Anxiety and achievement in a multicultural situation

Oral reading comprehension test scores are generally determined by a variety of factors, not only by task difficulty and examinee proficiency, but also by stimuli extraneous to the testing task itself — such as group vs. individual testing situation, scorer method and bias, and examinees' sociolinguistic background, motivational disposition, and anxiety during the test. By giving the same examination under the same testing conditions to students, teachers and administrators generally believe that they are thereby being fair in their testing practice. However, if the same test situation debilitates the motivation and performance of some groups while increasing the motivation and performance of others, the assumption of fairness in standardised test procedure may be only illusory. Indeed, while stimuli in the test situation appear to be held constant from the examiner's point of view, they may interact with background variables to differentially affect levels of test performance (Cole and Bruner, 1970). Two types of anxiety effects have been differentiated in the literature, namely debilitative and facilitative effects (Alpert and Haber, 1960). The present study sets out to examine the effects of the following factors on oral reading comprehension test scores: (a) anxiety under varying test conditions, (b) students' gender, (c) students' native language, (d) the testing situation (individual vs. group), and (e) the relative difficulty of the English language texts. Two studies were conducted during an advanced English reading course at Haifa University to examine these issues. Study One, based on a sample of 170 students, was designed to examine the affective, cognitive, and sociolinguistic correlates within the test situation. No meaningful relationship was observed between students' attitudes and test performance. Study Two, using 112 students, was undertaken to examine the roles of anxiety, sociolinguistic background, and text difficulty within the oral testing situation in greater depth. Test scores were affected by gender and ethnic affiliation but not by text difficulty, reported level of anxiety, or group vs. individual testing situation.