THE RELATIONSHIP BETWEEN LOCUS OF CONTROL, TEST ANXIETY AND SCHOOL PERFORMANCE

Abstract

Test anxiety is a specific situation frequently encountered in high school and university student population. Test anxiety is known as negatively affecting school/academic performance. Purpose: In the current study we have attempted to investigate the relationships between test anxiety and locus of control. Hypothesis: One hypothesis was tested, namely: students having a 'control' style based on externalism tend to show a higher test anxiety level. Participants: Data were obtained by response analysis of 198 students (139 girls and 59 boys), attending courses at the Faculty of Educational Sciences in Bucharest. Instruments: Participants filled up the Test Anxiety Inventory-Spielberger and the Internalism-Externalism Scale for Children and Youngers Chelcea, as well as a questionnaire referring to age, gender and academic level. Results: There were significant differences among gender in test anxiety scores. Female students exhibit higher statistic averages at all test anxiety measures, but only differences in Emotionality and total scores were significant. The results also indicated a statistically significant correlation between test anxiety and externalism scores.

Headnote

Abstract: Test anxiety is a specific situation frequently encountered in high school and university student population. Test anxiety is known as negatively affecting school/academic performance. Purpose: In the current study we have attempted to investigate the relationships between test anxiety and locus of control. Hypothesis: One hypothesis was tested, namely: students having a 'control' style based on externalism tend to show a higher test anxiety level. Participants: Data were obtained by response analysis of 198 students (139 girls and 59 boys), attending courses at the Faculty of Educational Sciences in Bucharest. Instruments: Participants filled up the Test Anxiety Inventory-Spielberger and the Internalism-Externalism Scale for Children and Youngers Chelcea, as well as a questionnaire referring to age, gender and academic level. Results: There were significant differences among gender in test anxiety scores. Female students exhibit higher statistic averages at all test anxiety measures, but only differences in Emotionality and total scores were significant. The results also indicated a statistically significant correlation between test anxiety and externalism scores.

Keywords: test anxiety, emotionality, worry, externalism, achievement motivation, gender.

Test anxiety - conceptual boundaries

Test anxiety is a variable specific to the school and academic environment, which many pupils and students experience. Thus, according to Moore1, approximately 30% of all American students suffered from a certain level of test anxiety. Among the symptoms characteristic to test anxiety we can find: associating the test result with personal value, fear of being rejected by parents or friends, time pressure, fear of losing control. Physical symptoms include rapid heartbeat, dry mouth, profuse sweating, stomach pain, dizziness, frequent urination. In this sense, some researchers have considered test anxiety as a chronic condition, which forces them to face negative results that do not correctly reflect their level of knowledge. Test anxiety negatively interferes with attention span and memory performance, making it difficult to remember topics studied for the test/exam.

A considerable number of research works have been carried out in order to study the consequences that test anxiety has at individual level, including performance on cognitive ability tests or school tests (achievement tests). The findings of these studies converge to the conclusion that test anxiety causes a decrease in performance in tests designed to evaluate cognitive and intellectual skills2. Also, research results have highlighted a positive correlation between the high level of test anxiety and school performance in the sense that a high level of test anxiety is negatively reflected in school results.3

It has been shown that test anxiety negatively interferes with performance in various tests that measure cognitive and intellectual functioning both in experimental laboratory situations4 and in real life ones - as is the case with tests and exams taken by pupils or students. Zeidner5 identified a direct relationship between the self-reported level of test anxiety and the information processing difficulties that the subjects experience when faced with various tasks. One approach explaining the low negative correlations between test anxiety and performance emphasizes the importance of some situational aspects. Bowler6 claims that test anxiety plays an indirect role, dependent on available time and test preparation. If a written exam is announced unexpectedly and therefore preparation for it is not possible, the variation in performance can be explained by individual differences as to the students' essential study and learning skills. On the other hand, if it is possible to prepare for the test in time, anxious students will focus on developing coping mechanisms to overcome the possible threat that the assessment situation represents for them. Students with a high level of test anxiety have difficulties in cognitive processes, other than those related to failure to recall information, as classical research claimed. Thus, students with a high level of test anxiety show problems in the processes of encoding and retaining (keeping) information, which lead, in most cases, to an inadequate conceptual representation of the contents they have to learn.7

