Keywords: virtual reality; test anxiety; radial arm maze; state anxiety; trait anxiety; worry; emotionality; task-irrelevant cues; task-relevant responses; time constraint

Cattell and Scheier (1961) understood state anxiety as a temporary or situational specific form of anxiety that vacillated over time as a function of perceived threat, whereas trait anxiety was a more stable characteristic.  Five years later, Spielberger (1966) identified state anxiety as a transitory emotional state, in which a person is subjectively conscious of one’s perceived emotions such as tension, nervousness, apprehension, worry, and heightened autonomic arousal.  In contrast, Spielberger (1966) defined trait anxiety as relatively stable individual differences in anxiety, i.e., the differences between individuals and their tendencies to respond with higher levels of state anxiety when placed in a situation that is perceived to be threatening.  Individuals with higher levels of trait test anxiety are predisposed or prone to perceive situations as more threatening than individuals low in trait test anxiety.

Liebert and Morris (1967) conceptually identified worry as a cognitive expression of concern, which consisted of self-reported ratings of fear of failure, negative expectations and consequences, lack of confidence, and distress about performance in relation to one’s peers.  Worry also seemed to be the key factor or component that was directly related to test performance.  The worrying individual directed his or her attention to the self and not to the task at hand, lacked confidence, and often felt insecure about his or her abilities.  In contrast to worry, Liebert and Morris (1967) referred to emotionality as autonomic arousal or one’s subjective perception of physiological arousal caused by examination stress.  Emotionality refers to unpleasant feelings such as nervousness, tension, rapid heartbeat, and upset stomach.  However, it is very important to understand that emotionality refers to one’s self-perceived autonomic arousal and not to just the physiological arousal itself.  Worry can last longer and may not decrease during a testing period unless a student’s confidence increases during that time. However, emotionality appears to dissipate once the testing period has commenced, while worry appears to continue if perceived test difficulty remains constant.  Therefore, worry consistently correlates negatively with performance, whereas emotionality is either unrelated or not as consistently linked to performance.

Research conducted by Sarason et al. (1952) looked at the effects of task-irrelevant and relevant responses in relation to task difficulty with high anxious and low anxious individuals.  According to Sarason et al. (1952), it was suggested that  “When a stimulus situation contains elements which specifically arouse test or achievement anxiety this increase in anxiety drive will lead to poorer performance in individuals who have task-irrelevant anxiety responses in their response repertory” (Sarason et al., 1952, p. 561).  Therefore, high test-anxious individuals when placed in a stressful situation would display more task-irrelevant responses in relation to the difficult task than low test-anxious individuals.  Thus, task-irrelevant responses interfere with test performance.  However, individuals who respond to a testing situation with task-relevant responses perform better on difficult tasks than individuals with task-irrelevant responses.

Since the time of Sarason et al. (1952), recent research has suggested that task-generated interference or task-irrelevant thinking can impede or hinder test performance (Wine, 1971; Deffenbacher, 1978a; Sarason, 1984).  When tests are administered under highly stressful or evaluative situations, high test-anxious individuals perform more poorly than low test-anxious individuals.  Highly test-anxious individuals divide their attention between self-relevant responses and task-relevant responses.  A self-relevant response is a response that is not related to the successful completion of the task at hand.  Task-irrelevant responses include worrisome thoughts such as fear of failure and the effects of time constraint on test performance.  In contrast, task-relevant responses are behaviors or cognitions that enhance test performance and do not distract the individual from the task at hand.

Research has suggested that highly test-anxious individuals are generally more self-preoccupied with their own thoughts and negative cognitions that than low test-anxious individuals (Wine, 1971; Deffenbacher, 1978a).  As a result, self-preoccupied thoughts and worry cognitions interfere with task completion and are activated during a testing situation.   In contrast, low test-anxious individuals focus their attention on the task at hand and do not divide their attention in comparison to high test-anxious individuals (Wine, 1971; Deffenbacher, 1978a; Sarason, 1984). Consequently, low test-anxious individuals perform better on tests than high test-anxious individuals because they do not divide their attention and display task-irrelevant responses.

Research has indicated that time constraints can have either a negative or positive effect on an individual by increasing or decreasing arousal (Mandler & Sarason, 1952; Siegman, 1956; Morris & Liebert, 1969; Onwuegbuzie & Seaman, 1995; Kellogg et al., 1999).   When time is a factor taking an exam, test anxious students perform more poorly on the exam than low anxious students (Siegman, 1956; Morris & Liebert, 1969; Onwuegbuzie & Seaman, 1995).  Research has found that low test-anxious students perform better on timed tests because time is seen as a motivating force (Siegman, 1956).  However, research has also suggested that high and low test-anxious students perform the same when under a time constraint (Kellogg et al., 1999).  In addition, time constraints have been found to be a task-irrelevant cue, which can distract a student from successfully completing a test (Sarason & Stoops, 1978a).  Thus, performance decreases because attention is divided between both the task-irrelevant and the task-relevant cues.  Though the research is not consistent, time constraints appear to play an integral part in test anxiety and can affect test performance.

