The Effect of Virtual Reality Immersion on Evoked Anxiety

Abstract

In the past couple of decades, the use of Virtual Reality (VR) has become increasingly prevalent. Virtual reality is mainly used today to treat anxiety disorders through exposure therapy, but it is also increasing in frequency regarding research. The goal of the present study is to determine the effect of VR immersion on evoked anxiety. The study consists of 20 undergraduate volunteers between 18 and 30 years of age. The results showed a marginal interaction of group and time point, in particular, the experimental group showed an increase in STAI scores post treatment.

Introduction

Virtually reality (VR), a new yet highly used technology, has taken the world by storm. VR immersion aims to give the user a sense of presence by controlling what they see and hear by blocking out any other sight and sound. (Rizzo, Buckwalter, & Neumann, 1997; Price & Anderson, 2007) It is slowly being used throughout households and ages. It has gotten widespread to the point where it is being used in medical settings. (Moline, 1997; Li, Yu, Shi, Shi, Tian, Yang, & Wang, 2017; Riva, 2003; Claudio & Maddalena, 2014) Perception of it in medical settings is overall positive. (Keller, Park, Cunningham, Foulandian, Chen, & Spiegel, 2017) It is transforming patient care. It has even made its way into treatment of mental disorders, particularly anxiety-based disorders, such as post-traumatic stress disorder, generalized anxiety disorders, seasonal affective disorders, general phobias, and schizophrenia. (Powers & Emmelkamp, 2007; Rothbaum, Hodge, Ready, Graap, Alarcon, 2001; Ku, Han, Lee, Jang, Kim, Park, Kim, Kim, Kim, & Kim, 2007) VR is most commonly used for exposure therapy. (Powers & Emmelkamp, 2008; Parsons & Rizzo, 2008; Opris, Pintea, Garc?­a-Palacios, Botella, Szamoskozi, & David, 2011) It is also being used for assessing mental health disorders. (Freeman, Reeve, Robinson, Ehlers, Clark, Spanlang & Slader, 2017; Freeman, Antley, Elhers, Dunn, Thompson, Vorontsova, Garety, Kuipers, Glucksman, & Slater, 2014) With the use of virtual reality becoming prevalent, we must look at the effects it can have on physiological methods as well. Virtual reality exposure therapy (VRET) has been proven to have benefits in mental health settings particularly in relation to anxiety disorders and specific phobias. (Parsons & Rizzo, 2008; Opris, Pintea, Garc?­a-Palacios, Botella, Szamoskozi, & David, 2011)

Gorini et al., (2010) proposed to improve treatments using modern technologies, in particular, biofeedback enhanced VR.

Methods

Participants

Participants consisted of a convenient sample of 20 volunteers between the ages of 18 and 30. 14% of the participants were male. One outlier was removed due to age. All participants were recruited from psychology classes at University of Massachusetts Dartmouth. There was offered extra credit for participation as compensation for their time. Participants were randomly assigned to either an immersive virtual-reality based experimental group, or a non-immersive group in which stimuli was presented on a laptop. They were informed about how the procedure would progress and what would happen. Participants signed a consent form with a summary of what was expected. They were informed that their participation was confidential, and their consent forms are stored in a secure location. This research was reviewed by the Institutional Review Board at the University of Massachusetts Dartmouth prior to data collection.

Measures

State Trait Anxiety Inventory (STAI) - STAI was used to measure state anxiety before the start of the procedure, and at the end after all stimuli had been presented. (Spielberger et al., 1983) While there are multiple variations, the most common version is the Y one. The STAI consists of two scales. The first scale is the Y-1 which measures how the person feels at the moment. The final scale is Y-2 which measures how a person feels currently. Each part has 20 statements about how a person feels. The person has to score from 1 (Not at all) to 4 (Very much so.) A few questions are reverse-coded to determine consistency. The scores were calculated. The higher the score, the more anxious the person was feeling. For our study, only the Y-1, or state, scale was used. (Appendix A)
Fitbit Charge 2 - A Fitbit Charge 2 was used to measure initial heart rate before the intervention was used and final heart rate. Heart rate was used as the physiological indicator of anxiety.

Oculus Rift - An Oculus Rift was for the immersion intervention. A computer screen will be used for non-immersion intervention. Three videos were used as stimuli for both conditions. The videos allowed for the participant to look around. Video 1 was a shark attack scenario with audio and video to set the scene. The participants point of view is from deep under water within an enclosed space, as a shark attempts to break through. The glass starts to shatter as they shark attempts to enter. This video used the common fears of claustrophobia and nyctophobia and lasts 2 minutes. Video 2 represented a story from the first-person point of view of a child in bed during a lightening storm. As the child, player, reads a book, the character in the story comes alive. It crawls around the room with its glowing eyes and the person has to track it to move the story along. The creature will do something mundane and then look at the player and charge at them. This video combines the common fears of nyctophobia, and xenophobia. It is also filled with jump scares and lasts 5 minutes. This video had potentially anxiety-inducing sounds and images related to horror. Video 3 was from the perspective of someone on the top of an unstable building while there is a robot alien attack. The wall in the skyscraper the player is in is ripped away. The player is then grabbed by a giant massive robot and pulled out. The player is then set on a latter that is dangling over a city as the robot falls to the ground. If the player looks around as the robot falls, they will see a battle between planes and a spaceship ensuing. As the battle is occurring, the player will get the sensation that they are being beamed up. Suddenly, the ship is destroyed and the player than falls rapidly through the city creating the sense of falling. This video used the common fear of acrophobia and only lasted 3 minutes. This is the only stimulus that the participants were asked to stand for.