In order to explain the relationship between test anxiety (fear of assessment) and academic performance, Salame proposes an interactional model that takes into account two variables: achievement motivation and failure orientation.8 Achievement motivation is defined by the conjunction between the level of aspiration and the importance that the pupil or student attaches to the examination, while the orientation towards failure is described by the negative perception that the pupil or student has regarding the level of their own competences, the expectations regarding the failure and the aversion towards the exam. Combined achievement motivation and failure orientation determine, on the one hand, the emotional reactions that the pupil or student manifests when they are about to face an exam (for example, increasing the level of anxiety as the exam approaches) and, on the other hand, the direction of the behaviors intended for adjustment and, at the end of the exam, the performance that the pupil or student will achieve.

Gender is considered one of the variables that have an important impact on the development and manifestation of test anxiety. Numerous studies have demonstrated/indicated that female subjects tend to score higher than male subjects on inventories or scales measuring test anxiety.9 The metaanalysis undertaken by Hembree10 highlighted the fact that differences depending on the gender variable, regarding the level of test anxiety, begin to appear in the population of students corresponding to the middle years of elementary school, becoming more noticeable for the populations of high school students, respectively college students. The same author showed that the size of the difference between the scores obtained by girls and those obtained by boys in the Emotionality component is considerably larger than that of the difference between the scores in the Worry component, this pattern indicating that the differences according to gender concern especially the affective dimension of test anxiety.

Hill and Wigfield11 identified, both for elementary school and high school students, as well as for college students, significant differences between girls' and boys' test anxiety scores.12 For each school level, girls obtained higher scores than boys. In an attempt to explain the observed differences between the scores obtained by female subjects and those obtained by male subjects regarding test anxiety, Zeidner showed that female subjects tend to be much more sensitive to evaluative stimuli, especially to those announcing a negative evaluation. Men tend to a greater extent than women to perceive a test situation as a personal challenge rather than a threat, which is why the anxious reactions that occur during a test situation are interpreted by them in a positive way.

Another explanation for the differences observed between female and male subjects in scores on scales or inventories measuring test anxiety was related to the different growth and socialization patterns of girls and boys. Girls are educated to recognize and express their anxiety - socially perceived as a specifically feminine trait. In this sense, the reactions from society towards the anxiety expressed by women tend to be of a more supportive and comforting nature, this prompting them to recognize rather than hide the anxiety towards the test that they feel. Men are educated to see evaluative situations as challenges that they must face, while women tend to see evaluative situations as threats, which they can cope with by resorting to strategies based on focusing on their own emotions or on avoidance behaviors.

Therefore, the data found regarding the level of test anxiety do not reflect a real difference between women and men, but are rather the expression of differences related to self-presentation and the availability of female and male subjects to admit openly that they suffer from anxiety13.

Procedure

Purpose and hypothesis

By means of the study we carried out, we set out to identify the relationships between test anxiety, anxiety and locus of control among students at the Faculty of Educational Sciences. The relationships were approached from the perspective of possible associations and the contribution that the last two variables have in explaining individual differences in terms of dysfunctional negative cognitions (the Worry factor), respectively the emotional responses specific to test anxiety (the Emotionality factor). We sought to identify the meaning and magnitude of the correlations between the measured variables.

Taking into consideration the results of the studies presented above, we started from the assumption that students for whom a predominantly externalist attribution style prevails tend to register higher levels of test anxiety.

Participants

The research batch consisted of a number of 198 people, students at the Faculty of Educational Sciences, Bucharest. The ages of the students varied between 19 and 32 years (m = 26 years; s = 1.18 years). The distribution of students according to the gender variable was: 59 boys and 139 girls. The 198 students completed two instruments, namely: the Test Anxiety Inventory (TAI), the Locus of Control Scale.

