| This study investigates the jeopardy in content-area testing for English as a Second Language (ESL) students in university settings. A theoretical model is proposed which suggests that language proficiency, test anxiety, and test preferences are variables which predict ESL students' performance on content-area tests. Language proficiency was measured using the Michigan Test of English Proficiency, and a writing assessment. Test anxiety was comprised of both state- and trait-anxiety measures. Testing preferences assessed students' attitudes toward true/false, multiple-choice, short-answer/completion, restricted-response, and extended-essay formats using a semantic differential scale.;Both qualitative and quantitative data were gathered. First, thirteen ESL students were interviewed, and a content analysis of the data provides a general overview of the problems, preferences, strategies, and teacher practices related to content-area testing for ESL students. Second, quantitative data relating to language, anxiety, preferences, and test performance were gathered in a specific course--American Heritage--with 38 ESL students at Brigham Young University. Data were analyzed using a multiple regression approach investigating the relationships among variables in the theoretical model (i.e., path analysis). In addition, four American Heritage ESL students were interviewed.;The qualitative and quantitative findings suggest that the testing jeopardy for ESL students is a multidimensional phenomenon, with aspects internal and external to the student's control. In the course-specific context, the theoretical model accounted for 60% of the variance in test performance. Language proficiency and testing preferences did not account for a significant amount of variance; however, test anxiety was found to be the strongest predictor of students' content-area test performance. This study provides qualitative and quantitative evidence that the American Heritage professor's instructional and testing accommodations minimized the language proficiency and testing preference jeopardy in testing for ESL students. Implications for theory, pedagogy, and research are presented. |
