Sunshine volunteers helped students with illegal drug use problem in schools. Each city and county had a volunteer group. After we sent an invitation letter to each of the 17 Sunshine volunteer groups, 5 of them agreed to participate. We invited all volunteers in the 5 groups (approximately 200) to participate in our study and randomized them into the experimental and control groups by assigning a number and flipping a coin. The selection criteria of participants were (1) helping at least 1 student with drug use problems, (2) having served for less than 2 years as a Sunshine volunteer, (3) having experience in using technology products, and (4) being able to operate SVVR. Recognizing service experience as a confounding factor in our sampling, we selected only the volunteers with less than 2 years of experience to minimize its effect. Training these novice volunteers was expected to produce greater improvements than enlisting experienced volunteers. There was no cybersickness incident that led to the exclusion of participants. [25]. After provision of their signed informed consent forms, 68 selected volunteers participated in the study and were allocated randomly into experimental and control groups by tossing a coin, leading to 35 and 33 participants in the experimental and control groups, respectively.

The participants in the control group received the SVVR intervention after completion of the study. The research team conducted one-on-one interviews to obtain the survey data. All participants successfully completed the program within approximately 3 hours.

According to Kirk [26], for an estimated effect size of 0.80, the approximate sample size required is 26 for each group when the power is set at 0.80 and type I error at 0.05. A previous 3D VR study yielded significant preintervention and postintervention improvements in psychological health with a similar sample size [19]. Another SVVR study with 32 participants in the experimental and control groups showed that providing childbirth education with SVVR resulted in superior learning performance [21]. Therefore, the sample size in this study (N=68) was large enough to detect training effects. Moreover, all participants finished the training program successfully.

A flowchart outlining participant enrollment and assessments is presented in Figure 1. After selecting the five Sunshine volunteer groups, the research team approached the executive director and staff to explain the research purpose, method, and protocol. After obtaining permission to conduct the study, we posted recruitment messages and held meetings to invite potential participants who met the inclusion criteria so that the participants could fully understand the purpose of the study before providing their written consent. After the meeting, we conducted a 15-minute operation session with SVVR to evaluate the feasibility and acceptance of the program. Participants indicated their acceptance of this arrangement and reported that the SVVR program is very novel and different from their previous training experiences. The research team members collected their baseline data in a quiet room of a local school. During the implementation period, a 3D VR technical professional, staff of the volunteer group, and several trainers of the SVVR program were available to ensure the training proceeded smoothly. After the SVVR training, participants of the experimental group were involved in a 20-minute group discussion. After the group discussion, a counseling professional hosted a 30-minute panel discussion to answer the questions regarding how to effectively help adolescents quit substance use. The 20-minute group discussion and 30-minute panel discussion were conducted to assist participants to organize what they learned into 5 VR scenarios and use them during the future accompanying process with students. After the panel discussion, the participants provided written and verbal feedback to the research team. The participants were encouraged to contact the counseling professional in future to work together in supporting students to quit using illegal drugs. The control group did not receive the SVVR training at the same time but received it after completion of the study.

Before developing the SVVR program, the research team interviewed five volunteers face to face to understand how to help students quit substance use. Verbatim transcripts of the interviews were recorded. Consequently, the research team developed the SVVR program involving drug education, educational psychology and counseling, social work, nursing, and health education professionals to ensure the appropriateness of the SVVR program for the volunteers. The development of scenarios adopted a narrative approach in which each scenario included elements such as the setting, character, and plot. In addition, the context was derived from previous interviews of volunteers. This presentation format could make the scenarios more complete and better meet the needs of the volunteers. After the first draft of the training program was completed, the research team invited the volunteers and professionals to examine the five scenarios and provide their comments. The five scenarios are shown in Figure 2.

The training program offered 5 VR scenarios and invited participants to actively engage in a 20-minute group discussion and a 30-minute panel discussion. Table 1 summarizes the five scenarios. Each scenario has a corresponding activity led by a drug use prevention professional to guide participants in having proper interactions with students. For example, the first scenario focused on ice breaking, greeting, mutual introduction, and welcome messages. In the second scenario, students chose three Tarot cards and interpreted the meanings of these cards. The third scenario was about how to face a situation when students play mobile games and neglect the participants. The fourth scenario was to use Kahoot, a game-based learning platform, to design 10 questions for exploring ways of facilitating communication between students and close peers using drugs. The fifth scenario taught the participants how to become friends or followers of the students on social media platforms (Line, Facebook, and Instagram) to assess the social groups and lifestyles of the students, regularly or irregularly post positive messages, offer students timely attention, and interact with the students when needed. The relationships of the training program components and outcome variables are presented in Table 1.

The research team conducted a 15-minute operation demonstration to introduce the purpose of the training before the SVVR training was formally launched. In this orientation, the 3D VR professional taught the participants how to wear the 3D VR helmets and select the VR scenes. The participants practiced multiple times to avoid dizziness.

The sociodemographic variables of the participants assessed at baseline included their gender, age, the number of years completed as a Sunshine volunteer, marital status, employment, and the number of served students. The performance impact was assessed based on three variables including problem-solving skills, self-efficacy, and teamwork.

The problem-solving ability scale was modified from a previously evaluated instrument [29] comprising 6 items and was scored on a Likert-type scale ranging from 1 to 5, with higher scores indicating a higher level of problem-solving abilities in terms of drug use prevention.

The self-efficacy scale was a modified form of a previously tested scale [30] containing 6 items. Each item was scored on a Likert-type scale ranging from 1 to 5, with higher scores indicating a higher level of confidence in drug use prevention.

