The acrophobia experiments

This chapter will describe the experiments that were done using the environments described in the previous chapter. The first section will give a description of the hardware used and the setup. The second section will describe the first group experiment that was done, and the third section will describe the case study that was done later. The final section in this chapter will discuss the results from the two experiments and the conclusions that can be drawn from this concerning presence.

Setup used for experiments

The place where the testing was done (see Figure 7.1) has no privacy at all. People can look inside through the windows, or when they want to walk to their workplace, they have to pass the setup. This was acceptable during testing. With the demonstrations, it was even desirable that as many people as possible could see it. But with the experiments, the level of privacy had to be increased.

Figure 7.1: Virtual Environments

In one of the corners of the GVU center there is a usability room. This room has two cameras mounted in the corners and a one way mirror to the GVU center. The area in front of the mirror can be closed off from the rest of the GVU center (see Figure 7.2). Using this room, it was possible to guarantee the privacy of the people participating in the experiment.

Figure 7.2: Setup of experiment

The operator would sit somewhere else in the GVU center and has no visual contact with the usability room. To be able to know what is happening and to "teleport" the subject to the next level when necessary, the audio from the usability room is sent to the operator.

Outside the usability room, a member of the Virtual Environments group would look after the video recorder. The video recorder was used to tape the subject in the usability room, the virtual environment the subject is looking at and the audio from the usability room. Whenever there was a problem with the communication between the psychologist inside the usability room and the operator, this person could also be used as a runner between the operator and the psychologist.

The subject stood in the mock up of the elevator wearing the HMD inside the usability room. The information from the tracker was now sent directly to the SGI Reality Engine, which calculated the images. Unlike the testing of the environments, there was no longer a second machine necessary. The disadvantage of this setup was that it was no longer possible to use audio in the environments, the SGI Reality Engine has no audio capabilities.

The psychologist was also in the usability room. He could watch how the subject was doing in the virtual environment and watched the virtual environment on the TV screen. When the subject was ready, the psychologist would tell the operator to "teleport" the subject to the next level.

Also in the usability room was a normal TV camera that was operated by the person watching over the video recorder. This camera was pointed at the subject. There was also a microphone in the usability room. Output from this microphone was sent to both the operator and the video recorder.

Group experiment

The first experiment was started in October 1993 and lasted until December of the same year. This experiment used two groups, one group that would get treatment by using VR Graded Exposure and one group that would not be treated. The following section will describe how the experiment was set up and how the people were selected. The next section will give some results. The results can be split in two groups, those that have to do with the efficacy of the treatment, and those that have to do with presence.


This section will describe the setup of this experiment. First the subjects were selected and put in a group. Next the experiment was done and finally the results of the two groups were collected.


Among students taking introductory classes in psychology and computer science at two universities in Atlanta, a questionnaire was given. 478 students returned this questionnaire and were screened for their fear of heights. Out of these 478 students, 46 were identified as possible acrophobes. 41 students were contacted by phone, 31 of these students were offered entry into the study. Out of the 20 subjects that actually entered the study, 17 completed the study. Of those who entered, 12 were male, 18 were Caucasian, and the average age was 20 (SD=4.2). Students from psychology courses (n=8) were given credit for participating in this study. Students with concomitant panic disorder, agoraphobia, or claustrophobia were excluded as wearing the HMD might cause them distress.

The subjects that entered the study got a number. This number would be their subject number and was used to insure confidentiality. The subjects were randomly assigned to either a treatment (VRGE) group or a wait-list control (WL) group. The VRGE group was treated for acrophobia using the environments described in the previous chapter. After the treatment, the results were compared with the WL group who did not get any treatment. After the experiment, the WL group was offered the possibility to be treated with VR.


Both the VRGE group and the WL group had to do a pre and a post test. During the pre test, the subject had to fill in a Client Information sheet (see appendix A-1), and to protect both Emory Medical university and Georgia Tech, the subjects had to sign a waiver (see appendix A-2). After filling in these forms, the subject was given three questionnaires that had to do with height. These questionnaires will be described in more detail in the following three paragraphs. After the treatment, or in the case of the WL group when the VRGE group was finished, the same three questionnaires had to be filled in again. To see whether or not the treatment was a success, the pre and post tests of the VRGE group were compared with WL group.

The Acrophobia Questionnaire (see appendix A-3), [cohen 77] describes twenty situations with rating scales for anxiety (0-6) and avoidance(0-3). The Acrophobia Questionnaire yields a total score ranging from 0-180 and anxiety (0-120) and avoidance (0-60) subscale scores. Adequate consistency and test-retest have been previously demonstrated, and it has been shown to discriminate between phobic and nonphobic subjects [cohen 77]. The questionnaire has been proven to be more sensitive to treatment outcome than behavioral tests [cohen 77].