Research in the past has indicated that manifest anxiety can have a negative effect on memory and maze performance (Farber & Spence, 1953).  Farber and Spence (1953) found that high anxious students performed more poorly in remembering a stylus maze than low anxious students.  However, in 1956, Axelrod et al. replicated the study by Farber and Spence (1953) and found no significant differences between anxiety groups and the number of errors or trials needed to successfully complete the maze.  Both the high and low anxious participants took the same amount of time to complete the maze and made the same number of errors as well.  Though the literature was inconsistent, more recent research has been conducted, which has found that high test-anxious children perform better than low test-anxious children when attempting to solve a stylus maze (Nijhawan & Cheema, 1971).  More recently, research by Moffat et al. (2001) examined the effects of age differences in spatial memory in a virtual environment.   Results from this study on the effects of age differences on spatial memory in a virtual reality maze were consistent with previous research conducted in a naturalistic or real world maze (Kirasic, 1991; Kirasic, 1992; Newman & Kaszniak, 2000).  Therefore, it appears that virtual reality is a reliable way to measure spatial memory and maze performance.

Based on the literature reviewed, the effects of time constraint on high test-anxious individuals can have a detrimental effect on test performance (Sarason & Mandler, 1952; Siegman, 1956; Morris & Liebert, 1969; Sarason & Stoops; 1978a; Eaton, 1979; Onwuegbuzie & Seaman, 1995; Kellogg, et al., 1999).  Research has also indicated that anxiety can have a detrimental effect on spatial memory when attempting to remember and solve a maze in both the real world as well in a virtual environment (Farber & Spence, 1953; Moffat et al., 1998; Moffat et al., 2001).  Thus, the use of virtual reality and its effectiveness to accurately replicate a naturalistic environment has become increasingly popular in recent years (Knox et al., 1993; Wiederhold, 1998).

Though recent research studies have looked at the effects of time constraint on test performance (Sarason & Mandler, 1952; Siegman, 1956; Morris & Liebert, 1969; Sarason & Stoops; 1978a; Eaton, 1979; Onwuegbuzie & Seaman, 1995; Kellogg, et al., 1999), as well as the effects of stress and anxiety in relation to maze learning (Farber & Spence, 1953; Axelrod et al., 1956; Nijhawan & Cheema, 1971), no published study has been conducted on the effects of time constraint on high and low test-anxious individuals in a virtual reality setting.  It is important to assess whether or not high and low test-anxious students continue to experience test anxiety when attempting to complete a cognitive task in the virtual world.

Research has already shown that test anxiety can be detrimental to test performance on traditional paper and pencil based exams, however, no studies have been conducted on test anxiety in virtual reality.  It is apparent that virtual reality may be a new way of teaching or taking tests in the near or distant future because many colleges and universities have already done away with paper and pencil-based courses and tests.  Therefore, virtual reality may be the next step in education and testing.  In addition, virtual reality might also be used as a way of dealing with or treating test anxiety (Knox, et al., 1993).  Consequently, it is imperative to assess the effects of time constraint on high and low test-anxious individuals in a virtual reality setting.

In the present study, test anxiety will be evaluated in participants after each participant has attempted to complete an 8-arm virtual reality radial arm maze on a computer. The first purpose of the study is to determine if individuals who normally experience test anxiety on a traditional paper and pencil test will also experience test anxiety in a virtual reality setting.   The second purpose is to ascertain if time constraint will have a detrimental or beneficial effect on high and low test-anxious individuals based on their maze performance in a virtual environment.  The hypothesis is that high test-anxious individuals will need more trials to successfully complete the virtual maze when placed in the time constraint group than low test-anxious individuals.

Procedure
        Prior to participation, each participant was given instructions for the study (see Appendix A) as well as a consent form (see Appendix B).  The consent form discussed the purpose of the study, how long it would last, a description of the study including its procedures, the benefits as well as potential risks, how the confidentiality of every person involved would remain anonymous, and the rights of the participants would be protected.  The study was conducted according to the American Psychological Association code of ethics and was approved by the International Review Board.
        Once the participants consented to being in the study, each participant signed up to be tested in the virtual reality maze at various times in room 3414 in the psychology department.  When each participant arrived for his or her scheduled appointment, each participant was given a demographics questionnaire to fill out (see Appendix C).  Once the questionnaire had been completed, each participant was randomly assigned to either the experimental group (timed) or the control group (untimed) by a flip of a coin.  Instructions for the maze were then given to the participants depending on which group each individual was assigned to (see Appendices D and E).
        Once seated at the computer and after instructions had been read, each participant was individually tested on a virtual reality 8-arm radial maze.  The participants in the control group had an unlimited amount of time to successfully complete the maze.  The participants in the experimental group were administered the same test as the control group, but were asked to complete the maze in 15 minutes.  For the participants in both the experimental and control groups, a regular digital time clock was placed adjacent to the computer in order to indicate how much time was left in the testing period.
Once time had expired or once each participant had completed the maze, each individual had to then fill out a posttest questionnaire (see Appendix F) as well as the Test Anxiety Scale (Sarason, 1978b; see Appendix G).  When the questionnaires were completed, each participant was debriefed on the purpose and goals of the experiment (see Appendix H). Each person was compensated for participating in the study by receiving course credit.