The team cooperation scale used was a modified version of a previously tested scale [31] with 4 items. Each item was scored on a Likert-type scale ranging from 1 to 5, with higher scores indicating a higher level of cooperation with team members. Analyses of the Cronbach α and exploratory factor analysis of each variable indicated that a single factor on each scale accounted for 64.26% to 65.20% of the variance, as shown in Table 2.

We conducted descriptive analysis of the demographic variables. In addition, two-tailed χ² tests were conducted to compare differences based on the gender, age, number of years completed as a Sunshine volunteer, marital status, employment, and number of served students. A generalized estimating equation (GEE) was used to explore the effects of problem-solving, self-efficacy, and teamwork. The GEE enables the understanding of the changes and their effects at the individual and group levels by estimating the average response of the population and is more advantageous than regression analysis that would enable prediction of the effect [32]. Statistical analyses were conducted using SPSS 22.0 (IBM Corp).

The study received approval from the Research Ethics Review Committee of National Taiwan Normal University (201805HS007).

Most of the participants in the experimental and control groups were female, aged 51 to 60 years and married. They had served as Sunshine volunteers for 7 or more years and were unemployed. The participants in the experimental and control groups had no statistically significant differences in terms of their demographic variables, as observed in Table 3. The participants in both groups were treated equivalently when comparing the intervention effects.

Group differences in the patterns of change over time are shown in Figure 3. GEE analyses indicated a significant group × time interaction for problem-solving (β=3.055; Wald χ²₁=4.757; P=.03) and self-efficacy (β=6.135; Wald χ²₁=19.033; P<.001) but not for teamwork (β=1.025; Wald χ²₁=1.172; P=0.28), as indicated in Table 4.

After the SVVR training, we invited the experimental group to discuss related responsive strategies. As shown in Table 5, the participants suggested responsive strategies including target setting, value identification, problem-solving skills, peer or parental influence, and environmental changes to assist students. During the panel discussion, the invited professionals and participants discussed responsive strategies including analyzing current situations, making multiple attempts, providing social support, and increasing social connections for students after quitting drugs considering their future life.

To the best of our knowledge, this was the first study using SVVR in a volunteer training program. We adopted a randomized controlled design and recruited 68 participants from 5 cities to have enough power to detect the training effects. Although the score improvement aspects among the participants in the experimental group were superior to those shown by their counterparts in the control group in terms of problem-solving skills, self-efficacy, and teamwork after training, the GEE results revealed that the participants in the experimental group showed better improvements in problem-solving skills and self-efficacy compared with their counterparts in the control group but not in teamwork capabilities.

The SVVR training program improved the participants’ problem-solving abilities when compared to their counterparts in the control group. Problem-solving is one of the life skills that was advocated by the World Health Organization [33]. Problem-solving skills involve identifying a problem, developing possible solution paths, and taking the appropriate course of action. Problem-solving skills can improve adolescents’ problematic behaviors such as substance use [34]. The use of VR technology in the training program offers a novel opportunity for the development of problem-solving skills by providing learners with richer situations; this made the learning process more interesting and interactive and could improve the learners’ motivation and attention, thus helping them explore new possibilities. A recent study supported the use of VR technology stating that it could increase users’ interest and motivation, and potentially assist students in developing problem-solving skills [35]. Apart from the SVVR training program, our study conducted group and panel discussions. It contributed to the development of participants’ problem-solving skills through discussion processes so that the volunteers could understand the importance of having discussions with the other participants.

Another important finding of the study is that the SVVR training program was effective in promoting volunteers’ self-efficacy to help students quit drugs. Self-efficacy is one of the core concepts in Bandura's social learning theory, which refers to the degree of personal confidence in assessing whether a particular behavior is performed in a specific situation. It is a critical variable in predicting and interpreting future behavior [10]. A previous study pointed out that self-efficacy can stimulate behaviors and is the source of motivation [36]. The higher the self-efficacy, the higher is the number of expected healthy behaviors [37]. Self-efficacy could be a significant predictor for drug quitting among adolescents [5]. In addition, self-efficacy reflects the proficiency in personal abilities. An individual’s self-efficacy can be improved through personal experiences of success [38]. Our findings advocate the use of SVVR for improving participants’ self-efficacy, and they are consistent with the findings of previous studies [39,40]. A recent study indicated that using VR-based learning environments with student teachers helped students increase their self-efficacy and allowed them to be more innovative and creative [41]. Thus, using VR for training adolescents in self-efficacy offers high potential at present and in future.

A previous study identified significant differences between virtual and live simulations in terms of teamwork attitudes and communication skills in a randomized controlled trial that was conducted with 120 undergraduate medical and nursing students. These findings supported the potential of VR as a substitute for conventional team-based simulation training [42]. Numerous learning and performing tasks require teamwork. Team members may work concurrently and meet during some occasions. A recent review article summarizing trainings for health professionals indicated that most review studies evaluated the usability and acceptability of VR simulations whereas very few studies have measured the effects of VR simulations on the development of nontechnical skills such as teamwork [43]. Researchers indicated that although the importance of teamwork in health care is recognized, limited consensus exists regarding what it is, how it can most effectively be learned, and how it should be assessed [44]. We suggest that collaborative virtual environments should be developed in future training programs to allow participants to come together to complete a task. VR might have different educational or training effects on teamwork, suggesting that more research is needed in future to clarify its effectiveness.

We conclude that it is feasible to apply 3D VR technology in training volunteers who help with substance use prevention. The results of this study revealed that SVVR can improve participants’ problem-solving and self-efficacy for assisting students in quitting illegal drug use. However, future studies are suggested to develop effective SVVRs to assist inexperienced volunteers in enhancing their teamwork abilities. We believed that it is promising to introduce this training program at more sites to enhance volunteer training so that volunteers can be better companions and help students quit drugs.