Instruments

l.The Test Anxiety Inventory (TAI) was built up and developed with the aim of measuring individual differences concerning the specific reactions to test anxiety, seen as a personality trait related to an assessment situation14 . The inventory contains 20 items with four answer options (A - almost never, B - sometimes, C - often and D - almost always), from which the subject must choose the one that suits him. The respondents are asked to indicate how often they manifest certain specific symptoms of test anxiety before, during and after an evaluative situation. In addition to measuring individual differences related to the predisposition to anxiety in various test situations, the scales of the inventory allow the assessment of the two components of test anxiety described by Liebert and Morris15 : negative cognitions related to performance and failure (the Worry factor), respectively the emotional reactions resulting from the intensification of the activity of the autonomic nervous system (the Emotionality factor). All 20 items are included in the calculation of the total score, considered the indicator of the level of test anxiety present in a subject. Since the answer to an item can be marked from 1 to 4, the minimum total score that a subject can obtain is 20, and the maximum is 80. The high total score obtained by a subject indicates a heightened level of anxiety against testing (as a latent trait, manifested in evaluative situations). On each of the scales, by which the two components of test anxiety are evaluated, a subject can obtain a score between 8 and 32. In our research, we calculated the internal consistency (a-Cronbach coefficient) for each of the scales of the inventory, as well as for the set of items. The obtained data are the following: Emotionality scale - 0.86, Worry scale 0.79 (the whole batch); the set of inventory items -0.91.

2. The locus of control, characteristic of an individual's cognitive style that can be distributed on a continuum between internalism and externalism, was measured with the Internalism-Externalism Scale for Children and Young People (IE-CT), adapted in Romania by S. Chelcea16 according to the Nowicki-Strickland Locus of Control Scale/N-SLCS.17 In the version we used in our study, the scale included 40 items, to which students answered 'yes' or no'¦ Each answer was rated with zero points or one point, depending on the correction grid. The total scores of the students were obtained by summing the scores of the 40 items. Following the indications provided by Chelcea, we considered a high total score as expressing externalism. In the current study, we obtained the following internal consistency values: 0.66.

Results Descriptive statistics

In the table below we have shown the averages, standard deviations and indicators of the shape of the distributions for the variables we measured.

From the standard deviation values, we find out that girls obtained, compared to boys, more heterogeneous distributions of scores on the TAI inventory. Instead, for the 'locus of control' variable, the boys obtained a more heterogeneous distribution of scores than the one obtained by girls. In the case of the scores for study skills and habits, the distribution obtained by girls was slightly more heterogeneous than the one obtained by boys.

Comparisons according to gender and age variables From the comparisons we made, we found that girls obtained significantly higher scores than boys, on the Emotionality scale of the TAI inventory (t = 2.88, p < 0.01, d = 0.45), as well as on the entire inventory (t

= 3.13, p < 0.01, d = 0.46). On the Worry scale, girls also obtained a higher average, but the difference was not statistically significant.

For the locus of control variable, the difference between girls and boys was statistically insignificant (t = 1.53, p = 0.13, d = 0.24).

As far as the age variable, the data were obtained after performing an analysis of variance (One-Way ANOVA), taking age as the independent variable and the scores on the two scales of the TAI inventory as dependent variables. The results indicated the tendency of scores to increase on both scales with increasing age, students over 30 years old having obtained higher scores for both scores of the TAI scale.

Correlations between the measured variables

From Table 2, we can see significant correlations between test anxiety and the other variables we measured. The correlation values were between - 0.51 and 0.95, predominating correlations with moderate absolute values.

Positive and high correlation values (over 0.80) were recorded between the scores on each of the scales of the TAI inventory and the total scores. This result is natural, if we take into account the fact that the scores on the items of the scales are included in the calculation of the total score on the inventory.

The positive and moderate correlation between the scores on the scales of the TAI inventory is due to the fact that both items that measured cognitive concerns (the Worry factor) and those that measured emotional reactions (the Emotionality factor) were built up so that they could target facets (dimensions) of a factor related to test anxiety.

On the other hand, the scores on the TAI inventory correlated positively, moderate in absolute value and statistically significant with the scores on the scale for measuring locus of control (externalism). The correlation value between the Worry scale scores and the locus of control scores was higher than the correlation value between the Emotionality scale scores and the locus of control variable.

Discussions

The data we obtained in our study are consistent with the data reported in the studies that dealt with the relationships between test anxiety and other variables related to the cognitive and emotional functioning of the individual.

Thus, the girls obtained, compared to boys, significantly higher scores on the Emotionality scale, as well as on the entire TAI inventory. These data confirmed those reported by American authors.18 It is shown that, in general, women are considered to be more sensitive in relation to evaluative situations and, consequently, show the tendency to be more anxious about a possible negative evaluation/failure.