Measures
The demographics questionnaire (Appendix C) was developed by the primary investigator of the study and thus does not have any reliability or validity scores.  The demographics questionnaire was used to gather general information about each participant such as age, sex, gender, etc.  In addition, the primary investigator of the study developed the posttest questionnaire (Appendix F) as a manipulation check to determine if anxiety was or was not induced.  Thus, there are no reliability or validity scores for the posttest questionnaire.  The Test Anxiety Scale (Sarason, 1978b) was also used to determine how test anxious each individual was when attempting to complete the maze (see Appendix G).

Instruments
Derek Hamilton  (2003) developed the 8-arm virtual reality maze.  The maze was administered on an IBM-compatible Pentium 4 Processor.  The computer was equipped with a 1024MB ECC RDRAM and a 20-inch flat panel monitor.  The radial arm maze began with the participant in the middle of the maze.  Each participant was able to turn 360 degrees in the maze, which enabled the individual to view all 8-arms of the maze.  By using the arrow keys on the keyboard, each individual was able to change direction in the maze.  Each participant had to go down each arm of the maze and hit the wall at the end in order to determine if there was an object at the end of an arm.  If there was an object at the end of the arm, a red number 1 would appear at the top of the computer screen.  The goal was be to determine and remember where each of the four objects was in the maze.

Once all of the objects were discovered, the participants had to remember where each of the objects was and had to do the maze correctly twice, consecutively. If a participant hit the end of the wall where an object was not located, the participant would have to start the maze all over again until it was completed in the correct fashion two times in a row.  In the maze, there were various environmental or spatial cues that would help the participants to remember where each of the objects was located (e.g., a smokestack).

A series of Pearson r Correlations were performed on high and low test-anxious, grouping, trials, and all ten posttest questions (Appendix F).  There were a total of twenty-three significant correlations between various posttest questions, if the student was low or high test-anxious, and with the number of trials it took to successfully complete the maze.  Table 3 not available!

The results were inconsistent with research conducted on test anxiety.  According to previous research (e.g., Liebert & Morris, 1967; Morris & Liebert, 1969; Sarason, 1984), high test-anxious individuals when placed in a timed condition perform more poorly than low test-anxious individuals.  High test-anxious individuals divide their attention between task-irrelevant cues, such as watching the clock or attending to worry cognitions, as opposed to displaying task-relevant thinking (Sarason, 1984).  According to the literature, having a timed test can have an inhibitory effect on high test-anxious students, whereas time can have a facilitating effect on low test-anxious students (Morris & Liebert, 1969).  Also, time can have a negative effect on test performance in high test-anxious students because it increases self-reported worry cognitions, which correlates with poor performance (Liebert & Morris, 1967; Morris & Liebert, 1969).

In addition, results from the study showed that both high and low test-anxious individuals made the same number of trials to successfully complete the maze in both the timed and untimed test conditions.  These results were consistent with research by Kellogg et al. (1999), who found that both high and low test-anxious students made the same number of errors when placed in a timed and untimed math test.  Additionally, research by Onwuegbuzie and Seaman (1995) demonstrated that both high and low test-anxious individuals performed comparatively the same when taking an untimed test.  In contrast, research has also indicated that when high and low test-anxious individuals are placed in a timed test, low test-anxious individuals received higher grades than high test-anxious students (Morris & Liebert, 1969; Onwuegbuzie & Seaman, 1995).
In the present study, high test-anxious students worried significantly about how much time was left in the testing period than students who were low test-anxious.  This is consistent with previous research that suggests that high test-anxious individuals experience more self-reported worry cognitions than low test-anxious individuals (Liebert & Morris, 1967; Morris & Liebert, 1969).  In addition, results from the present study indicated that students placed in the timed condition worried significantly more about how much time was left in the testing period than students in the untimed condition.  These results were also supported by previous research (Morris & Liebert, 1969).

Results from the study also found relationships between various posttest questions that were based on self-reported worry cognitions.  For example, there were significant correlations between “My anxiety distracted me from remembering the maze” and between the question “I worried about how much time was left during the testing period.”  And finally, two other significant correlations were found between the question, “I found myself looking at the clock when attempting to complete the maze” and between “I found myself thinking about how well others in my class did on the maze.”  Though none of these relationships can show cause and effect, it is still important to note that each of the worry cognitions were associated with a task-irrelevant behavior such as looking at the clock or worrying about how well others did on the maze.  These correlations support previous research by Mandler and Sarason (1952), which found that individuals exhibit either a learned response or an anxiety response when placed in a test situation.  Individuals who exhibit a learned response will display behaviors that lead to completion of a task.  However, individuals who exhibit an anxiety response will engage in behaviors that do not lead to task completion.  In this case, an anxiety response consisted of looking at the clock and worrying about how well others did on the maze.  Thus, time constraint and negative cognitions are task-irrelevant cues that distract the individual from the task at hand, which correlates with test performance.

