The Canadian Journal for the Scholarship of Teaching and Learning Volume 6|Issue 3 Article 3 12-11-2015 Perceptions and Incidence of Test Anxiety Travis G. Gerwing University of New Brunswick, Department of Biology, [email protected] Joshua A. Rash University of Calgary, Department of Psychology Alyssa M. Allen Gerwing University of New Brunswick, Department of Biology Bev Bramble University of New Brunswick, Center for Enhanced Teaching and Learning Jeff Landine University of New Brunswick, Faculty of Education Follow this and additional works at:http://ir.lib.uwo.ca/cjsotl_rcacea Part of theEducational Psychology Commons,Higher Education Commons,Other Psychology Commons, and theSchool Psychology Commons Recommended Citation Gerwing, Travis G.; Rash, Joshua A.; Allen Gerwing, Alyssa M.; Bramble, Bev; and Landine, Jeff (2015) "Perceptions and Incidence of Test Anxiety,"The Canadian Journal for the Scholarship of Teaching and Learning: Vol. 6: Iss. 3, Article 3. Available at:http://ir.lib.uwo.ca/cjsotl_rcacea/vol6/iss3/3 Perceptions and Incidence of Test Anxiety Abstract Test anxiety (TA) can lower student GPA and increase dropout rates in populations of university students. Despite numerous treatment options, many students still suffer from TA. The stigma attached to this type of anxiety and the incidence rates and perceptions of TA were quantified through surveys distributed to 1,099 students at a Canadian university. Results of this study indicated that 38.5% of students (30.0 % of males, 46.3 % of females) suffered from self-reported TA at some point over the course of their university career. The prevalence of TA varied by faculty, with the highest incidence among those students enrolled concurrently in Arts and Science, and Nursing students. While student perceptions varied by age, sex, and experience with TA, one third of students expressed negative and inaccurate views about TA. These negative perceptions may explain why 11.3% of surveyed students indicated they would not seek help for their TA as, for many, to do so would make them seem weak in the eyes of their colleagues. Further, 20.5% of students surveyed reported that they believe professors would be unable or unwilling to help. It may be the case that this negative perception towards TA makes it difficult for faculty and helping professionals to identify and intervene effectively. Faculty specific educational campaigns designed to educate students about TA, in particular about its prevalence and severity, are suggested as a method to circumvent the negative stigma surrounding this condition. Implementation of such educational policies will likely improve the educational experience and performance of students with TA, as well as improve student retention. L’anxiété due aux examens peut faire baisser la moyenne cumulative des étudiants et augmenter le nombre de décrocheurs parmi les étudiants universitaires. Malgré les nombreuses options qui existent pour remédier à cela, un grand nombre d’étudiants souffrent de cette anxiété. La stigmatisation de ce type d’anxiété, ainsi que le taux d’incidence et les perceptions de l’anxiété due aux examens ont été quantifiés grâce à des sondages distribués à 1.099 étudiants dans une université canadienne. Les résultats de cette étude indiquent que 38,5 % d’étudiants (30,0 % d’hommes et 46,3 % de femmes) ont déclaré avoir souffert d’anxiété due aux examens à un moment ou à un autre au cours de leurs études universitaires. La prévalence de cette anxiété variait selon les facultés et la plus haute incidence se trouvait parmi les étudiants inscrits simultanément en lettres et sciences et en soins infirmiers. Les perceptions des étudiants variaient selon l’âge, le sexe et l’expérience préalable d’avoir souffert d’anxiété, toutefois un tiers des étudiants ont exprimé des opinions négatives et inexactes à propos de l’anxiété due aux examens. Ces perceptions négatives pourraient peut-être expliquer pourquoi 11,3 % des étudiants qui ont participé au sondage ont indiqué qu’ils n’allaient pas chercher à se faire aider pour réduire leur anxiété car, s’ils le faisaient, cela les ferait paraître encore plus faibles aux yeux de leurs collègues. De plus, 20,5 % des étudiants interrogés ont rapporté qu’ils pensaient que leurs professeurs seraient incapables de les aider ou peu disposés à le faire. Il est possible que cette perception négative