The Significance of Context in Testing Reading of Comprehension Intermediate Level ESL Subjects

Two research questions formed the basis for this study: 1) Is the amount of context associated with reading comprehension assessment significant when testing intermediate level ESL subjects? (2) Is the test performance of these subjects affected by the types of test items used, regardless of the amount of referent context available? The experimental test instrument presented two forms of referent context: minimum (one sentence or less) and maximum (two or more related sentences). Three test types were included: cloze, paraphrase and synthesis. Each item type was prepared in both context forms, in equal proportions, for a total of 60 test items. All test questions were presented with the multiple-choice response feature. Analysis of the data indicated no statistical difference between the results based on context form. However, significant differences were noted between the test item types, with the highest mean scores for paraphrase, middle level scores for doze, and lowest scores for synthesis.

Thesis Author: Chock, Roberta F.

Year Completed: 1985

Thesis Chair: Harold Madsen