Test anxiety effects on African-American students' SAT performance

Construct Development and Scale Creation: Test Anxiety Effects on the Performance of African-American Students on the SAT I

Construct Development and Scale Creation

Construct and operational definition of test anxiety

Test anxiety has been identified as a significantly detrimental factor upon student performance, both in the testing context and beyond. While everyone experiences a certain level of anxiety when confronted with the prospect of an assessment, there can be no doubt that some individuals experience significantly higher levels of anxiety than others. Previous literature has indicated that more than one third of the students in U.S. schools experience some form of scientifically-defined test anxiety. This has only increased with the greater prevalence of standardized testing and testing requirements for college entrance exams such as the SAT. Moreover, because of the lack of self-confidence that either spawns test anxiety or derives from lower-than-expected test performance, test anxiety can have long-standing negative effects on students' learning and academic performance, as test-anxious students who do not perform well on standardized achievement tests may gain entry into less competitive schools and programs, as well as receive poorer grades, be more likely to be retained, and even drop out of school out of frustration at their underperformance in relation to their abilities (Lowe et. al, 2008, p. 216).

Once, test anxiety was defined simply as an expression of general anxiety in evaluative situations in which test-anxious students made more incorrect responses as a result of their levels of anxiety. However, a new model has emerged called the cognitive-attentional model which suggests that, contrary to 'normal' student anxiety' test-anxious students are more likely to divide their attention between task-relevant thoughts like problem-solving, and task-irrelevant thoughts like more holistic concerns about their general worth and competence. "The task-irrelevant thoughts interfere with the students' ability to focus on the test, thus lowering their [the chronically anxious students'] test performance" (Lowe et. al, 2008, p. 216). While certainly manifest as a cross-cultural phenomenon, it has been noted that little attention has been done to study the greater prevalence of test anxiety in certain population groups as opposed to other groups.

It is hypothesized that it is possible that because of negative cultural messages, certain groups may be more likely to have internalized negative messages about their self-worth and as a result may be more likely to manifest the cognitive-attentional model of test anxiety, which results in negative feelings of stress that inhibit test performance. Also, current models to measure test anxiety may be inadequate to fully assess the degree to which such anxiety is manifest within the individual in a cross-cultural fashion. For example, qualitative data of a cross-section of Indian youth "indicated culture-specific elements of test anxiety in Indian youth," deriving from the high stakes associated with exam performance and future schooling in India (Bodas et.al., 2008, p.387). However, the Indian student's anxieties in interviews were more manifest in somatic ways than their Western counterparts, and voiced in terms of concern about social derogation to a greater degree than personal shame, even though quantitatively-measured test anxiety "failed to confirm the importance of high-stakes environments on test anxiety" in their articulated responses on the questionnaires they were given (Bodas et.al., 2008, p.387). Researchers hypothesized that students might be unwilling to articulate their personal sensations of anxiety and to use expressions of somatization (such as feeling ill before a test) as a way of expressing their level of anxiety about testing (Bodas et.al., 2008, p.387).

The importance of a culturally relevant model is important, given the greater significance given to showing a strong performance on standardized tests in social mobility today. For example, even with 'all things being equal' -- preparation, grades, and socio-economic status, regarding the SAT I, a notable discrepancy has been identified in the ability of SAT I scores to predict freshman grades, undergraduate class rank, college graduation rates, and attainment of a graduate degree for African-American students as opposed to their white counterparts. It would be a valuable addition to the existing research literature to see if test anxiety while taking the SAT I was greater amongst African-American students, regardless of previous preparation (such as taking a preparatory course), high school grades, quality of high school attended, and socio-economic status (SAT I: A faulty instrument for predicting college success, 2007, Fair Test). It is also noteworthy that literature suggests that tests of the relationship of anxiety to performance alone shows that low confidence about one's general ability to succeed "reflects almost one full point increase in anxiety ratings; being unsure of one's ability (compared to high confidence) is also associated with a significant increase in anxiety" particularly in evaluative tests of ability, such as the SAT I (Reeve et.al 2008)

How much to you agree with these sentiments?

A worry that my test performance will affect my ability to get into my desired college. (Social Humiliation) felt sick to my stomach and shaky before taking the test. (Physiological Hyper-arousal) worry that I did not prepare enough for SAT before I took the exam. (Worry) worry that I am not smart enough to have done well on the test (Obstruction) do better on a test if I am slightly worried before I take it. (Facilitation)

Self-report format justification

Although one-on-one qualitative interviews might have some value, at least during the preliminary stages of research it would be necessary to submit the questionnaire to a wide range of individuals of both African-American and non-African-American extraction to get a sufficiently varied sampling. The sampling should be drawn from all states of the union and include individuals both categories of different economic demographic groups, who used a variety of test preparation methods, who from a variety of high schools, and also be balanced in terms of gender. Then, different segments could be compared -- for example, African-Americans from an affluent background could be compared with white Americans from an affluent background. As wide discrepancies have been noted across racial categories about the performance of males and females on the SAT I, each category should be subdivided according to gender.

Part II: Analysis and Justification

Reliability for this large sampling could be assured by using a technique known as split-half reliability. Half of the results would be split evenly in a random fashion. However, the random division would also be balanced demographically. This way it could be determined if African-American students and white students in both randomized categories showed consistent results in terms of their scores on measures of the different measures of test anxiety (Trochim, 2006, Types of reliability).

The validity of the test may be justified, given in the fact that its approach is derived from a test previously determined to be useful in assessing positive and negative test anxiety, the Test Anxiety Inventory for Children and Adolescents (TAICA). This test is useful because it can bring to light certain aspects of anxiety that are manifested somatically, even if the student is unwilling to admit about worrying about test performance on the more overt questions. Then, if a strong correlation was shown between high levels of test anxiety amongst African-American students and a lower-than-expected performance on the SAT when compared with their white counterparts, more qualitative work could be done to do a deeper examination of why this might be the case.

Final note: while certain other racial and ethnic groups have also shown less of a correlation between college performance and SAT I scores, only African-Americans and Caucasians would be compared in this study, as certain other factors (such as first-generation Latinos and Asian immigrants) might come into play other than stress, such as language acquisition, regarding test results (SAT I: A faulty instrument for predicting college success., 2007, Fair Test).