The studies carried out by a number of authors 19 offer as a possible explanation for the gender difference in terms of test anxiety scores, the differences that exist between women and men from the point of view of the patterns of socialization. Women feel a greater tension in relation to school/academic (in)success, as they fear failure. Women tend to see every testing situation as another possible chance of failure, which is associated with a higher level of test anxiety among them.

On the other hand, the data we obtained revealed the tendency of the scores on the Worry scale from the TAI inventory, as well as the total scores on test anxiety, to increase as the age of the students in the research group increased. This result is consistent with the data reported by other authors who showed that the level of test anxiety increases with age (this increase is observed until the period corresponding to the college years), after which it begins to decrease according to Wheeler.20

The increasing trend of the level of test anxiety depending on the age variable that we found in our study could be based on several factors: 1) increasing pressure from the family and professional environment (most students after 30 are married, have children and work as substitute teachers in education) in terms of their academic results; 2) the accentuation of the defensive attitude among younger students, who were probably less willing to recognize the specific symptoms of test anxiety; 3) the increase in the complexity of school materials and tasks, an aspect that can contribute to the overload of older students, taking into account the complexity of their professional and personal lives.

This result is in accordance with a number of studies21, which stated that test anxiety increases with schooling level. This would be due to the stressful situations represented by the exams, the school tasks perceived to be more and more demanding and the increase in responsibilities from year to year. According to the cited author and his collaborators, students in their last year show a tendency to have high expectations regarding the results of the graduation exam, which depends on the performances during the years of study and which is considered very important for their future career. Al-Doughmi et al.22 cast doubt on the observation that the frequent exposure of pupils and students to evaluative situations contributes to reducing the level of test anxiety among them.

In the study we carried out, we recorded a positive, significant and moderate correlation in absolute value between the scores on the scales of the TAI inventory. This result confirmed the data of the studies cited by Zeidner23, according to which cognitive concerns and emotional reactions related to test anxiety correlate positively with each other. The results that we obtained in our study, contradictory to the expectations we had, can be explained by the fact that the subjects in the experimental batch are very concerned about their self-image and especially how they are perceived by others. The fact that they are teaching staff (working as substitute teachers) induces them a feeling of "fear" in front of the exam in the sense that a possible failure would affect their image and authority in front of their own students.

The data we obtained indicated positive, significant and moderate correlations in value between the scores on the TAI inventory and the scores on the locus of control variable (externalism), thus confirming the results reported in other studies.24 Thus, Moore appeals to Barlow's theory of apprehension to explain these results. According to Barlow's theory, people who believe they have limited control over the events they encounter in life are more prone to experiencing anxiety, compared to people who believe they can control the events they face throughout life. In the present case, we can say that when a student facing a test situation believes that he cannot control his own skills and the chance of success, he tends to show anxious symptoms. In this way, they end up being more prone to developing anticipatory failure anxiety and thus test anxiety.

Conclusions

The findings in our study are consistent with the results of other studies, which dealt with test anxiety and its relationships with other variables related to the individual's emotional functioning. The established hypotheses were partially confirmed but in the direction of those assumed in the sense that an externalist attribution style positively correlates with a high level of test anxiety, and a high level of self-confidence positively correlates with a high level of test anxiety contrary to our expectations. The value of the results must be analyzed, first of all, from the perspective of the relevance they have for the explanatory model of the causes and factors favoring test anxiety. Also, the value of our results must be interpreted in relation to the practice of diagnosing the symptoms of test anxiety in the school population and the intervention programs aimed at preventing this scourge, which negatively affects, along with other factors, the academic performance of students.

From a theoretical perspective, the results we obtained suggest that, in evaluating test anxiety among students, we must take into account their individual characteristics. Among these characteristics, locus of control (the style of attributing one's own successes/failures) can play an important role.

Along with these variables, in the evaluation of the complex etiology of test anxiety, the school psychologist must also take into account: certain stable dispositions of the personality (traits, such as emotional stability or conscientiousness, structures related to the motivation to achieve success in school, one's own image and self-esteem, etc.), certain factors related to school or family environment. These factors interact dynamically, determining the manifestation of certain behavioral patterns of students, related to academic performance, among which specific symptoms of text anxiety can be encountered.

