Professor Petger Schaberg Academic Article The Effects of a

Professor Petger Schaberg
Academic Article
The Effects of a Positive Writing Task on Test Anxiety Immediately Before an Exam
Abstract
Test anxiety is a common experience on college campuses. Many treatments for test
anxiety have been developed, but few have focused on creating a brief intervention that could be
implemented immediately before an exam. Studies have focused on test anxiety as it pertains to
tasks performed in a lab, rather than in realistic testing environments, and on long-term
treatments instead of brief interventions. A brief, effective intervention for test anxiety, able to
be administered immediately before an exam, could be easily included at the beginning of any
test. This type of intervention has the potential to help large numbers of test anxious students in
a short amount of time. The present study tested the effect of writing about a successful
experience on symptoms of test anxiety immediately preceding a regularly scheduled exam. 30
undergraduate students arriving for a chemistry exam participated. Subjects were randomly
assigned to a control or treatment group. The control group was directed to write about their
morning routine, while the treatment group was directed to write about successfully overcoming
a challenge. Anxiety in males versus females was also examined. No significant difference was
found between the treatment and control groups nor was a significant difference found between
males and females. Possible explanations for these findings and future research directions are
discussed.
Introduction
Test Anxiety
Test anxiety has come to be a common experience in modern society as academic
achievement is increasingly used to evaluate individuals in the professional world. The adverse
effects of test anxiety often result in an underestimation of one’s abilities and reduced access to
educational and occupational opportunities (Spielberger and Vagg 1995). Most people
experience the physiological and psychological components of test anxiety at some point or
another. Common physiological symptoms include sweating, rapid pulse, “butterfilies” in the
stomach, and other uncomfortable sensations. Psychological symptoms include feeling
distracted, worried, incompetent, and self-critical. Test anxiety has been found to be negatively
correlated with performance in experimental and normal student populations; this correlation is
strongest for the cognitive components of test anxiety (Pekrun 2001). It is thought that selfcentered, task-irrelevant worry cognitions are detrimental to performance because they occupy
essential cognitive resources (Spielberger and Vagg 1995). Because cognitive aspects of test
anxiety are so influential to performance, the present study uses a writing task designed to alter
negative cognitions as a means of combating the negative impact of test anxiety. Negative selfview has been found to be a particularly strong predictor of test anxiety as well (Wong 2008).
This finding supports the hypothesis that writing about a positive experience and overcoming a
challenge will reduce test anxiety by increasing confidence in addition to reducing the cognitive
load.
Writing Interventions
Several writing exercises have been shown to alleviate test anxiety and improve
performance on exams. A previous study done at the University of Colorado found that a
writing exercise involving values affirmation helped female students perform better in physics
classes, narrowing the gender gap in science achievement (Miyake et al. 2010). Women who
wrote about their most important values at the beginning of the course scored an average of one
full letter grade above women in the control group. While Miyake’s study did not focus on the
relationship between the values affirmation task and anxiety, it did reveal that a simple writing
task can dramatically improve academic performance.
Lumley and Provenzano (2003) found that anxious students earned higher grades when
they wrote about their worries and fears at the beginning of the semester. This was thought to be
due to the fact that disclosing worries freed cognitive processes and improved mood. Similarly,
Ramirez and Beilock (2011) found that taking ten minutes before an exam to write about worries
improved exam scores in test anxious students. The Ramirez and Beilock study was especially
significant because the writing exercise was administered immediately before a regularly
scheduled exam. This finding indicates that other types of writing interventions for test anxiety
may also be successfully administered immediately before exams.
Nelson and Knight (2010) found that participants who completed a positive-thought
writing task before a quiz exhibited more positive affect, less test anxiety, better performance on
the quiz, and were more likely to appraise the quiz confidently than participants in a control
group. Their findings were in agreement with previous theories that positive affect, optimism,
and self efficacy may be effective coping mechanisms in stressful situations, and the writing
exercise they devised seemed to employ these mechanisms. Because Nelson and Knight found
such encouraging results from a simple writing exercise, this task was chosen to be tested in the
current study. One limitation to Nelson and Knight’s findings was that the stressor was a 15question pop quiz. This stressor is very different from the lengthy and heavier-weighted exams
that college students encounter more frequently. The present study chose to administer this
writing task to students in the minutes before a 160 point exam worth 15% of the total class
grade. This setting more accurately represents the test anxiety most typical to undergraduates at
the University of Colorado.
