Running head: TEST ANXIETY 1

The Affect of Test Anxiety on Student Performance on Tests

Alaa K. Abd-El-Hafez
Long Island University


April 2014
TEST ANXIETY 2
The Affect of Test Anxiety on Student Performance on Tests
Over the past decade, there has been increased testing in schools and pressures associated
with it (Lee, Lowe, & Sena, 2007; McDonald, 2012). Legislation, such as No Child Left Behind
and Race to the Top, have created test based accountability systems, increasing the importance
placed on student test outcomes. Since school and student performance are determined through
the use of high-stakes tests, it is critical to examine any variable which may interfere with the
valid measurement of student achievement and school effectiveness (Chamberlian, Daly, &
Spalding, 2011; Embse & Hasson, 2012; Lee et al., 2007). One such variable is test anxiety.
Test anxiety has been shown to impair test performance by many studies (DiBattista & Gosse,
2006; Embse & Hasson, 2012; Mcdonald, 2001; Putwain, 2008). However, the relationship is not
simply linear and often moderated by other factors and interactions (Chamberlian et al., 2011;
McDonald, 2001). This paper will explore the impact of test anxiety on student performance on
exams using the literature from quantitative research studies.
Test Anxiety
Test anxiety is the set of cognitive, physiological, and behavioral responses related to
concerns about possible failure or poor performance on an exam or a similar evaluative
situation (Bodas, Ollendick, & Sovani, 2008, p. 387). Test anxiety, however, is not a
diagnosable disorder; yet many students from elementary school to college have been reported to
experience test anxiety. One study has reported that 25% to 40% of school children experience
test anxiety (Bodas et al., 2008). Another study, Barterian, Carlson, Goforth, Embse, and Segool
(2013), have reported lesser percentage of school-aged children experiencing test anxiety, 10%
to 30%. The study also reported that about 10% of children are highly test anxious and
TEST ANXIETY 3
experience impairment in test performance as a result. DiBattista and Gosse (2006) stated that
the percent of college students who experience test anxiety vary from 10% to 15%.
Test anxiety was found to be more prevalent among minority ethnic groups (Putwain,
2008). Specifically, higher levels of test anxiety were found among African American
elementary school-aged children with a reported percentage of 41%. Female students report
higher level of test anxiety than male students (Barterian, Carlson, Goforth, Embse, and Segool,
2013). Students with learning disabilities were more likely to be test anxious. They were also
more likely to experience the debilitating effects of test anxiety. Test anxiety was found to be a
significant predictor of their reading and mathematics achievement scores (Lee et al., 2007).
Research has also shown that test anxiety is prevalent in many cultures across the globe with
higher levels of test anxiety reported from Islamic culture than from Western European cultures
(Bodas et. al., 2008; Putwain, 2008).
In educational testing, test anxiety may obstruct the true potential of students. With an
increase in testing and accountability of students, teachers, and schools, a variable such as test
anxiety must be explored to ensure the genuine measurement of performance (Chamberlian,
Daly, & Spalding, 2011; Embse & Hasson, 2012; Lee et al., 2007). Research on test anxiety and
its impact has flourished. In fact, research on test anxiety is more widespread than it appears
because test anxiety is sometimes hidden under broader constructs (Pekrun & Stober, 2004). Of
the multiple dimensions of test anxiety, it is the worry component that is related to lower
examination performance (Eum and Rice 2010; Putwain, 2008). Some studies have shown that
worry and sensitivity to distraction independently predict performance on examination (Lee et
al., 2007; Pekrun & Stober, 2004).
Historical Overview of Test Anxiety
TEST ANXIETY 4
Studies related to test anxiety were conducted as early as 1914 but first discussed in depth
in 1932 by Luria (Burns, 2004; Pekrun and Stober, 2004). It had received considerable attention
since then. The relation between test anxiety and test performance was first examined in 1952
and a negative relationship was reported (Pekrun and Stober, 2004). Higher test scores were
found to be correlated with lower test performance (Burns, 2004). Highly test-anxious
individuals were more self-critical and more likely to experience performance-interfering worry
during examinations than were individuals who were observed to be low in test anxiety (Burns,
2004, p. 119). Furthermore, highly test-anxious students were three times more likely to drop
out of college (Burns, 2004). The Test Anxiety Questionnaire was developed to assess
individual differences in adults and the Test Anxiety Scale for Children (Pekrun and Stober,
2004).
In the 1960s and early 1970s, several contributions were made to test anxiety research.