Some possible reasons or limitations as to why the present study did not support the predicted hypothesis might be because the sample size was too small and was not truly representative of the general population.  There were also an uneven amount of males and females in the study and the entire population was Caucasian.  All of the students in the study were from an undergraduate psychology statistics course, which is a course known to cause anxiety in many students.  Therefore, students when answering the Test Anxiety Scale (Sarason, 1978b) may have been thinking about their statistics anxiety and not their test anxiety in general.  This may have affected an individuals test anxiety score.  In addition, time spent on task was not measured or accounted for during the completion of the maze when placed in either the timed or untimed test condition.
Moreover, it is possible that the maze itself was not a true or valid measure of test anxiety.  In the past, mazes have been used to look at spatial cognitions and reference and working memory and not as a measure of test anxiety.  Previous research (e.g., Onwuegbuzie & Seaman, 1995) has looked at test anxiety both before and after an academic paper and pencil exam, and not when attempting to complete a virtual reality maze.  The maze used in the present study was not a standardized test of cognitive ability, meaning that it had not been used on the general population.

It is also possible that students did not experience test anxiety when attempting to complete the maze because the maze was not a left hemisphere brain task.  The maze was a visual or spatial task, which is more directly related to functioning in the right-brain hemisphere.  According to the literature, the left hemisphere is more closely related to language sounds, verbal memory, and arithmetic (Matlin, 1995).  In contrast, the right hemisphere is more closely related to recognition of faces, visual memory, and geometry (Matlin, 1995).  It is possible that anxiety was only induced in the right hemisphere; however, no instrument was used in the present study to determine if anxiety was induced in either the right of left-brain hemispheres.
Recent research has found that trait anxiety leads to enhanced left hemisphere activation during negative motivational states (Derryberry & Reed, 1998).  Negative mood states such as frustration are closely associated to right-brain functioning (Rodriguez & Rodriguez, 1994).  Recent research has also suggested that the left hemisphere predominantly modulates the symptoms involved in parasympathetic arousal, whereas the right brain is more closely linked to sympathetic arousal (Everhart et al., 2002).  However, results from a study by Williams (1992) looked at the differences in anxiety levels between individuals who had either a right or left hemisphere stroke.  Individuals who had a right hemisphere stoke reported less anxiety than individuals who had a left hemisphere stroke.  Because the literature is inconsistent, future research could assess the effects of anxiety on both right and left hemisphere tasks in a virtual environment.

Though there was a significant correlation between an individual having test anxiety and feeling nervous when attempting to complete the maze.  Many students stated that the radial arm maze reminded them of a computer game and not of a traditional academic test.  For many students, computers are more representative of Instant Messenger and downloading music as opposed to test taking.  As a result, many of the cues associated with computers may not have evoked a sense of anxiety like a classroom.  However, researchers are now looking into virtual reality, as a way to alleviate test anxiety (Knox et al., 1993).
Aside from the limitations, the present study could have been improved if the sample size was larger and there was an equal distribution of males and females.  It would also be more beneficial if the test were a valid measure of test anxiety.  Mazes are sometimes used for testing reference and working memory and not test anxiety.  So, instead of using the radial arm maze to assess test anxiety, the study should have used a virtual reality math or English test to determine if an individual experienced test anxiety on a virtual reality math or English test.  In addition, another dependent variable, such as accounting for time spent on task, should have been added to the study.  For example, the experimenter should have told one group that they had an unlimited amount of time to complete the test and then stopped them once time had expired.  Thus, accounting for time would have created a more valid study by eliminating time as a confounding variable.

Though the present study did not support the hypothesis, one interesting finding was discovered.  During the first half of the study, many of the students were becoming restless, irritable, and especially frustrated when attempting to solve the maze.  As a result, two questions were added to the posttest questionnaire halfway through the study.  The first question asked whether or not the maze caused the individual to be frustrated.  The second question asked if the individual’s frustration interfered with his or her performance or his or her ability to solve the maze.  Results from a Pearson r Correlation indicated a strong positive relationship between the two-posttest questions.  Some literature has examined the relationship between task difficulty and frustration (e.g., Schmeck & Bruning, 1978) and on error produced frustration (e.g., Schmeck & Anderson, 1974).  Unfortunately, no literature has directly assessed the effects of frustration on either test or maze performance, especially its effects on memory in humans.  Therefore, finding a significant correlation between “The maze caused me to be frustrated” and “My frustration interfered with my performance or my ability to solve the maze” is an important finding because frustration may be found to have a detrimental effect on test performance and more specifically, memory.

Future research could be conducted to assess the effects of frustration on test or maze performance as well as its effects on reference and working memory.  Additionally, future research could look at the effects of time constraint on test performance in a virtual reality setting.  Previous research has only examined virtual reality as an alternative to alleviate test anxiety (e.g., Knox et al., 1993).  Another possible area of interest would be to assess whether individuals with a repressive coping style perform more poorly on a test in comparison to both high and low test-anxious individuals.  Research has been conducted on people who are repressors in relation to examination stress (e.g., Rohrmann et al., 2003), but no published study has examined the differences in test performance between repressors and high and low test-anxious students on a traditional test or in virtual reality.

In addition, with the recent development and widespread use of computer software as well as the development of virtual technology, it is now plausible to simulate a real world experience in the virtual world.  For example, airline pilots now spend hours at a time learning how to fly planes in a virtual reality flight simulator without even stepping foot inside of a plane.  Virtual reality could therefore be used to simulate an actual test experience for a test anxious student.  Virtual reality could also be used to treat test anxiety through systematic desensitization (Knox et al., 1993), which is the most popular way of dealing with and treating test anxiety (Tyron, 1980).  An individual can learn to cope with test anxiety by always being in that situation.  Therefore, future research could assess whether treating test anxiety through systematic desensitization is beneficial in decreasing test anxiety and increasing confidence and test performance.