Design - It was a randomized, experimental design. Numbers were randomized using a random number generator to determine order (ex. 16, 2, 12, 9, etc.) The numbers were then divided into two groups for the intervention (ex. Number > 10 goes into experimental group.) The experiment was conducted in a single session per participant of approximately 20 minutes in two rooms of the Liberal Arts building at University of Massachusetts Dartmouth.

Statistical Analysis A MANOVA was run to check for interactions between group and time point. An ANOVA was run for each dependent variable to check for interactions between group and time point.

Proposed Results- We expected an interaction between time point and group. We expected the VR group would see an increase in both STAI scores and HR with no effect seen in the control group. We set our alpha level at 0.05 for significance.

Procedures

The participants were greeted by a researcher and led to the area where the study was conducted. They were asked to sit down and had the procedure explained to them; what was going to happen, what was being measured, that they had the ability to end the experiment at any time, and what group they would be in. They were told that their participation is confidential and were safely secured in a different location separate from their surveys. Then the FitBit Charge 2 was connected the left wrist and heart rate will be measured. The participants then they filled out the STAI Y-1 questionnaire to assess their baseline state anxiety. The participants were then placed in their group. If they were in the experimental group, the Oculus Rift was placed on their head. If they were in the control, they were moved to sit in front of a computer scene with speakers raised to a comfortable volume while one of the researchers wore the Oculus Rift. After being established, the researchers started the sequence of videos. Following the final video, the headset was removed, or they were moved away from the computer scene. They were then asked to take a second STAI Y-1 form to assess state anxiety following the intervention. After completion of the STAI Y-1 questionnaire, a final heart rate was measured, and the FitBit Charge 2 was removed. The participants were thanked for their participation and escorted out of the room.

Results

Average heart rate and STAI scores were entered into a repeated-measures multivariate analyses of variance (MANOVA) with group and time point as fixed factors. Of interest was a significant group x time point interaction. The analysis was conducted using SPSS (Version 25). The multivariate results indicated only a marginal group x time point interaction [Wilks' ?» = 0.724, F(2,17) = 3.236, p = 0.064, ?·2p = 0.538]. However, follow-up univariate tests, performed for the dependent variables separately, showed a significant group x time point interaction for the STAI [F(1,18) = 6.73, p = 0.018, ?·2p = 0.689], but not for heart rate [F(1,18) = 0.970, p = 0.338, ?·2p = 0.154]. Specifically, STAI scores (Fig. 1) significantly increased after the VR intervention. Figures 1 and 2 illustrate these results.

Discussion

VR immersion is being used more frequently within medical settings. The main goal of the study was to determine if VR immersion has an effect on physiological and behavioral factors. Our primary focus was to examine the short-term effects that VR immersion has on heart rate and perceived, state, anxiety. We randomly assigned 20 participants to either the VR immersion group or the control group, same videos but on a computer screen. We expected to find no change in heart rate and perceived anxiety levels in the control group. We expected to find an increase in both heart rate and perceived levels of anxiety in the VR group from baseline to post-intervention. Results showed that there on average was a decrease in heart rate in both the control and VR groups post intervention, however, the VR group showed a greater decrease in heart rate. (Fig. 2) The difference came from when we compared the STAI results of both groups. The average STAI scores in the VR immersion group increased while the average STAI scores decreased in the control group. (Fig. 1) These results support previous research that VR immersion influences perceived anxiety. (These results only partially supported our primary hypothesis. VR immersion did affect perceived anxiety levels. However, it didn't affect heart rate in the way we hypothesized.

There are a few possible reasons for the results the study produced. The VR group felt an increase in perceived anxiety because the VR immersion gives them a sense of presence. The feeling that they were actually there. They felt the ghost charge at them or the sensation they were falling. Those videos made them feel more anxious post intervention. Each video consisted of one to three of the most ten common fears in the United States. These videos successfully induced anxiety, regarding the STAI, in the VR group. Which leads us to talk about how the HR in the VR group decreased. There are a few possible reasons why HR would decrease in the VR group. First, while perception influences our physiological response to stress, the amygdala filters out that the threat is not imminent causing the prefrontal cortex to further evaluate that threat. All of this is happening subconsciously regardless of the perception of the person. These subconscious processes are represented in the physiological measure of heart rate steeply decreasing. (Fig. 2) Another potential reason for the decline in HR in the VR is the order of videos. Most participants reported finding the second video more frightening than last. The time between the second videos ending and the final heart rate was approximately 5 minutes. There is also the possibility that the VR group just had a naturally high heart rate that decreased as they relaxed, physiologically, and felt more comfortable with the experience in spite of the increase in STAI scores.