The first contribution was a distinction between test anxiety as a transitory state and as a stable
personality trait. A second contribution is a distinction between the four dimensions of test
anxiety: (a) worry; (b) emotionality; (c) interference; and (d) lack of confidence (DiBattista &
Gosse, 2006; Pekrun & Stober, 2004; Lee et al., 2007). Advances in test anxiety and its effect on
cognitive performance model construction occurred in the 1970s and 1980s. Since 1952, the
number of publications on test anxiety exceeded 1000. It has been shown by educational
psychologists that test anxiety is an important factor in primary, secondary, and tertiary
education (Pekrun & Stober, 2004).
Test Anxiety and Working Memory
Test anxiety is a complex multi-dimensional, and dynamic construct (Bodas et. al,
2008, p. 388). After reviewing the literature, Mowbray (2012) found that individuals experience
TEST ANXIETY 5
test anxiety in two dimensions of emotionality and cognitions (worry). The cognitive aspect of
test anxiety was found to have the most impact/connection with test performance. According to
attentional control theory, anxiety impacts the ability to (a) suppress task-irrelevant stimuli
(inhibition); (b) shift attention between multiple tasks; and (c) the phonological loop. Anxiety, in
general, is believed to promote bottom-up attentional processes. That is, anxiety results in an
involuntary attention shift toward relevant sensory events such as perceived threat related
stimuli. Unable to disengage from such stimuli, ones attention is diverted and loses focus on a
task. This is referred to in the literature as retrieval failure test anxiety. Students who know the
material fail to recall it on a test due to cognitive interference (Mowbrary, 2012).
Attentional control theory states that anxiety reduces inhibition, overwhelming the
working memory with task-irrelevant thoughts. Also, highly anxious students can only work at
one task at a time. Participants with high levels of test anxiety were found to take longer to
complete tasks including examination tasks, especially if that task required a great deal of
attention. Students with the same level of test anxiety are not affected similarly, those with
greater working memory capacities are better buffered against performance deficits of test
anxiety. Highly anxious students also showed deficits in encoding study material. This is due to
interference with the phonological loop. The relationship between test anxiety and exam
performance can only be explained by both failure to obtain, encode, and organize exam material
and failure to retrieve previously learned information during a testing situation (Broader &
Musch, 1999; Mowbrary, 2012).
There are several limitations to Mowbrarys (2012) study. First, attentional control
theory addresses anxiety in general rather than specifically test anxiety. Anxiety is a diagnosable
disorder but test anxiety is not (Bodas et al., 2008). Research has shown that instruments
TEST ANXIETY 6
measuring anxiety in general are not necessarily applicable to test anxiety (Burns, 2004;
Mowbrary, 2012). Second, working memory has been shown to be unrelated to test anxiety.
Third, the interventions (cognitive therapies) offered in the study do not have a lasting effect,
further reduce efficiency, and can actually be detrimental to exam performance (Mowbrary,
2012).
Test Anxiety at the College Level
Burns (2004) found no significant relationships between anxiety at the time of an exam
and performance on that exam. There was also no significant relationship between anxiety at the
time of a test and the final grade for a course. However, a positive relationship was found
between performance expectations at a test and levels of anxiety at that time, higher expectations
resulted in higher anxiety levels. Positive relationships were also found between the scores on
previous exams and the level of anxiety on the subsequent exam. That is, the higher the scores
on previous exams, the higher the anxiety level on the next exam. The study also did not find a
significant relationship between poor preparedness and number of absences and anxiety at the
time of an exam (Burns, 2004).
In contrast to the findings found in Burns (2004), DiBattista and Gosse (2006) found a
negative relationship between the level of test anxiety, as measured by Revised Test Anxiety
Scale (RTAS) scores, and various measures of test performance. However, DiBattista and Gosse
(2006) found that immediate feedback, through using the Immediate Feedback Assessment
Technique (IFAT) answer form, on an exam did not increase the levels of test anxiety in students
but actually reduced it for the majority of them. Furthermore, students preference for the IFAT
was not related to test anxiety or any measure of test performance (DiBattista and Gosse, 2006).
TEST ANXIETY 7
Broder and Musch (1999) compared how two rival models of test anxiety (the
interference model and the deficit model) explained academic performance on a statistics test. In
the interference model, test anxiety is an interfering agent. Highly test anxious students know
the content but unable to recall or retrieve it during the test. In the deficient model, anxiety is
merely an emotional reaction that accompanies the awareness of not being prepared for a test.
