of the Likert scale either. Moreover, even if reading a text (with or without pictures) is a procedure classically used in studies with the TAM which focus on technology acceptance before a first effective use (e.g., Li et al. 2019; Mascret et al. 2022), a potential priming effect of the text cannot be totally ruled out. To control for this effect in the present study, the text was factual and descriptive, and the two photos were neutral, representing the use of a VR-HMD in American football, while none of the coaches participating in the study belonged to this sport. Secondly, due to a method based on a nonprobability snowball sampling (Kosinski et al. 2015), the number of coaches per sport varied widely, making it impossible to compare acceptance of the VR-HMD among coaches in different sports. To mitigate the limitations of snowball sampling and to overcome potential bias in future studies, a more systematic sampling approach could be conducted with targeting coaches from a wider range of sports and competitive levels. Moreover, since answers to the online questionnaire were anonymous, we cannot state with certainty that each participant was a coach. Thirdly, the present study focused on the initial psychological blockages, examined through the TAM variables only. A broader comprehensive assessment of initial blockages may be conducted in future studies, including other variables such as cost, accessibility, social isolation, or safety (Düking et al. 2018). Finally, the present study was conducted to investigate coaches’ acceptance of the VR-HMD prior to first use to highlight potential initial psychological blockages among this specific population. In future studies, it would be interesting to see how scores of the various TAM variables evolve after an effective test of the VR-HMD by coaches, but also after a longer use as part of a longitudinal follow-up throughout a training period. The comparison of acceptance before use and acceptance after use is classically used in the literature based on the TAM (e.g., Venkatesh and Bala 2008). However, this perspective necessarily requires the creation of VR applications that are adapted to the specificities, requirements, and constraints of each sport, so that VR training enables effective and specific transformations to be envisaged in athletes (Le Noury et al. 2022; Wood et al. 2021).

5 Conclusion

Based on the TAM, the present study aimed to identify possible initial psychological blockages by measuring coaches’ acceptance of VR-HMD prior to first use. The results showed that VR-HMD was fairly well accepted by coaches, even before it is used for the first time. From a technology acceptance point of view, there seems to be no initial contraindication to the use of VR-HMD by coaches for sport performance optimization purposes. A practical application of the results of the present study is to consider that field measurement of VR acceptance before use is relevant to identify any initial blockages that might arise in some coaches, and therefore lead to VR not being used by them despite all the potential of this technology for sport training and improving sport performance. Future research directions may be considered, particularly those focusing on coaches’ VR acceptance after its effective use, acceptance of sport-specific VR applications, longitudinal examination of VR acceptance throughout a training season, and the relationships between coaches’ VR acceptance and sport performance of their teams and/or athletes. These lines of research promise to provide a better understanding of the psychological processes involved in the adoption of technology by coaches.