Methods
Procedure
A convenience sample of 30 students (11 females and 19 males) from an introductory
chemistry class at the University of Colorado, Boulder was used. Participants were all
undergraduates with ages ranging from 18 to 24 years old. Participation was voluntary and
anonymous. All research methods were approved by the Institutional Review Board for class
projects involving human research subjects. Participants were randomly assigned to one of two
writing activities. The treatment group was assigned to write about a past experience in which
they successfully overcame a challenge and experienced feelings of pride and joy. The control
group was assigned to write about their morning routine. After the writing assignment
participants completed a brief survey measuring physiological and cognitive symptoms of
anxiety. The writing activities were completed immediately before a regularly scheduled exam
in order to gain an accurate measure of test anxiety in a realistic environment.
Measurement
The Symptoms of Test Anxiety Scale (STAS) created by Berk and Nanda (2006) was
used as a measure of anxiety. This scale incorporates aspects of five of the most frequently used
measures of test anxiety in a very concise checklist format. The various measures used to
formulate the STAS were the Test Anxiety Scale, Reactions to Tests Scale, Revised Test Anxiety
Scale, and Cognitive Test Anxiety Scale (Berk and Nanda 2006). These scales all measure
various physiological and cognitive dimensions of test anxiety, but are lengthy and complex.
The comprehensive STAS was used because its brevity and ease allows it to be administered
immediately before an exam. Subjects were directed to put a check next to any symptoms they
were currently experiencing. Each check was scored as 1 and each blank as 0. A few control
symptoms were also included (e.g. relaxed, confident) and scored as -1 if checked. Subjects
were also asked to report if they were sick the day of the exam. Because symptoms of anxiety
can be similar to symptoms of illness, results from subjects who reported being sick were not
included in the study.
Statistical Analysis
A two-way ANOVA was used to examine the relationships between the writing task and
gender with anxiety scores. The statistical program R was used to run the ANOVA. Post- hoc
tests were used to reveal more about specific relationships. A challenge involved in the
statistical analysis of this data was that the sample was unbalanced. Balance refers to the number
of subjects in each subgroup of the study. The female control group had five subjects, the female
treatment group had three subjects, the male control group had six subjects, and the male
treatment group had eight subjects. This lack of balance may slightly affect the outcome of the
ANOVA, but correcting this problem requires much more advanced software than was available
for this study.
Results
There was no significant difference found between the treatment group and the control
group (p=.5306). The mean number of symptoms of test anxiety reported by the treatment group
was 12.92 and the mean for the control group was 11.65. Additionally, there was no significant
difference found for the main effect of gender on test anxiety (p=.7953). The mean for males
was 12.0 and the mean for females was 12.57. There was also no significant interaction between
gender and writing task group on symptoms of test anxiety (p=.5578). Details of the results are
displayed in the following tables and graphs.
Mean Number of Test Anxiety Symptoms Reported
Control
Treatment
Females
12.8
12.33
12.57
Males
10.5
13.5
12.0
11.65
12.92
Table 1
Anova Table (Type II tests)
Sum Sq
Df
F value
Pr (>F)
Gender
2.83
1
0.0693
0.7953
Group
16.70
1
0.4090
0.5306
Gender:Group
Residuals
14.57
734.97
1
18
0.3568
0.5578
Table 2
Plot of Means
10
12
14
c
tx
8
mean of TestAnxiety1$score
16
TestAnxiety1$group
f
m
Figure 1 The x-axis depicts gender. The treatment group is represented by the red triangles and
the control group is represented by black
circles.
TestAnxiety1$gender
Discussion
Possible Explanations for Results
Because previous studies (Nelson and Knight 2010; Miyake et al. 2010; Lumley and
Provenzano 2003) have found similar writing exercises to be effective in reducing test related
anxiety and improving academic performance, it was surprising that the positive writing task in
this study had no effect. It appears that this specific writing task is not an effective intervention
immediately before an exam. One explanation for this may be that the task required too much
cognitive effort. Subjects had to remember a positive experience in which they overcame an
obstacle, describe their feelings, and explain what this experience meant for them. It is possible
that this task provoked already stressed cognitive resources. This could have resulted in added
stress or prevented subjects from thoroughly completing the task. It may be that this type of
treatment is most effective when administered in a state of lower anxiety, a few hours or days
prior to the exam.
Another explanation for this lack of treatment effect in this study is the possibility that
the control task had an unexpected effect on anxiety. One hypothesis is that in a state of high
anxiety, such as immediately before an exam, an easy task may reduce symptoms of anxiety and
increase confidence and feelings of accomplishment. It requires very little cognitive effort to
describe one’s typical morning routine. This task can also be completed in list form, rather than
complete sentences. Such an easy task may actually reduce the cognitive load and have a
calming effect.