Test anxiety is an undesirable scourge that affects the school population to a worrying extent. Therefore, concerted action on the part of all educational agents (teachers, school psychologists, school principals, parents) is strictly required in order to provide psycho-pedagogical assistance to students who are facing the troublesome symptoms of test anxiety, as well as in order to reduce the incidence and manifestation of this scourge. From the point of view of programs designed to prevent and reduce test anxiety in the school population, the results of the study we undertook suggest that permanent work with students (individual and group psychological counseling, monitoring school performance) is a necessity, even from the first year of high school. The students' approach can be differentiated, depending on the skills they have (cognitive-intellectual, motivational, certain personality traits) and the fluctuations in their school performance.

Footnote

1 M.M. Moore, Variations in Test Anxiety and Locus of Control Orientation in Achieving and Underachieving Gifted and Nongifted Middle School Students. Unpublished doctoral dissertation, University of Connecticut. http//www.gifted.uconn.edu, 2006.

2 R. Hembree, Correlates, Causes, Effects, and Treatment of Test Anxiety in Review of Educational Research, 58 (1), 47-77,1988.

3 R. Hembree, idem

4 M. Zeidner, Test Anxiety: The State of the Art. New York: Plenum, 1998.

5 M. Zeidner, idem.

6 R. Bowler, A Brief Review of Test Anxiety in West German Schools. In R. Schwarzer, H. M. Van der Ploeg, & CD. Spielberger (Eds.). Advances in Test Anxiety Research (Vol. 5., pp. 85-90). Lisse, The Netherlands: Swets & Zeitlinger, 1987.

7 J.C. Cassady, The impact of cognitive test anxiety on text comprehension and recall in the absence of external evaluative pressure. Applied Cognitive Psychology, (18, 311-325), 2004.

8 R.F. Salame, Test anxiety: its determinants, manifestations, and consequences. In H.M. Van der Ploeg, R. Schwarzer, & CD. Spielberger (Eds.) Advances in Test Anxiety Research (Vol. 3, pp. 83-118). Lisse, The Netherlands: Swets & Zeitlinger, 1984.

9 CD. Spielberger, Manual for the State-Trait Anxiety Inventory (Form Y). Palo Alto, CA: Consulting Psychologists Press, Inc, 1983.

10 R. Hembree, Correlates, Causes, Effects, and Treatment of Test Anxiety, op. cit.

11 K.T. Hill & A. Wigfield, Test anxiety: a major educational problem and what can be done about it. The Elementary School journal, 85 (1), 105-126,1984.

12 T.S. Hall, Is test anxiety a form of specific social phobia? Unpublished Master's thesis. College Park, University of Maryland, 2005.

13 M. Zeidner, Test Anxiety: The State of the Art, op.cit.

14 CD. Spielberger, Manual for the State-Trait Anxiety Inventory, op. cit.

15 R. M. Liebert & L.W. Morris, Cognitive and emotional components of test anxiety: A distinction and some initial data. Psychological Reports, 20(3, PT. 1), 975-978,1967.

16 S. Chelcea, S., The Locus of Control and the Emergence of Rumours. In Personality and Society in Transition. Social Psychology Studies. Bucharest: Science and Technology Society, 1994.

17 B. Strickland, The Gale Encyclopedia of Psychology (Second edition). Farmington Hills: Gale Group, (Ed.) (2001).

18 A. Fiore, Gender differences in test anxiety. Unpublished master's thesis. College of Human Resources and Education, West Virginia University, 2003.

19 M. M. Moore, Variations in test anxiety and locus of control orientation in achieving and underachieving gifted and nongifted middle school students, 2006.

20 J.M. Wheeler, Anxiety levels of school age students prior to and following high-stakes testing. Unpublished dissertation submitted in partial fulfillment of the requirements for the degree of educational specialist in school psychology, 2005.

21 G. Sarason, Stress, anxiety, and cognitive interference: reactions to tests. Journal of Personality and Social Psychology, 46 (4), 929-938,1984.

22 Al-Doughmi, M.A., Shuriquie, N.T., Abdulhamid, M, Al-Ghwairy, A., Shuneigat, W. (2006). The impact of exams on anxiety levels among university students. Journal of the Royal Medical Services, 13 (1), 14-19.

