Date of Award

2000

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Educational Leadership, Research and Counseling

First Advisor

Eugene Kennedy

Abstract

Recent developments in the philosophy of validity, highlighting the importance of investigating the consequences of assessment use, provide theoretical support for the move toward performance assessment. Numerous issues related to the choice of assessment approach (e.g., multiple-choice or performance-based) are not well understood today. This study investigated the potential interaction between assessment approach and the English skill of the examinee and the similarity of the information provided by multiple-choice and essay writing tests. Four hypotheses were examined in this study. Hypothesis 1 results indicated that students who were native speakers of English (regular students) achieved higher scores on the writing test than did the LEP students. Regular students were significantly different from LEP students in support/elaboration/organizations, sentence structure, and usage. However, there were no significant differences between these groups in writing mechanics and responsiveness to assignment. The results for Hypothesis 2 indicated that regular students scored significantly higher in all the eight subskills of the multiple-choice test than did the LEP students. Hypothesis 3 indicated a substantial interaction between English skill and assessment approach. The differences between the two groups on the higher-order elements of composition were much greater on the writing test than the multiple-choice test. Finally, the results for hypothesis 4 indicated that the factorial structure of the two tests were essentially identical for the two groups. It was concluded that the multiple-choice test consistently differentiated between the two groups of students studied in the direction consistent with the English skill levels of the students. In contrast, the performance-based writing test yielded expected differences for higher-order skills, but failed to separate the groups on lower level writing skills. This suggests that the multiple-choice test is more general in its utility than is true of the performance-based writing test. Also, it suggests that the writing test may be superior to the multiple-choice test at differentiating students on the higher-order elements of compositions. Conversely, it may be the case that the writing test exaggerates the differences between the groups raising the possibility of bias. The implications of this for high-stakes testing programs are discussed and recommendations are offered.

ISBN

9780599853652

Pages

111

Share

COinS