Possible Causes of Type I Error
One possible source of error in this study was the lack of control over the environment.
Because the limited time frame required several people to complete the study at the same time,
subjects would occasionally talk to one another during the study, make jokes, or fail to
thoroughly complete the task. Many subjects were also in a hurry to find a seat or do some last
minute studying. Clear directions were provided on the worksheets, yet six participants failed to
complete the writing task, only checking symptoms of anxiety. The results from these six
participants were not included. After throwing out these scores, and those of participants
reporting illness, the total number of subjects in the study was brought to 22, causing the sample
to be rather unbalanced. Of the participant who followed directions, only half of the treatment
group thoroughly completed the writing task, possibly masking any benefit thorough completion
may have had.
Limitations
Time and money were the main limitations of this study. The short time frame allotted
for this experiment, required the fastest type of approval from the Institutional Review Board.
The class project approval was granted very quickly, but required that all participant information
be kept anonymous. This meant that data on test performance could not be measured. In similar
studies, researchers were able to offer students extra credit on the exam and obtain 100%
participation rates. Because this was an undergraduate class project, offering extra credit was not
possible. With zero funding for this experiment, there was little else to offer participants. There
was also little incentive for volunteers to help run the study. As a result, many test-anxious
students declined to participate and research volunteers with no background in conducting
psychological studies were used.
Future Research Directions
Perhaps the most interesting finding in this study was the indication that the control task
may reduce anxiety. Future investigations could examine this possibility more closely. It is
unclear whether the specific task of writing about one’s morning routine is calming, or if the
simple act of performing any easy cognitive task reduces anxiety. If the latter is true, placing a
few easy questions at the beginning of an exam could prove to be an effortless way to reduce
anxiety and improve performance.
Future studies may also look to compare the effectiveness of several brief interventions
or combinations of interventions that could be performed immediately before and exam. An
example of this would be a comparison of values affirmation, worry disclosure, and positive-
thought writing tasks. While this study may have failed to identify an effective and brief
intervention for test anxiety, it has provided interesting insight into the direction of further
research, and added to the increasing knowledge about the complex nature of test anxiety.
Appendix A. Symptoms of Test Anxiety Scale (STAS; Berk and Nanda 2006)
References
Berk, R.A. & Nanda, J. (2006). A randomized trial of humor effects on test anxiety and test
performance. Humor: International Journal of Humor Research, 19(4), 425-454.
Hembree, R. (1988). Correlates, causes, effects, and treatment of test anxiety [Abstract]. Review
of Educational Research, 58(1), 47-77.
Lumley, M.A. & Provenzano, K.M. (2003). Stress management through written emotional
disclosure improves academic performance among college students with physical
symptoms. Journal of Educational Psychology, 95(3), 641-649.
Miyake, A., Kost-Smith, L.E., Finklestein, N.D., Pollock, S.J., Cohen, G.L., & Ito, T.A. (2010).
Reducing the gender achievement gap in college science: A classroom study of values
affirmation. Science, 330(6008), 1234-1237.
Nelson, D.W. & Knight, A.E. (2010). The power of positive recollections: Reducing test anxiety
and enhancing college student efficacy and performance. Journal of Applied Psychology,
40(3), 732-745.
Pekrun, R.(2001). Test anxiety and academic achievement. In International Encyclopedia of the
Social and Behavioral Sciences (Vol. 23, pp. 15610- 15614). Oxford, UK: Elsevier
Science, Ltd.
Putwian, D.W., Woods, K.A., & Symes, W. (2010). Personal and situational predictors of test
anxiety of students in post-compulsory education. British Journal of Educational
Psychology, 80, 137-160.
Putwain, D.W., Connors, L., & Symes, W. (2010). Do cognitive distortions mediate the test
anxiety—examination performance relationship? Educational Psychology, 30(1), 11-26.
Ramirez, G. & Beilock, S.L. (2011). Writing about testing worries boosts exam performance in
the classroom. Science, 331(6014), 211-213.
Sapp, M. (1999). Test anxiety: Applied research, assessment, and treatment interventions (2nd
ed.). Lanham, MD: University Press of America.
Spielberger, C.D. & Vagg, P.R. (1995). Test anxiety: Theory, assessment, and treatment. Bristol,
PA: Taylor & Francis.
Wong, S.S. (2008). The relations of cognitive triad, dysfunctional attitudes, automatic thoughts,
and irrational beliefs with test anxiety. Current Psychology, 27, 177-191.
Zeidner, M. (1998). Test anxiety: The state of the art. New York, NY: Plenum Press.