Highly test anxious students have difficulty acquiring content to begin with due to poor study
habits. The study found that the academic performance of highly test-anxious students is
affected directly by their lack of knowledge (deficit model) and indirectly by distractive thinking
during the test situation itself. Thus, the study found that a mixture of those two models best
describe the academic performance of highly test anxious students (Broader & Musch, 1999).
Eum and Rice (2010) explored the relationship between test anxiety (cognitive),
perfectionism, goal orientation, and academic performance, as assessed by GPA and word recall
task. Most test anxious students performed poorer on the recall task. This is supported by the
interference model (Broder and Musch, 1999). There was a negative relationship between test
anxiety and academic performance. Those findings are similar to those found by DiBattista and
Gosse (2006). Eum and Rice (2010) also found a positive relationship between cognitive test
anxiety and maladaptive perfectionism (high expectations and extreme self-blame for failing to
meet those expectations) and avoidance goal orientations. Furthermore, most test anxious
students were found to be female. These findings were supported by Barterian (2013). It is
important to note however, that the majority of participants were females (62%). The study was
also limited in that it only assessed one dimension of test anxiety (cognitive).
Similar to DiBattista and Gosse (2006), Broader and Musch (1999) found a statistically
significant negative relationship between test anxiety and exam performance. However, Broader
TEST ANXIETY 8
and Musch (1999) found that test anxiety explained only about 5% of the variance. Furthermore,
math skills not test anxiety was the most important factor, explaining 9% of the variance, in
exam performance. The authors concluded that in a statistics performance exam, academic skills
(math skills) are more important than task-irrelevant thoughts during a test. Ability, not negative
thoughts during the test, was the most important factor in predicting math test performance. The
results of Broader and Musch (1999) are actually more significant than they appear to be. As the
authors have pointed out, most test anxiety take place on a statistics exam. Even then, test
anxiety only explained 5% of the variance of exam performance. These results imply that test
anxiety would be even lower for other disciplines (Broader & Musch, 1999).
It is difficult to generalize the results of Broader and Musch (1999) to other disciplines.
The authors picked the discipline where college students have reported the most test anxiety on.
Second, the three variables examined in the study (test anxiety, math skills, and study habits)
explained only 25% of the variance in exam performance. The authors could have explained
more of the variance had they used more variables as previous studies have done. Third, as
previous research has shown, using evaluative measures such as the Test Anxiety instrument and
the Study Habits instruments put more pressure on highly-test anxious student. They induce
more anxiety and so may not be true reflections of the students. Furthermore, relaying on self-
reports such as the Study Habits instruments put the validity of this instrument into question
(Broader & Musch, 1999).
Test Anxiety at the K-12 Level
Barterian et al. (2013) was the first study to examine the differences in elementary
students reported test anxiety between a high-stakes test (NCLB achievement testing) and
classroom testing. Three hundred thirty-five students from Grades 3 to 5 participated in the
TEST ANXIETY 9
study. The study found that students reported a higher level of test anxiety on the high-stakes test
than on the classroom tests. More cognitive and psychological symptoms of test anxiety were
reported about the NCLB assessment (Barterian et al., 2013).
Contrary to these findings, Bodas et al. (2008) found no statistically significant
relationship between high stakes testing and test anxiety in Indian elementary school children.
The difference in findings may be due to cultural differences (western versus eastern). Bodas et
al. (2008) reported that parents and available resources play a big role in reducing any test
anxiety associated with high stakes testing. Higher expectations of Indian children could have
led to desensitization to test anxiety (Bodas et al., 2008). Bodas et al. (2008) was also restricted
in that the instrument used after translation was weakly reliable. The reliability coefficients of
the instruments were in the range of moderate to lower than those reported for the original
instruments (Bodas et al., 2008).
Barterian et al. (2013) also found that students who experienced higher levels of test
anxiety also experienced associated impairments in test performance. Specifically, the cognitive
symptoms of test anxiety more than the psychological symptoms were associated with the
impairments in test performance. The findings of this study also validated teachers perceptions
of their students test anxiety levels on the NCLB exam. Although 25% of the students reported
high levels of test anxiety on the NCLB testing versus classroom testing, 15% of the students
reported less test anxiety about the NCLB assessment (Barterian et al., 2013).