à l’égard de l’anxiété due aux examens explique pourquoi il est difficile pour les professeurs et les autres professionnels aidants d’identifier le problème et d’intervenir efficacement. Les campagnes éducatives qui s’adressent spécifiquement aux professeurs et qui sont conçues pour éduquer les étudiants sur l’anxiété due aux examens, en particulier sur la prévalence et la gravité de cette anxiété, ont été suggérées comme moyen de circonvenir les stigmates négatifs qui entourent cette condition. La mise en oeuvre de telles politiques éducatives pourrait probablement améliorer l’expérience éducative et la réussite des étudiants qui souffrent d’anxiété due aux examen, elle pourrait également améliorer la rétention des étudiants. Keywords test anxiety; exam anxiety; incidence; perceptions; stigma This research paper/rapport de recherche is available in The Canadian Journal for the Scholarship of Teaching and Learning: http://ir.lib.uwo.ca/cjsotl_rcacea/vol6/iss3/3 Cover Page Footnote We would like to thank the Center for Enhanced Teaching and Learning for funding this project, as well as the instructors who gave up class time for our survey. This research paper/rapport de recherche is available in The Canadian Journal for the Scholarship of Teaching and Learning: http://ir.lib.uwo.ca/cjsotl_rcacea/vol6/iss3/3 Gerwing et al.: Perceptions and Incidence of Test Anxiety In the typical academic environment, often heavily dependent on the practice of formal testing and entrance examinations, tests and examinations represent a potent stressor that can impact a student’s performance in school (Cassady & Johnson, 2002; Vitasari, Wahab, Othman, Herawan, & Sinnadurai, 2010). Given the potential negative consequences associated with poor test performance or failure, it is not surprising that students consistently cite exam and grade-related worries as their greatest sources of anxiety (Furr, Westefeld, McConnell, & Jenkins, 2001). The experience of high levels of exam-related anxiety is referred to in the literature as test anxiety (TA) and is characterized by cognitive, behavioural, and physiological responses induced by academic testing (Zeidner, 1998). Common behavioural expressions of TA can include poor study skills and procrastination. These behaviours may be experienced before, during, or after an examination (Gregor, 2005; Liebert & Morris, 1967; Spielberger & Gonzalez, 1980; von der Embse, Barterian, & Segool, 2013). The relationship between TA and other anxiety disorders is currently unclear. Some evidence suggests TA may be related to, and interact with, general anxiety disorders (Hopko, Hunt, & Armento, 2005; In'nami, 2006), while other studies suggest that TA may be a separate condition (Hopko et al., 2005; Wittmaier, 1972). Further complicating these relationships, TA can range from mild to severe, and, in some cases, anxiety, termed facilitative anxiety, can improve performance (Cassady & Johnson, 2002). Regardless, TA has been identified as a serious issue affecting the academic performance of college students (Cassady & Johnson, 2002; Chapell et al., 2005; Szafranski, Barrera, & Norton, 2012; Vitasari et al., 2010). Previous research suggests that 15-40% of university students experience functionally impairing levels of TA, depending on the survey methodology used and population studied (Abolghasemi, Golpour, Narimani, & Ghambari, 2009; Hill, 1984; Szafranski et al., 2012; Vitasari et al., 2010; von der Embse et al., 2013). In fact, TA may be a contributing factor to 15- 29% of student withdrawals from post-secondary institutions (Khalaila, 2014; Spielberger, 1962). More specifically, females report higher levels of TA than males (Chapell et al., 2005; Egbochuku & Obodo, 2005; Szafranski et al., 2012; Zaheri, Shahoei, & Zaheri, 2012), and there is evidence to suggest that the negative association commonly found between TA and poor academic performance is stronger in females (Chapell et al., 2005; Ferrando, Varea, & Lorenzo, 1999; Szafranski et al., 2012; Zaheri et al., 2012). However, the relationship between TA and academic performance may be more complicated, as females may exhibit higher levels of test anxiety than males, but do not always suffer from reduced performance (Zeidner, 1990). Personal and situational variables may also contribute to the incidence of TA. For example, in one study, socioeconomic