Axelrod, H.D., Cower, E.L., & Heilizer, F. (1956).  The correlates of manifest anxiety in stylus maze learning.  Journal of
        Experimental Psychology, 131-138.

Cassady, J.C. & Johnson, R.E. (2002).  Cognitive test anxiety and academic performance.  Contemporary Educational
        Psychology, 270-295.

Cattell, R.B. & Scheier, I.H. (1961).  The Meaning and Measurement of Neuroticism and Anxiety.  New York: The Ronald
        Press Company.

Covington, M.V. & Omelich, C.L. (1987).  “I Knew It Cold Before the Exam”: A Test of the Anxiety-Blockage Hypothesis.
        Journal of Educational Psychology, 393-400.

Cruise, R.J., & Wilkins, E.M. (1980).  Stars:  Statistical Anxiety Rating Scale.  Unpublished manuscript, Andrews University,
        Berrien Springs, MI.

Deffenbacher, J.L. (1978).  Worry, Emotionality, and Task-Generated Interference in Test Anxiety: An Empirical Test of
        Attentional Theory.  Journal of Educational Psychology, 248-254.

Deffenbacher, J.L. & Deitz, S.R. (1978).  Effects of Test Anxiety on performance, Worry, and Emotionality in Naturally
        Occurring Exams.  Psychology In the Schools, 446-450.

Deffenbacher, J.L. & Hazaleus, S.L. (1985).  Cognitive, Emotional, and Physiological Components of Test Anxiety.
        Cognitive Therapy and Research, 169-180.

Derryberry, D, & Reed, M.A. (1998).  Anxiety and Attentional Focusing: Trait, State, and Hemispheric Influences.
        Personality and Individual Differences, 745-761.

Eaton, W.O. (1979).  Profile Approach to Longitudinal Data: Test Anxiety and Success-Failure Experience.  Developmental
        Psychology, 344-345.

Everhart, D.E., & Harrison, D.W. (2002).  Heart Rate and Fluency Performance Among High-and-Low-Anxious Men
        Following Autonomic Stress.  International Journal of Neuroscience, 1149-1171.

Farber, I.E., & Spence, K.W. (1953).  Complex Learning and Conditioning as a Function of Anxiety.  Journal of
        Experimental Psychology, 120-125.

Glazeski, R.C., Hollandsworth, J.G., & Jones, G.E. (1986).  An investigation of the role of physiological arousal in test
        anxiety.  Educational and Psychological Research, 67-81.

Gourgey, A.F. (1983).  The relationship of beliefs about mathematics, mathematical self-concept and arithmetic skills to
        mathematics anxiety and performance in basic statistics.  Dissertation Abstracts International, 2284-2285.

Hamilton, D. (2003).  Virtual Radial Arm Maze.  [Computer Software].  Manchester, NH: Saint Anselm College.

Hampson, E., & Kimura, D. (1992).  Sex Differences and Hormonal Influences on Cognitive Function in Humans.  In J.B.
        Becker, S.M. Breedlove, & C. Crews (Ed.), Behavioral Endocrinology.  Cambridge: The MIT Press.

Herman, W.E. (1990).  Fear of Failure as a Distinctive Personality Trait Measure of Test Anxiety.  Journal of Research and
        Development in Education, 180-185.

Hong, E. & Karstensson, L. (2001).  Antecedents of state anxiety.  Contemporary Educational Psychology, 348-367.

Hong, E. (1999).  Test Anxiety, Perceived Test Difficulty, and Test Performance: Temporal Patterns of Their Effects.
        Learning and Individual Differences, 431-455.

Hunsley, J. (1987).  Cognitive processes in mathematics anxiety and test anxiety: The role of appraisals, internal dialogue, and
        attributions.  Journal of Educational Psychology, 388-392.

Kellogg, J.S., Hopko, D.R. & Ashcraft, M.H. (1999).  The Effects of Time Pressure on Arithmetic Performance.  Journal of
        Anxiety Disorders, 591-600.

Kirasic, K.C. (1991).  Spatial Cognition and Behavior in Young and Elderly Adults: Implications for Learning New
        Environments.  Psychology of Aging, 10-18.

Kirasic, K.C., Allen, G.L., & Haggerty, D. (1992).  Age Related Differences in Adults Macrospatial Cognitive Processes.
        Experimental Aging Research, 33-39.

Knox, D., Schacht, C.,& Turner, J. (1993).  Virtual Reality: A Proposal for Treating Test Anxiety in College Students.
        College Student Journal, 294-296.

Lazarus, R.S., & Folkman, S. (1984).  Stress Appraisal and Coping.  New York: Springer.

Liebert, R.M., & Morris, L.W. (1967).  Cognitive and emotional components of test anxiety: A distinctive and some initial
        data.  Psychological Reports, 975-978.

Linn, M.C., & Peterson, A.C. (1985).  Emergence and Characteristics of Gender Differences in Spatial Abilities: A
        Meta-Analysis.  Child Development, 1479-1498.

Mahoney, M.J. (1978).  Experimental Methods and Outcome Evaluation.  Journal of Consulting and Clinical Psychology,
        660-672.