While Barterian et al. (2013) focused on elementary school children, Chamberlian, Daly,
and Spalding (2011) focused on high school students, ages 16 to 19. The two studies also
differed in their instruments. While Chamerlian et al. (2011) did not use a high stakes test
(NCTM) as Barterian et al. (2013) did, the study did use a highly stressful oral exam, a mock
TEST ANXIETY 10
French language speaking test. Chamberlian et al. (2011) examined the impact of test anxiety on
academic performance, a mock exam, and on heart rate. The study found a negative relationship
between test anxiety and academic performance, as measured by UK General Certificate of
Secondary Education (GCSE) scores. However, no statistically significant relationship was
found between test anxiety and performance on the mock exam. This is inconsistent with the
findings from Barterian et al. (2013).
Chamberlian et al. (2011) was limited in that the exam used in the study was a mock
exam rather than an actual exam. Had the actual exam been used, the impact of test anxiety on
test performance could have been statistically significant. Furthermore, the sample size was
relatively small compared to Barterian et al. (2013). However, the instruments used to measure
test anxiety in both studies were restricted in that they only measured two dimensions (worry and
emotionality) of test anxiety (Chamberlian et al., 2011).
Putwain (2008) examined the impact of test anxiety on test performance among high
school students. The study found a negative relationship between test anxiety and test
performance, as measured by GCSE examination scores. Higher levels of test anxiety were
associated with lower GCSE examination performance. As in Chamberlian et al. (2011), the
instrument used in Putwain (2008) to assess test anxiety only measured two dimensions:
cognitive and emotionality. The cognitive component of test anxiety accounted for 7% of
variance in examination performance. These findings are similar to those found in Chamberlian
et al. (2011) in terms of the relationship between test anxiety and GCSE scores.
Putwain (2008) also found that the impact of test anxiety on performance varied with
socioeconomic background. Students from lower socio-economic background had lower GCSE
scores partially due to test anxiety. As discussed in Putwain (2008), these findings cannot be
TEST ANXIETY 11
generalized to other nationalities. The study was also limited in that prior ability was not
controlled for (Putwain, 2008).
In comparison to other previously mentioned studies (Barterian et al., 2013; Bodas et al.,
2008; Chamerlian et al., 2011; Putwain, 2008), the participants in Lee, Lowe, and Sena (2007)
were both elementary and secondary school students. Lee et al., (2007) had the largest number
of participants (774 students) from among the previously mentioned studies. The study
examined test anxiety-assessment performance relationship among students with and without
learning disabilities (LD). The study differentiated between the four different dimensions of test
anxiety, an aspect that Chamberlian et al. (2011) and Putwain (2008) both lacked.
Lee et al. (2007) found that students with LD had higher scores on the worry and
cognitive obstructions/inattention parts of the test anxiety assessment. Students with LD scored
lower on performance enhancement/facilitation. That is, students with LD are less likely to
perform well on exams. Besides LD status, the study found that age and gender were significant
predictors of test anxiety. This is contrary to the findings of Putwain (2008) which found no
significant relationship between gender and test anxiety. Overall, older age predicted higher
cognitive obstruction/inattention scores and lower physiological hyperarousal scores. Female
students with LD predicted lower lie scale scores. That is, they have a more accurate prediction
of their scores (Lee et al., 2007). However, as in Putwain (2008), Lee et al. (2007) stated that the
findings of this study cannot be generalized to other nationalities.
Criticism of Research on Test Anxiety
Research on test anxiety has been criticized. Measurement of test anxiety has not been an
easy task. Studies found that the physiological measures of test anxiety are unrelated and there is
no preferred measure of anxiety. Self-report inventories of anxiety are unreliable due to the
TEST ANXIETY 12
assumption of the awareness of test anxiety as a conscious process (Burns, 2004). Similarly,
Chamberlian et al. (2011) and Eum and Rice (2010) have mentioned that self-report inventories
often use mood induction procedures that have induced an anxious mood in participants. Also,
self-report measures of test anxiety often use negative wording for the majority of items. Ideally,
measures should use both negative and positive wording to assess both negative and positive
thoughts. Test anxiety scales are often structured and specific eliciting particular responses.
Furthermore, administering test anxiety measures before an actual exam can affect students test
performance (Chamberlian et al., 2011).
There have been questions on whether general trait anxiety measures rather than test
anxiety measures should be used. It has been found that test anxiety is related to but different
from trait anxiety (Burns, 2004; DiBattista & Gosse, 2006). DiBattista and Gosse (2006) found
an inverse relationship between test anxiety and performance on exams but no relationship
between trait anxiety and test performance. The linear relationship often shown between anxiety
and performance has been criticized (Burns, 2004; Chamberlian et al., 2011; Mcdonald, 2001).