status, age, sex, and ethnicity were reported to predict TA (Putwain, 2007). Moreover, features of the classroom environment, such as positive regard expressed by the professor towards students, have been shown to influence students’ motivations, attitudes, and competence (Eccles et al., 1993; Vallerand, Fortier, & Guay, 1997). These factors may interact with TA and influence student performance. Despite these studies into the factors that may contribute to TA, there is little research looking at the relationship between student perceptions and TA. The identification of this relationship is important for several reasons. First, perceptions of TA may differ depending on demographic and academic factors. Determining how perceptions of TA vary may offer insights into the thought processes underlying TA. Published by Scholarship@Western, 2015 1 The Canadian Journal for the Scholarship of Teaching and Learning, Vol. 6, Iss. 3 [2015], Art. 3 Second, identifying perceptions of TA has the potential to inform policy makers on how students view this condition which may aid in the design of effective educational and treatment policies. For instance, information on perceptions about TA may offer insight into the student populations for which education is most needed. For this study, two surveys were administered to students attending a Canadian university with the objective of getting a broad understanding of student perceptions of TA. The research goals of the study were to: (a) determine student perceptions of TA (how students view TA and those who suffer from this condition); (b) investigate whether or not a negative stigma exists around this condition; and (c) quantify the prevalence of TA on campus. More specifically, we wanted to quantify the perceptions students hold towards TA and those who suffer from it, student perceptions of the institutional commitment to addressing TA, and the perceived treatability of TA. We also wanted to determine if perceptions varied by age, sex (Putwain, 2007), or the amount of time in school. With regards to the negative stigma potentially surrounding TA, we wanted to determine what proportion of the student population held negative views about TA. We also attempted to quantify the proportion of students who would not seek treatment for TA as a result of this stigma. Method This study was conducted at a major Canadian university (~11,000 undergraduate students), located in a small Canadian city. Two questionnaires were distributed to students during class time and students were asked to fill in both surveys, though participation was voluntary. Classes were selected for participation based upon the willingness of the instructors to give up class time, and represented a wide variety of faculties (Table 1). Approximately 2% of instructors allowed us access to their class. All faculty and instructors on campus were contacted via email or in person, and were given the opportunity to participate. Data were collected from the classes of the instructors who responded. First year students were surveyed in their second semester, so that all students had experienced university level exams by the time they completed the questionnaires. The study protocols were approved by the institution’s Research Ethics Board. All participants provided written consent. http://ir.lib.uwo.ca/cjsotl_rcacea/vol6/iss3/3 2 Gerwing et al.: Perceptions and Incidence of Test Anxiety Table 1 Percentage of Students Reporting Severe1 Test Anxiety (TA) by Faculty Faculty % TA (N) Arts/Science2 53.6 (19) Nursing 52.3 (130) Computer Science 42.9 (14) Kinesiology 42.3 (104) Science 40.1 (133) Forestry 38.9 (36) Engineering 38.2 (131) Business 37.8 (114) Arts 36.5 (104) Unspecified 9.7 (31) Note.1 Students self-reported that their TA lowered their performance at some point in their academic career. 