Evaluation (Test) Anxiety

CORRELATES OF TEST ANXIETY

An excellent meta-analysis conducted by Hembree (1988) is the source of many of the conclusions presented here, and Hembree source gives a more complete coverage of TA. Where appropriate, these conclusions are updated with more recent research findings.

Normative data on rates of TA are difficult to estimate because as of 2008 large-scale studies have not been conducted. However, a study by Methia (2004) found that more than one-third of school age children experience at least some TA. As mentioned earlier, it is assumed that TA has increased in the early 2000s, possibly due to an increased emphasis on testing in schools. Whatever the precise prevalence rate may be, it is clear that TA remains a concern for educators.

Not surprisingly, TA has been found to be negatively correlated with tests of IQ and various types of achievement and aptitude (Hembree, 1988). Again, these data are correlational and not causal. Thus, it is as plausible that low achievement causes TA as it is to say that TA causes low achievement. In terms of motivational factors, although there is a negative correlation between need for achievement and TA in the elementary years, there appears to be no relationship during the high school and college years. There is a strong negative correlation between self-esteem and TA. Moreover, students with high TA tend to have an external locus of control (i.e., feeling as if they have little control over the events of their lives). TA is positively related with defensiveness and both general state and trait anxiety, whereas it is negatively related with dominance.

One study tested the hypothesized negative impact of cognitive TA in the test preparation, performance, and reflection phases (Cassady, 2004). Students with high-cognitive TA reported lower study skills, rated tests as more threatening, and prepared less effective test notes. The high-anxiety students performed worse on tests and reported higher levels of emotionality. Cassady concluded that cognitive TA is associated with detrimental perceptions and behaviors in all phases of the learning-testing cycle.

Females typically are more prone to TA than are males, with the difference peaking during grades 5 through 10, and declining thereafter. However, this difference is not associated with lower test performance. In one study investigating differences between males and females, Chappell and colleagues (2005) found a small inverse relationship between TA and GPA in both groups. Low-test-anxious female graduate students had significantly higher GPAs than high-test-anxious female graduate students, but there were no GPA differences between low- and high-test-anxious male graduate students. Female undergraduates had significantly higher TA and higher GPAs than male undergraduates, and female graduate students had significantly higher TA and higher GPAs than male graduate students.

Both African American (in elementary school) and Hispanic students exhibit higher TA than do Caucasian students. Later-born children exhibit higher TA than do first-borns. Some researchers have proposed that TA among minority students may be partly explained by the stereotype threat hypothesis (Steele & Aronson, 1995). Based on the perception that there are negative stereotypes regarding certain groups' performance on standardized tests, group members experience higher anxiety on tests due to fear of being stereotyped or confirming the negative group stereotype. According to some, this hypothesis may help to explain why girls perform worse than boys on high-stakes math tests and why African American and Hispanic students perform worse on achievement tests than do Caucasian and Asian-American students (Osborne, 2007).

MEASURING TEST ANXIETY

Many scales are used for measuring test anxiety. Sarason's 1984 Reaction to Tests (RTT) questionnaire consists of four 10-item scales: two that measure the cognitive component (worry and test irrelevant thinking) and two that measure the emotional component (emotionality and bodily symptoms). The Friedban Scale (Friedman & Bendas-Jacob, 1997) of TA has 23 items covering three subdimensions: social derogation (eight items, e.g., “If I fail a test, I am afraid I shall be rated as stupid by my friends.”), physical tenseness (six items, e.g., “I am very tense before a test, even if I am well prepared.”), and cognitive obstruction (nine items, e.g., “In a test, I feel like my head is empty, as if I have forgotten all I have learned.”). Internal consistency reliability coefficients for the three subscales in one study were .93, .88, and .94, respectively (Sawyer & Hollis-Sawyer, 2005), indicating satisfactory levels of internal consistency. The Test Anxiety Inventory for Children and Adolescents (Lowe & Lee, 2005) is a 45-item self-report measure for grades 4 through 12. Coefficient alphas for the TAICA have ranged from .81 to .94. There is also the Test Anxiety Inventory (Spielberger, 1980) used solely with adults. For a shorter scale that can be also be used with older students, the Westside Test Anxiety Scale (Driscoll, 2007) has 10 items covering self-assessed anxiety impairment and cognitions that can impair performance and is designed to identify students with anxiety impairments who may benefit from an anxiety-reduction intervention.

TREATING TEST ANXIETY

Most early treatments of TA were behavioral and mainly systematic desensitization—a process in which the person is trained in relaxation techniques in the presence of increasing levels of the aversive, anxiety-producing stimulus (here testing environments). Unfortunately, although the early treatments were successful in reducing TA, they were less successful in improving test performance (Allen, Elias, & Zlotlow, 1980). Tryon (1980) suggested that students also need study counseling combined with desen-sitization to improve test performance. However, Hem-bree (1988) concluded that TA reduction programs generally result in higher test performance and GPA. Ergene (2003) conducted a meta-analysis that synthesized results from TA reduction programs. The most effective treatments appear to be those that combine skill-focused approaches with behavior or cognitive approaches. Individually conducted programs, along with programs that combined individual and group counseling formats, have produced the greatest changes. More research is needed to clarify the causal role TA reduction programs may have with regard to achievement and test performance.

Regardless of its prevalence and severity, TA will continue to be a concern of parents and educators. In an era of heightened school accountability, schools (especially administrators and teachers) will be interested in ways to reduce TA in students so as to enhance test performance. Merely treating TA by desensitizing students to the testing situation may not improve their test performance because some students' struggles may be simply due to inadequate preparation. It is important to assess both TA and academic preparedness before deciding on an intervention program. Interventions that provide training in study skills and test taking, combined with reducing the emotionality of the testing situation, are likely to be more successful in facilitating test performance.