Mandler, G. & Sarason, S.B. (1952).  A Study of Anxiety and Learning.  Journal of Abnormal and Social Psychology,
        166-175.

Matlin, M.W.  (1995).  Psychology Second Edition.  New York, NY: Harcourt Brace College Publishers.

Minor, S.W. & Gold, S.R. (1986).  Behavior of test anxious students across time.  Personality and Individual Differences,
        241-242.

Moffat, S.D., Hampson, E., & Hatzipantelis, M. (1998).  Navigation in a “Virtual” Maze: Sex Differences and Correlation
        With Psychometric Measures of Spatial Ability in Humans.  Evolution and Human Behavior, 73-87.

Moffat, S.D., Zonderman, A.B., & Resnick, S.M. (2001).  Age Differences in Spatial Memory In a Virtual Environment
        Navigation Task.  Neurobiology of Aging, 787-796.

Morris, L.W. & Liebert, R.M. (1969).  Effects of Anxiety on Timed and Untimed Intelligence Tests: Another Look.  Journal
        of Consulting and Clinical Psychology, 240-244.

Morris, L.W., Davis, M.A., & Hutchings, C.H. (1981).  Cognitive and emotional components of anxiety: Literature review
        and a revised worry-emotionality scale.  Journal of Educational Psychology, 541-555.

Newman, M., & Kaszniak, A. (2000).  Spatial Memory and Aging: performance on a Human Analog of the Morris Water
        Maze.  Aging , Neuropsychology and Cognition, 86-93.

Nijhawan, H.K. & Cheema, P. (1971).  Maze Learning Under Stress In Normal and High Test Anxious Children.  Journal of
        Indian Academy of Applied Psychology, 23-29.

Onwuegbuzie, A.J., & Seaman, M.A. (1995).  The effect of time constraints and statistics test anxiety on test performance in a
        statistics course.  Journal of Experimental Education, 115-120.

Osterhouse, R.A. (1972).  Desensitization and Study-Skills Training as Treatment  for Two Types of Test-Anxious Students.
        Journal of Counseling Psychology, 301-307.

Rodriguez, I., Rafael, N., & Rodriguez, L. (1994).  Hemispheres’ modes of cognitive functions in a Finnish school: Grades and
        Behaviour.  Educational Psychology, 207-217.

Rohrmann, S., Netter, P., Hennig, J., & Hodapp, V. (2003).  Repression-Sensitization, gender, and Discrepancies in
        psychological reactions to examination stress.  Anxiety, Stress, & Coping: An International Journal, 321-329.

Sarason, S.B., Mandler, G., & Craighill, P.G. (1952).  The Effect of Differential Instructions on Anxiety and Learning.
        Journal of Abnormal Social  Psychology, 561-565.

Sarason, I.G. (1972).  Experimental Approaches to Test Anxiety: Attention and the Uses of Infornmation.  In C.D.
        Spielberger (Ed.), Anxiety: Current Trends in Theory and Research (Vol. 2).  New York: Academic Press.

Sarason, I.G., & Stoops, R. (1978a).  Test Anxiety and the Passage of Time.  Journal of Counseling and Clinical
        Psychology, 102-109.

Sarason, I.G. (1978b).  Test Anxiety Scale: Concept and Research. In C.D. Spielberger & I.G. Sarason (Eds.), Stress and
        Anxiety (Vol. 5). Washington D.C.: Hemisphere Publishing Corp.

Sarason, I.G. (1984).  Stress, anxiety, and cognitive interference: Reactions to Tests.  Journal of Personality and Social
        Psychology, 46, 929-938.

Schmeck, R.R., & Bruning, J.L. (1968).  Task Difficulty and the Frustration Effect.  Journal of Experimental Psychology,
        516-520.

Schmeck, R.R.,& Anderson, V. (1974).  Error-Produced Frustration: Additional Data.  Perceptual & Motor Skills, 1012.

Schwarzer, R., Van der Ploeg, H.M. & Spielberger, C.D. (1982).  Advances in Test Anxiety Research (Vol. 1).  Hillsdale,
        NJ: Lawrence Erlbaum Associates, Inc.

Schwarzer, R. (1984).  Worry and emotionality as separate components in test anxiety.  International Review of Applied
        Psychology, 205-220.

Siegman, A.W. (1956).  The Effect of Manifest Anxiety on a Concept Formation Task, a Nondirtected Learning Task, and
        on Timed and Untimed Intelligence Tests.  Journal of Consulting Psychology, 176-178.

Smith, M.K. (1985).  The evolving structure of statistical knowledge in high and low math anxious students.  Dissertation
        Abstracts International, 3983.

Taylor, J.A. (1953).  A Personality Scale of Manifest Anxiety.  Journal of Abnormal Social Psychology, 285-290.

Tyron, G.S. (1980).  The Measurement and Treatment of Test Anxiety.  Review of Educational Research, 343-372.

Spielberger, C.D. (1966).  Theory and Research on Anxiety.  In C.D. Spielberger (Ed.), Anxiety and Behavior.  New York:
        Academic Press, 1966, p. 3-20.

Spielberger, C.D. (1970).  Test Anxiety Inventory: Preliminary Professional Manual.  Palo Alto, CA: Consulting
        Psychologists Press.