2Arts/Science represents the concurrent degree where students obtain bachelor’s degrees in both the arts and science faculties. Questionnaires The Test Anxiety Inventory (TAI) (Spielberger & Gonzalez, 1980; Szafranski et al., 2012) is a 20-item self-report psychometric measure designed to assess individual differences in TA. Participants respond on a 4-point Likert scale with anchors 1 “almost never” to 4 “almost always.” The TAI generates a total score (TAI-T) as well as worry (TAI-W) and emotionality (TAI-E) subscales. The worry subscale reflects cognitive concerns about the consequences of failure while TAI-E reflects reactions of the autonomic nervous system. The TAI, and its subscales, has proven both reliable and valid in assessing TA among samples consisting of high-school students, university students, and community members (Spielberger & Gonzalez, 1980; Szafranski et al., 2012). The TAI also demonstrated reliability in the present sample (α = .96). The Perceptions of Test Anxiety Survey (PTAS) was created by the authors for this study to measure perceptions of TA severity, the students’ view of the University’s commitment to TA, and the treatability of TA. Participants were asked to respond to 22 items using a 1 “strongly disagree” to 7 “strongly agree” Likert scale. The PTAS was reliable in the present sample, α = .72. Ratings on the PTAS were subjected to a principal components analysis to extract latent components underlying the survey. The PTAS also contained a written section asking about experiences with TA (e.g., Have you ever experienced TA at any point over the academic career which negatively influenced academic performance? Have you ever sought help for TA? What TA services are available on campus?). As part of the survey package, students completed a demographics questionnaire asking for information about age, sex, ethnicity, level of education, and faculty. Data Analysis Test Anxiety Inventory. We used student scores on the TAI-T, TAI-W, and TAI-E to characterize our sample. Beyond sex differences, further investigation into TAI data (faculty, Published by Scholarship@Western, 2015 3 The Canadian Journal for the Scholarship of Teaching and Learning, Vol. 6, Iss. 3 [2015], Art. 3 age, etc.) was not possible, as in order to maximize confidentiality, TAIs were not paired to perception surveys. Perceptions of TA. A principal components analysis was performed on the 22-item PTAS to derive the latent constructs underlying student perceptions. Several iterations of principal components analysis were performed until the best solution was uncovered. Participant scores on the components extracted were generated using regression in SPSS Factor, and these scores were used as dependent variables in subsequent analyses. Independent samples t-tests were used to assess if perceptions of TA varied by sex, experience with TA, or knowledge of institutional TA services. As not all participants responded to all survey items, degrees of freedom vary between tests. Independent hierarchical linear regressions were used to assess whether age and years of education were predictors of perceptions of TA after controlling for sex. Inflation of familywise error was corrected using the Holms-Bonferroni correction procedure. Principal components extraction. Principal components extraction was used to estimate number of factors, presence of outliers, absence of multicollinearity, and factorability of the correlation matrix. With an α = .001 cut-off level, there were no univariate outliers, however 50 cases produced scores that were identified as multivariate outliers exceeding χ2(22) = 48.27 and were removed from principal components extraction. The final principal components extraction was performed on 16-items from the PTAS. Results indicated that the sampling was adequate and inter-correlations were strong enough to merit principal components extraction; Kaiser-Meyer-Olkin measure of sampling adequacy = .86. Further, the null hypothesis that variables are uncorrelated was rejected using Bartlett’s test of sphericity χ2(120) = 4725.43, p < .001. Three principal components were extracted. Communality values (h2), as seen in Table 2, were moderate in magnitude. The lowest communality extracted was .27 and the highest was .72, indicating that between 27% and 72% of item variability was predictable by the underlying components. Component one (Importance of TA) was composed of ten items, component two (Institutional Commitment to TA) of four items, and component three (Treatability of TA) of two items. There is some concern that component three may not be reliable, as it only consisted of two items (Tabachnick & Fidell, 2012). As such, component three will be interpreted with caution. When orthogonal rotation was required, components interpreted as “importance of TA” and “institution commitment to TA” were correlated .34. Further, components interpreted as “institution commitment to