The relationship includes some interactions such as time pressures, difficulty of the tasks, exam
format, and low performance on previous exams. All these interactions and more should be
taken into consideration when measuring test anxiety (Burns, 2004; Putwain, 2008).
Addressing Test Anxiety
Armstrong, Boxer, Ford, and Ford (2012) found that college students who were exposed
to humorous material (cartoons) prior to taking a difficult math test exhibited lower levels of test
anxiety associated with the anticipated test and their performance was enhanced. The authors
concluded that humorous materials are effective in inhabiting the amount of test anxiety
associated with a test, and thus enhance performance. This finding was supported by other
TEST ANXIETY 13
research which found that humor lessens test anxiety if it is presented prior to the anxiety-
evoking event such as test while the event is being anticipated (Armstrong et al., 2012).
There were several limitations of Armstrong et al. (2012). First, the study assumed that
there is one type of test anxious students, those that know the material but unable to retrieve it
during the test. The authors assumed that test anxiety acted as an interfering agent rather than a
mere emotional reaction that accompanies the awareness of not being prepared for a test. The
study simply adopted the interference model with no reference to the deficit model (Broader &
Musch, 1999). Second, the study did not take into consideration other psychological
mechanisms besides test anxiety in explaining the enhancement of math performance. Perhaps
an increase in judgment rather than a decrease in test anxiety contributed to the enhancement of
math performance. Third, the study took place in a laboratory rather than in a realistic
environment (Armstrong et al., 2012).
Unlike Armstrong et al.(2012), Mowbrary (2012) focused on interventions that were
effective for the two types of test anxious students. One cognitive intervention given was the
attentional Cognitive Bias Modification (CBM-A), an attention training procedure. CBM-A is
designed to reduce negative attentional bias by training the participants to manage or remove
negative feelings or thoughts associated with an anxiety-related stimuli and direct their attention
toward an anxiety-free stimuli. With enough training, this intervention allows attentional
resources to be applied to relevant on task stimuli. The second cognitive intervention offered is
the acceptance and commitment therapy (ACT). The purpose of ACT is to help an individual
accept negative feeling and thoughts, to allow those feelings and thoughts to pass by rather than
be occupied with them (Mowbrary, 2012).
TEST ANXIETY 14
The above cognitive therapies do not have lasting effects. The literature shows thoughts
and behaviors return to their pre-restructured form, particularly in times of stress and test-related
situations are perceived as very stressful events (Mowbrary, 2012, p. 150). As was discussed in
the article, cognitive therapies may actually use more working memory resources and further
reduce efficiency. Also, CBM-A has been shown to negatively impact exam performance by
accidently removing important exam information. Studies using ACT have been limited by
small sample size making results tentive (Mowbrary, 2012).
Mowbrary (2012) focused on one dimension of test anxiety, cognitive (worry). However,
Stober (2004) focused on all four dimensions of test anxiety. Mowbrary (2012) was a review of
the literature while Stober (2004) was a study examining the relationship between different
dimensions of test anxiety and coping strategies. The study found significant gender differences
in test anxiety and in ways of coping with test anxiety. Overall, female students showed higher
levels of test anxiety (higher total scores on the TAI-G assessment). This finding is similar to
those found in Barterian et al. (2013). Also, overall, test anxiety was related only to coping by
social support (Stober, 2004).
Stober (2004) found that females scored higher on two dimensions of test anxiety (as
measured by TAI-G assessment): worry and emotionality. Both males and females used social
support to cope with those two dimensions. Females only coped with pre-exam anxiety related
to those two dimensions using task-orientation and preparation. With interference, males
significantly related to avoidance coping while females related to social support. Lack of
confidence was found to be statistically significant only for females and avoidance coping was
found to be related to it. Overall, females used task-orientation and preparation and social
support and less avoidance more than males (Stober, 2004).
TEST ANXIETY 15
Stober (2004) had few limitations. Although this study offered more coping strategies
than both Armstrong et al.(2012) and Mowbrary (2012), the study pertained to only three points.
As stober (2004) discussed, the study focused on pre-examination phase and so does not
accurately replicate test anxiety and coping strategies during the actual exam. Similar to
Armstrong et al.(2012) and Mowbrary (2012), the study focuses on university students. It is thus
difficult to generalize the results considering that test anxiety is found at a very early age (Stober,
2004).