23 Zeidner, Test Anxiety: The State of the Art, op. cit.

24 M.M. Moore, op. cit.

References

REFERENCES

Al-Doughmi, M.A., Shuriquie, N.T., Abdulhamid, M., Al-Ghwairy, A., & Shuneigat, W., (2006), The impact of exams on anxiety levels among university students. Journal of the Royal Medical Services, Iordania, 13 (1), 14-19.

Arch, E.C., (1987), Differential responses of females and males to evaluative stress: anxiety, self-esteem, efficacy, and willingness to participate. In R. Schwarzer, H.M. Van der Ploeg, & CD. Spielberger (Eds.). Advances in Test Anxiety Research (Vol. 5., pp. 97-106). Lisse, The Netherlands: Swets & Zeitlinger.

Cassady, J.C., (2004, The impact of cognitive test anxiety on text comprehension and recall in the absence of external evaluative pressure. Applied Cognitive Psychology, 18, 311-325.

Chelcea, S., (1994), The locus of control and the emergence of rumours. In Personality and Society in Transition. Social Psychology Studies. Bucharest: Science and Technology Society, 1994.

Corcoran, K., & Fischer, F., (Eds.) (1987), Measures for Clinical Practice. A Sourcebook. New York: The Free Press.

Ergene, T., (2003), Effective interventions on test anxiety reduction. A meta-analysis. School Psychology International, 24 (3), 313-328.

Fiore, A., (2003), Gender Differences in Test Anxiety. Unpublished master's thesis, College of Human Resources and Education, West Virginia University.

Hall, T.S., (2005), Is Test Anxiety a Form of Specific Social Phobia? Unpublished master's thesis. University of Maryland, College Park.

Herrmann, C, Liepmann, D., Otto, J., (1987), Problem-solving and action control as determinants of test anxiety. In R. Schwarzer, H.M. Van der Ploeg, & CD. Spielberger (Eds.). Advances in Test Anxiety Research (Vol. 5., pp. 87-96). Lisse, The Netherlands: Swets & Zeitlinger.

Moore, M.M., (2006), Variations in Test Anxiety and Locus of Control Orientation in Achieving and Underachieving Gifted and Nongifted Middle School Students. Unpublished doctoral dissertation, University of Connecticut, http/ /www.gifted.uconn.edu.

Pintrich, P.R., & De Groot, E.V., (1990), Motivational and self-regulated learning components of classroom academic performance. Journal of Educational Psychology, 82 (1), 33-40.

J J -J CJ -J \ /' Popa, M., (2008), Statistics for Psychology. SPSS Theory and Applications. IasT: Polirom Publishing House.

Robu, V., (2012), Predictors of test anxiety among high school students. Implications for the teaching and learning process. The Educational Journal, 60 (1), 35-53.

Robu, V., (2011), The Psychology of Test and Exam Anxiety. Theoretical Milestones and Applications, IasT: Performantica Publishing House.

Robu, V., (2008), Test Anxiety: gender differences. In M. Milcu, W. Griebel, R. Sassu (Coordonatori). Modern Psychological Research: directions and perspectives (pp. 116-131). Bucharest: University Publishing House.

Robu, V., Morara§u, C.V., (2008), Psychological correlates of test anxiety. In M. Milcu, W. Griebel, R. Sassu (Coordinators). Modern Psychological Research: directions and perspectives (pp. 132-142), Bucharest: University Publishing House.

Sarason, I.G., (1984), Stress, anxiety, and cognitive interference: reactions to tests. Journal of Personality and Social Psychology, 46 (4), 929-938.

Spielberger, CD., Gorsuch, R.L., & Lushene, R.E., (1970), Manual for the State-Trait Anxiety Inventory (Self-Evaluation Questionnaire). Palo Alto, CA: Consulting Psychologists Press, Inc.

Spielberger, Ch. D., (1980), Test Anxiety Inventory. Preliminary Professional Manual. Redwood City, CA: Mind Garden, Inc.

Strickland, B., (Ed.) (2001), The Gale Encyclopedia of Psychology (Second edition). Farmington Hills: Gale Group.

Wheeler, J.M., (2005), Anxiety Levels of School Age Students Prior to and Following High-Stakes Testing. Unpublished educational specialist dissertation, Graduate College of Marshall University. http/ / www.marshall.edu.

Zeidner, M., (1998), Test Anxiety: The State of the Art. New York: Kluwer Academic Publishers.