TA” and “treatability of TA” were correlated .22. Oblique rotation was chosen because correlations were modest and persisted across components. Loadings of variables on components, communalities, and percent of variance accounted for are shown in Table 2. Variables are ordered by item number and component loading to facilitate ease of interpretation. Loadings under .33 (10% of variance) were suppressed. http://ir.lib.uwo.ca/cjsotl_rcacea/vol6/iss3/3 4 Gerwing et al.: Perceptions and Incidence of Test Anxiety Table 2 Principal Component Analysis of Student Perception Survey – Pattern Matrix Item Factor 1 Factor 2 Factor 3 h2 Importance of Institution Treatability TA Commitment to TA of TA 1) The consequence of test anxiety is as great as the .63 .42 consequences of other anxiety disorders 4) Test anxiety should be considered a true anxiety disorder .78 .62 5) Test anxiety deserves a high level of attention from the .78 .61 university and its professors 6) Test anxiety may prompt students to leave post .53 .30 secondary education 7) Test anxiety is not a severe enough form of anxiety to -.67 .46 merit recognition or treatment 8) Individuals with a comprehensive understanding of .70 .51 course material may perform poorly during exam situations due to feelings of anxiety related to testing 9) Professors should go out of their way to identify students .67 .45 with test anxiety? 10) Professors should identify test anxiety in a confidential .52 .27 manner 12) People who identify themselves as suffering from test .65 .45 anxiety should be accommodated in some manner (i.e., given additional time to write, write the test in a separate room) 13) Test anxiety causes individuals to perform poorly .64 .43 during academic testing Published by Scholarship@Western, 2015 5 The Canadian Journal for the Scholarship of Teaching and Learning, Vol. 6, Iss. 3 [2015], Art. 3 Table 2 (continued) Item Factor 1 Factor 2 Factor 3 h2 Importance of Institution Treatability TA Commitment to TA of TA 3) Students can feel confident that the university has a .74 .56 policy surrounding test anxiety 16) The university raises adequate awareness into the .81 .65 services available for test anxiety 18) Professors are adequately trained to recognize the signs .67 .45 of test anxiety in students 21) The university has adequate services to accommodate .81 .66 test anxiety 14) Test anxiety can be treated easily .84 .65 19) The treatment of test anxiety would be a difficult task -.79 .72 % of Variance 29.22 14.18 7.97 Note. N = 1099 (460 males) http://ir.lib.uwo.ca/cjsotl_rcacea/vol6/iss3/3 6 Gerwing et al.: Perceptions and Incidence of Test Anxiety Results TAI Characteristics of the population of student respondents for each survey are summarized in Table 3. Table 4 summarizes the results of the TAI-T, TAI-W, TAI-E, which were similar to that of the normative distribution of college undergraduates. Females had higher average TAI – T, t(1040) = 7.40 p < .01, TAI-W, t(1065) = 7.86, p < .01, and TAI-E, t(1065) = 5.11, p < .01, TAI-T scores than males. Table 3 Summary of Respondents for the Test Anxiety Inventory (TAI) and the Perceptions of Test Anxiety Survey (PTAS) Total Sample1 Survey Years Post- secondary Age education Ethnicity TAI PTAS 86% 20.32 (3.37) 1.50 (1.65) Caucasian Sample Size 1083 1099 Proportion Male 0.43 0.44 Proportion Undergraduate 0.13 0.13 Population Sampled Note. 1 Means and standard deviations are reported for age and years in post-secondary education. Demographic information is based on the PTAS respondents. Table 4 Means and standard deviations for the Test Anxiety Inventory (TAI) TAI Score Total Male Female Total 43.00 (14.47) 39.34 (13.07) 45.67 (14.81) Worry 15.83 (5.98) 14.75 (5.40) 16.61 (6.23) Emotionality 18.19 (6.36) 16.47 (5.80) 19.47 (6.45) Perceptions of Test Anxiety In the perceptions survey, 38.5 % (30.0 % of males, 46.3 % of females) of students self-reported having suffered from episodes of TA serious enough to reduce performance at some point in their academic career. Table 1 presents the prevalence of TA by faculty. Of the surveyed students, 11.3% stated that they would not seek help for TA as this would suggest weakness. Similarly, 20.5% of students would not approach their professors about TA, reporting that their instructor would be unable or unwilling to assist. Finally, 32.2% of the Published by Scholarship@Western, 2015 7