The Attitude Towards Heights Questionnaire (see appendix A-4) contains six items assessing attitudes towards heights and was adapted from Abelson and Curtis. These include the following dimensions rated on a 0-10 semantic differential scale: good-bad, awful-nice, pleasant-unpleasant, safe-dangerous, threatening-unthreatening, and harmful-harmless.

The Fear Questionnaire (see appendix A-5) was constructed for use in this study. One item from the Marks and Mathews' Fear Questionnaire assessing the degree of distress related to acrophobia. The three situations to be used for VR Graded Exposure, glass elevator, outdoor balconies, and bridges, were rank ordered and rated for the amount of discomfort they produced.


After filling in the pre test, the VRGE group was taken upstairs and the first treatment was given. In the first treatment session, the subjects were familiarized with the VR equipment. First one of the members of the Virtual Environments group explained the equipment and showed the subjects a small demo. Each of the subjects was allowed to take a turn in wearing the HMD and was encouraged to interact with the environment. The environment was a model of the GVU center with a big light switch close to them. Touching this switch with the virtual hand turned the light on and off in the virtual GVU center. This demo had no height clues in it, the capabilities to fly upwards were turned off.

During this small demo, one of the subjects touched a virtual monitor and the monitor changed color. The change of color was due to aliasing but the subject did not know about aliasing and reacted to this by telling everybody he could turn the monitor on and off.

After this demo, the subjects were sent home and were individually treated over the next seven weeks. The treatment started with the environment rated least scary on the Fear Questionnaire and after seven weeks should end with the environment rated most scary.

The sessions in the first week lasted 50 minutes. But after people reported nausea, one subject even had to vomit, sessions were cut down to 35-45 minutes. The nausea was not correlated to the height experience but more to what is called simulator sickness. Simulator sickness arises from discrepancies between what the body is perceiving visually and kinetically and feels similar to motion sickness. After the first week, people were asked how they were feeling and that they could rest at the first sign of nausea.

To measure the level of discomfort, and to us the level of presence, they were asked to give SUD, Subjective Unit of Discomfort, ratings. These ratings were asked every five minutes and were between 0 and 100, where 0 means relaxed and 100 near panic. The SUD's were used by us to get an idea about the level of presence. When a subject does not feel present, his SUD will be low, also when the subject is "teleported" from one location to the other, his SUD will not change. If the subject does feel present, his SUD will start off high and go down slowly, and back up high again when he is "teleported".


The experiment gave a lot of results in the form of SUD's and difference between pre and post tests. Also the comparison between the WL group and the VRGE group showed an enormous difference. This section will show some of the results and some of the more remarkable comments of the subjects during treatment. More numbers can be found in [hodges 94a], [hodges 94b] and, [rothbaum 94a].


The following table (Table 7.1) shows the means and standard deviations between the pre and post tests from both the VRGE group and the WL group.

Table 7.1: Means and standard deviations Pre- and Post-Treatment

This table (Table 7.1) shows clearly that the WL group stayed the same while the VRGE group had made enormous progress. The distress from fear had gone down from 4 (definitely troublesome) to 2 (slightly troublesome). The same goes for the attitude towards heights questionnaire. The attitude has gone down from around 6 (negative) to less than 4 (slightly positive).

From these numbers, we can conclude that it is possible to help people afraid of heights by using VR.


The subjects were videotaped while they stood in the mock up of the elevator. Very often it was visible that the subject held on to the railings, letting go of the railings when he became more comfortable. Then when he was taken up to the next level of the hierarchy, he grasped the railings again. This showed that the subject felt he was present in the environment.

Comments made by the subjects also showed that most subjects were immersed in the virtual environment. Some sample comments:

"I'm feeling a little weak in the knees. My chest is getting tight. My palms are sweaty" (session 2).
"A lot easier than last week. Last week I was terrified" (session 4).
"This is the first time I can look down and not get that weak in the knees feeling." (session 6).
Another, small, clue that the subjects felt present in the environments is that 7 out of the 10 subjects went out and exposed themselves to height situations without specifically asking them to do so.

Case study

The case study used the same environments as the group study, but instead of a group of people, now one person was used. The following section will describe the setup in more detail. The next section will describe some of the results.