Spielberger, C.D. (1980).  Test Anxiety Inventory: Preliminary Professional Manual.  Palo Alto, CA: Consulting Psychologists
        Press.

Spielberger, C.D. & Vagg, P.R. (1995).  Test Anxiety Theory, Assessment, and Treatment.  Washington D.C.: Taylor &
        Francis.

Wentzel, C.A. (1999).  A comparison of statistics anxiety and mathematics anxiety among graduate students in the social
        sciences.  Dissertation Abstracts International, 2349.

Wiederhold, B.K. (1998).  A Review of Virtual Reality as a Psychometric Tool.  Cyberpsychology & Behavior, 45-52.

Williams, A.M. (1992).  Self-report of indifference and anxiety among persons with right hemisphere stroke.  Research in
        Nursing and Health, 343-347.

Wine, J. (1971).  Test Anxiety and Direction of Attention.  Psychological Bulletin, 92-104.

Wise, S.L., Plake, B.S., Eastman, L.A., Boettcher, L.L., et al. (1986).  The effects of item feedback and examinee control on
        test performance and anxiety in a computer-administration test.  Computers in Human Behavior, 21-29.

Zeidner, M. (1991).  Statistics and mathematics anxiety in social science students: Some interesting parallels.  British Journal
        of Educational Psychology, 319-328.

Table 1
Independent Samples t-test conducted on test anxiety score and posttest questions.

I worried about how much time                                    3.461                4.454                2.231             .036*
was left during the testing period.                                 (sd=1.198)         (sd=.934)

I found the maze very difficult to                                   3.200                 4.333                1.969             .080
complete.                                                                   (sd=1.303)          (sd=.5164)

Significant and marginal differences were found by an independent samples t-test on test anxiety score and the posttest questionnaire.

Table 2
Independent samples t-test was conducted on grouping and posttest questions.

I worried about how much time                                     3.500             4.333               1.820              .082
was left during the testing period.                              (sd=1.446)         (sd=.651)

Marginal significance was found by an independent samples t-test on grouping and a question on the posttest questionnaire.

Appendix A

Instructions to be read prior to informed consent:
This study being conducted is a virtual reality 8-arm maze test that has been considered by some to measure for non-verbal analytical skills and is highly correlated with mathematics ability.   Research has suggested that mathematics ability can predict statistics achievement (Gourgey,
1983; Smith, 1985; Wise et al., 1986; Hunsley, 1987; Zeidner, 1991;Wentzel, 1999).  In terms of course credit, participating in the research study is optional and not required.  You can either replace a quiz as long as you submit a one-paragraph report on what the study was about/what you learned.  Or, you can take a traditional stats quiz instead of being in the study.   If you would like to be in the study, please carefully read the informed consent and sign your name at the bottom.

Appendix B

INFORMED CONSENT AND RIGHTS OF RESEARCH PARTICIPANTS
IN THE DEPARTMENT OF PSYCHOLOGY AT SAINT ANSELM COLLEGE

All psychological research at Saint Anselm College is conducted according to strict ethical principals outlined by the American Psychological Association and is in full compliance with Federal law.  The Department of Health and Human Services, for example, specifies that informed consent must be given prior to research studies, that is, "…the knowing consent of an individual or his legally authorized representative so situated as to be able to exercise free power of choice without undue inducement or any element of force, fraud, deceit, duress, or other form of constraint or coercion."

Simply put, this means when you participate in any research study, you will be given a clear explanation of the procedures involved.  You may ask for clarification either before or during the procedure, and you may terminate the procedures at any time.

This research study requires you to fill out three questionnaires as well as complete a virtual reality task.  This study is using virtual reality, which may make an individual feel uncomfortable.  Students participating in this research study may feel nervous, anxious, dizzy, or nauseous.   While there is a chance for harm, no records have suggested that the virtual reality maze that is being used will evoke such harm. However, if you experience any of these uncomfortable symptoms, please notify the instructor immediately and stop what you are doing and the session will be terminated.

After having carefully read and considered the foregoing, I consent to participate in research activities according to the terms heretofore enumerated.  My signature indicates that I understand the instructions of this study as they have been read to or read by me.

Date_______________________                                  Signature___________________________
Class/Student I.D. #_________________________     Other_________________
 

=====================================================================

Appendix C

Demographic Information:
Please provide the following information:

I.  First:                              Last:

II.  Last 4 digits of your SS#:                                                Age:

III.  Expected Graduation Year:

IV.  Gender:            Male            Female

V.  What state are you from?

VI.  Race/Ethnicity:              White/Caucasian                                    Black
                                            Asian                                                    Hispanic
                                            Native American/Indian                        Other (Specify Please)

VII.  Religion:                       Catholic                                                Protestant
                                             Lutheran                                              Mormon
                                             Jewish                                                  Islamic
                                                                                                        Other (Specify please)

VIII.  Major:                         Undeclared                                          History
                                             Science (bio, chem.)                            A Foreign Language
                                             Psychology                                          Nursing
                                             Politics                                                 English
                                             Business/Economics                             Sociology
                                             Philosophy                                           Theology
                                             Computer Science                                Art
                                                                                                         Other (Specify Please)