We can compare this to the STAI and HR changes in the control group. Both the average STAI scores and HR saw decreases post-intervention. The decrease is the STAI can explained by the fact that they had to listen to one of the researchers wearing the helmet and possibly letting out unintentional noises due to the stimuli. It would allow the control to relax more which would explain the decrease. The decrease is the average heart rate was not a steep as the one in the VR group. This can be explained easily. The fact that the control group did not have to filter out the threat can account for the smaller drop in HR when compared to the VR group.

The interpretation was supported what we were told by participants following the experiment. Participants in the VR group indicated that they felt anxious following the intervention.. The control group had a different opinion regarding the videos from their perspective. They assumed that if they were in the VR group they would have felt more anxious as the videos looked anxiety inducing from the control groups' perspective. Another fact that must be noted is that even though there was an increase in perceived anxiety in the VR group nearly all participants said they would like to do it again. They enjoyed it even if it made them anxious. Most participants were excited about to participate in this study due to the VR aspect. Few had previous VR experience.

Additionally, there are other possible explanations for the contradictory results regarding HR compared to STAI between groups and time points. While heart rate is a good physiological indicator, it is not always effective, a more effective physiological measure is that of heart rate variability (HRV). A technique that measures the time between heart beats along with heart rate. Most studies related to the physiological measures of anxiety use HRV in lieu of heart rate for this specific reason. This issue is only exacerbated by the fact that the tool used was only a crude measure. A Fitbit Charge 2 is good at looking for heart rate at any given time. It's not effective in measuring heart rate continuously which would allow researchers to observe how heart fluctuates over time. Another issue was the timing of which we measured the initial heart rate. We measured the initial heart rate immediately after attaching the FitBit. We noted that initially the heart rate was much higher when you measured immediately after attached the FitBit then when you attached it and waited a minute. For consistency's sake, we measured immediately after attachment. Other limitation is the fact that our group consisted of only 20 participants with the majority being female. Which limits the study in two ways, one is females tend to have a faster heart rate naturally than males.

The other is we could not determine if the effect of VR immersion influenced heart rate and perceived anxiety the same way it would in a male. Which brings me to the age, the mean age of the participants was 20.7 years old which is another factor that cannot be extrapolated for the general population. This age group tends to have more experience with VR and technology in general than other populations. Our study was limited to University of Massachusetts Dartmouth students. While the study was conducted in a similar setting it was conducted at different times during the day, with most of the participants coming after classes. Approximately fourteen students came immediately after receiving a test back where the average score for the class was 68. The environment can influence how anxious one feels initially, especially a campus at night, where and when the final study took place, when most students and faculty have gone home. One final limitation that should be mentioned is the time in which this study was to be completed. The time from recruitment to study spanned at day at most and an hour at least. Further research is needed to assess if these limitations influenced the results the study produced.
We recommend a few possible directions for further research. The sample needs to be increased to include more males to replicate the general population.

The sample should be in general to see if the interaction that this study had is repeatable. It should be done with tools that allow for a precise measure of heart rate. (ex. EKG) There should be a down period in between videos. This study only allowed for a minute before taking the final heart rate. We also suggest taking HR after each video. This study did that initially but for simplicity sake, the HR measurements were removed from the analysis. The environment should be controlled to filter out possible environmental influences. It should be done at a consistent time during the day not spanning the day as this study did. The study should allot more time to collect data and recruitment. It should be done on a more relaxed schedule when participants are not potentially in a rush to get somewhere else and to allow for the videos to complete without the added anxiety of being late to a class. There should be a variety of VR immersion technology. This study used the Oculus rift for VR immersion. There are multiple forms of VR immersion. We do not know if these results would be repeated with different VR systems. There should other stimuli. This study used mainly scary videos, building on a few common fears (eg. heights, jump scares, etc.) comparing those to videos of what most find relaxing (eg. sitting on the beach, relaxing in the woods, etc.) Another suggestion is to alter the sequence of the videos from least frightening to most frightening shortening the time from the most frightening to the final HR measurement. A final suggestion is to establish trait anxiety and see it influences STAI and HR or affects an increase/decrease in state anxiety scores. Trait anxiety would allow for researchers to expand the results seeing if those with trait anxiety felt more anxious.

Conclusion

VR immersion is a new and increasingly popular tool. It's allowing people to see places they will never see in real life and overcome fears with exposure therapy. VR is doing a lot of good for society. Studies like this one are needed to help improve treatments because when we know what can influence perceived anxiety, we learn better ways to treat anxiety in general.

Cite this page