The setup of the hardware was the same as used in the group experiment. The subject was one of the people in the WL group who wanted to be treated by using VR. He had to fill in the forms (see appendix A1-A5) and got a demo to get used to the VR equipment. Besides the questionnaires, he was taken up the Marriott Marquis in downtown Atlanta, Georgia, accompanied by a psychologist. This test is called the Behavioral Avoidance Test. The subject had to ascend the glass elevator as high as the subject wanted to go. He gave a rating at various floors.

During the treatment with VR, only the elevator environment was used. The psychologist again watched the subject, who was standing in the elevator mock up, and the TV that showed the Virtual Environment as seen by the subject. The subject was again asked every five minutes to give a SUD rating.

After the treatment, the subject had to fill in the questionnaires again, and was taken to the Marriott Marquis for a ride in the elevator, the post Behavioral Avoidance Test.


A full description of this case study can be found in [rothbaum 94b]. This section will show some of the results found, and some conclusions that can be drawn from this case study.

This case study was the second time that subjects were treated from acrophobia by using VR Graded Exposure. Again the subject benefited from this treatment as can been seen in Table 7.2. Especially the difference in the BAT shows that the treatment was working for this subject. As with the group study, this subject went out and exposed himself to height situations in vivo without being asked by the psychologist.

Table 7.2: Scores Pre- and Post-Treatment


Both the group study and the case study support the idea that VR can be used to treat psychological disorders. In both studies, the subjects reduced their anxiety and, without being asked by the psychologist, exposed themselves to height situations in vivo.

Using VR Graded Exposure the psychologist does not need to leave his office for treatment of a patient, saving time and money, and a patient who has trouble imagining can be stimulated. Another advantage of VR Graded Exposure is the possibility to overcorrect the fear and avoidance of the patient.

Not only for psychologists was this study an success. The studies helped us to get a better insight in presence. The following assertions were deduced from the experiment. New experiments will be necessary to prove them.

Assertion 1
A person's experience of a situation in a virtual environment may evoke the same reactions and emotions as the experience of a similar real-world situation. This may be true even when the virtual environment does not accurately or completely represent the real-world situation.
The subjects reacted as if they were at a high place, although the virtual environment had some peculiarities in it. Subjects were teleported from one bridge to another. The subjects were standing in the mock up of the elevator all the time, even when they were on the bridges. Still with all these misrepresentations, the subjects felt like they were in a height situation. Either they gave a high SUD rating or told they "did not like this".

Assertion 2
Each person brings their own Gestalt into a VR experience.
Gestalt is used here to describe the person, who he is and all the previous experiences he had. Each person brings with him into the virtual environment all the experience he had in the real world. Most of the people, visitors and subjects, who went up the elevator did not want to step off the elevator, afraid they would fall down. Nobody told them there was gravity in the virtual environment, in fact there was no gravity in the virtual environment. Stepping off the platform would have left them floating in the air. Most people however did not dare to step off the platform, afraid they might fall down.

Also during the studies this Gestalt became apparent. People reacted like they would normally to height situations. One of the subjects even remarked that he used a previous experience he had in an elevator to increase the anxiety he got from the virtual elevator.

Another nice example of this Gestalt is the fact that a lot of people felt an impact on their knees when they came down with the elevator. The mock up of the elevator they stood in was standing solidly in the usability room, so no movement was possible. Still, when the elevator came down to a stop they could feel it, as they would in a real elevator, in their knees.

Assertion 3
A person's perceptions of real-world situations and behavior in the real-world may be modified based on his experience within a virtual world.

Both studies showed that it is possible to treat people suffering from acrophobia. Especially the group study showed an enormous difference between the VRGE group and the WL group. This assertion can have a great impact on the treatment of phobias in general. More experiments are necessary to prove the generalization of this assertion.

Assertion 4
A primary difference between the experience of an event in a virtual environment and the experience of the same event in a real environment is in the intensity or vigor of the experience.

The bridges the subjects were standing on were placed between 7 and 80 meters above the ground. Especially the top bridge was very unsafe. One of the wooden slats was missing and all that kept them from falling was a small rope. In real life, nobody, even people not suffering from acrophobia, would stand on a bridge like this. All the subjects, however, managed to reach this bridge and feel comfortable standing on it.

Assertion 5
Familiarity with a virtual environment does not necessarily increase the participants sense of presence.

Each time the subject was taken in a new scenario, the level of presence increased. After some time, this level of presence became less. Subjects told us they had to think about experiences they had before in the real world to keep up the level of presence.

The final chapter will link these five assertions to the research questions asked, as well as to the five sub questions asked and the previous research that has been done.

Designing environments (1) Conclusions TOC
Rob Kooper

Last modified: Wed Aug 9 12:36:55 GMT 1995