Appendix D

Instructions for the T Group
This study being conducted is a virtual reality 8-arm maze test that has been considered by some to measure for non-verbal analytical skills and is highly correlated with mathematics ability.   Research has also suggested that mathematics ability can predict success in statistics (Gourgey, 1983; Smith, 1985; Wise et al., 1986; Hunsley, 1987; Zeidner, 1991;Wentzel, 1999).  Therefore, it is imperative that you try your hardest to successfully complete the maze.
The computer in front of you is a virtual reality maze that has 8 arms.  In this version, four of the arms contain items.  You have an unlimited amount of trials to solve this maze.  Each trial ends when you have found the four items.  For each trial of the experiment, the items will always stay in the same arms.  To solve the maze, you must learn the arms in which the items are located.  You must do this by completing two trials in a row where you enter only the arms that contain items and do not re-enter the arms once you have retrieved that item.  It is important to note that you will not be able to determine if an object is at the end of each arm until you hit the wall at the end of each arm.  If you hit the end of the arm and the number 1 does not appear at the bottom of the screen, then no object exists in the arm.
When the maze begins, use the arrow key to move forward, and the left and right arrow keys to move left or right and to turn around.  You must go all the way to the wall at the end of each arm to determine if it contains an item.  If you have any questions please ask the administrator.  Virtual reality tasks can make a few people feel uncomfortable.  If you feel uncomfortable with this task at any time, please notify the administrator.  You have 15 minutes to successfully complete the radial arm maze.  Good luck!

Appendix E

Instructions for the UT Group
This study being conducted is a virtual reality 8-arm maze test that has been considered by some to measure for non-verbal analytical skills and is highly correlated with mathematics ability.   Research has also suggested that mathematics ability can predict success in statistics (Gourgey, 1983; Smith, 1985; Wise et al., 1986; Hunsley, 1987; Zeidner, 1991;Wentzel, 1999).  Therefore, it is imperative that you try your hardest to successfully complete the maze.
The computer in front of you is a virtual reality maze that has 8 arms.  In this version, four of the arms contain items.  You have an unlimited amount of trials to solve this maze.  Each trial ends when you have found the four items.  For each trial of the experiment, the items will always stay in the same arms.  To solve the maze, you must learn the arms in which the items are located.  You must do this by completing two trials in a row where you enter only the arms that contain items and do not re-enter the arms once you have retrieved that item.  It is important to note that you will not be able to determine if an object is at the end of each arm until you hit the wall at the end of each arm.  If you hit the end of the arm and the number 1 does not appear at the bottom of the screen, then no object exists in the arm.
When the maze begins, use the arrow key to move forward, and the left and right arrow keys to move left or right and to turn around.  You must go all the way to the wall at the end of each arm to determine if it contains an item.  If you have any questions please ask the administrator.  Virtual reality tasks can make a few people feel uncomfortable.  If you feel uncomfortable with this task at any time, please notify the administrator.  You have an unlimited amount of time to successfully complete the radial arm maze.  Good luck!

Appendix F

Post-Test Questionnaire:
The purpose of this questionnaire is to determine if anxiety was or was not induced while attempting to complete the virtual reality maze.  Please read thoroughly and answer each question as it appears on the page.

STRONGLY DISAGREE……...5
DISAGREE…………………….4
UNDECIDED………………….3
AGREE………………………...2
STRONGLY AGREE……….....1

1.  I traditionally get nervous during an evaluative
situation such as when taking a test…………………………………………..1     2     3     4     5

2.  I was nervous when attempting to complete the maze…………………….1     2     3     4     5

3.  My anxiety distracted me from remembering the maze………..…………..1     2     3     4     5

4.  I found myself looking at the clock when attempting to complete
the maze…………………………………………………………………....1     2     3     4     5

5.  I worried about how much time was left during the testing period………....1     2     3     4     5

6.  I was relaxed when attempting to solve the maze………………………....1     2     3     4     5

7.  I found myself thinking about how well others in my class
did on solving the maze……………………………………………………...1     2     3     4     5

8.  The maze caused me to be frustrated…………………………………….1     2     3     4     5

9.  My frustration interfered with my performance or my ability
to solve the maze…………………………………………………………....1     2     3     4     5

10.  I found the maze very difficult to complete……………………………....1     2     3     4     5

Appendix G

Test Anxiety Scale (Sarason, 1978b)

Appendix H

Debriefing Statement:
At this time I would like to thank you for participating in my experiment.  One of the purposes of the study was to determine if individuals who normally experience test anxiety on a traditional paper and pencil test would also experience test anxiety in a virtual reality setting.   The second purpose was to ascertain if time constraint would have either a detrimental or beneficial effect on high anxious and low anxious individuals in relation to test performance.  Therefore, this study was not about non-verbal analytical skills in relation to mathematics ability and statistics achievement. This study examined the effects of time constraint on high and low test-anxious psychology students in a virtual reality setting.
In order to keep this as being a valid study, I would ask that you do not disclose or discuss any of the information that I have now given to you about the study, until all of data are collected.  If the purpose, hypothesis, and goals of my study were to be known by a future participant, that individual might be biased to answering a question in a particular way.  This would in fact affect the results of my study.  If you have questions or comments regarding my study, or would like to know the results, you can contact me at pfrancin@anselm.edu.
Thank you for your cooperation and participation!