Summary/Conclusion
Research on the impact of test anxiety on test performance yielded mixed results. Studies
(Barterian et al., 2013; Broader and Musch, 1999; DiBattista & Gosse; 2006; Embse & Hasson,
2012; Eum & Rice, 2010; Lee et al., 2007; Mcdonald, 2001; Putwain, 2008) have found that
there is a negative relationship between test anxiety and exam performance. Other studies
(Bodas et al., 2008; Burns, 2004; Chamberlian et al., 2011) found no statistically significant
affect of test anxiety on exam performance. One thing for certain is that the relationship between
performance and test anxiety is nonlinear (Chamberlian et al., 2011; McDonald, 2001).
Instruments often used in the studies did not assess all four dimensions of test anxiety. Culture
played an important role in examining test anxiety and so findings often could not be generalized
to other nationalities. Perhaps future research can examine the impact of test anxiety on test
performance while using gender, age, nationality, learning disabilities, and socioeconomic
background as moderators. It is also important for studies to control for prior ability to yield
accurate results.

TEST ANXIETY 16
References
Armstrong, J. M., Boxer, C. B., Ford, B. F., & Ford, T. F. (2012). Effect of humor on state
anxiety and math performance. International Journal of Humor Research, 25(1), 59-74.
doi: 10.1515/humor-2012-0004
Bodas, J. B., Ollendick, T. O., & Sovani, A. S. (2008). Test anxiety in indian children: A cross-
cultural perspective. Anxiety, Stress, and Coping, 21(4), 387-404. doi:
10.1080/10615800701849902
Broder, A. B., & Musch, J. M. (1999). Test anxiety versus academic skills: A comparison of two
alternative models for predicting test performance in a statistics exam. British Journal of
Educational Psychology,69, 105-116. doi: 10.1348/000709999157608
Burns, D. B. (2004). Anxiety at the time of the final exam: Relationships with expectations and
performance.Journal of Education for Business, 80(2), 119-124. doi:
http://tlc.apa.uoit.ca/wp-content/uploads/2011/12/Anxiety-at-the-time-of-the-exam.pdf
Carlson, J. C., Barterian, J. B., Embse, N. E., Goforth, A. G., & Segool, N. S. (2013). Heightened
test anxiety among young children: Elementary school students anxious responses to
high-stakes testing.Psychology in the Schools,, 50(5), 489-499. doi: 10.1002/pits
Chamberlain, S. C., Daly, A. D., & Spalding, V. S. (2011). Test anxiety, heart rate and
performance in a-level french speaking mock exams: an exploratory study.Educational
Research, 53(3), 321330. doi: 10.1080/00131881.2011.598660
DiBattista, D. D., & Gosse, L. G. (2006). Test anxiety and the immediate feedback assessment
technique. The Journal of Experimental Education, 74(4), 311-327. Retrieved from
http://www.epsteineducation.com/home/articles/file/research/testanxietyandtheif-at.pdf
TEST ANXIETY 17
Embse, N. E., & Hasson, R. H. (2012). Test anxiety and high-stakes test performance between
school settings: Implications for educators. Preventing School Failure, 56(3), 180187.
doi: 10.1080/1045988X.2011.633285
Eum, K. E., & Rice, K. R. (2010). Test anxiety, perfectionism, goal orientation, and academic
performance. Anxiety, Stress, & Coping, 24(2), 167-178. doi:
10.1080/10615806.2010.488723
Lee, S. W., Lowe, P. A., & Whitaker Sena, J. D. (2007). Significant predictors of test anxiety
among students with and without learning disabilities. Journal of Learning Disabilities,
40, 360-376. doi:10.1177/00222194070400040601
Mcdonald, A. M. (2001). The prevalence and effects of test anxiety in school
children. Educational Psychology: An International Journal of Experimental Educational
Psychology, 21(1), 89-101. doi: 10.1080/01443410020019867
Mowbray, T. M. (2012). Working memory, test anxiety and effective interventions: A
review. The Australian Educational and Developmental Psychologist, 29(2), 141156.
doi: 10.1017/edp.2012.16
Pekrun, R. P., & Stober, J. S. (2004). Advances in test anxiety research. Anxiety, Stress, and
Coping,,17(3), 205-211. doi: 10.1080/1061580412331303225
Putwain, D. P. (2008). Test anxiety and gcse performance: the effect of gender and socio-
economic background. Educational Psychology in Practice,24(4), 319334. doi:
10.1080/02667360802488765
Stober, J. S. (2004). Dimensions of test anxiety: Relations to ways of coping with pre-exam
anxiety and uncertainty. Anxiety, Stress, and Coping, 17(3), 213-226. doi:
10.1080/10615800412331292615