Assessing Reading Comprehension in Bilinguals

Diane August,Han‐Ya Annie Hsu,David J Francis,Catherine E Snow

doi:10.1086/510656

Abstract

A new measure of reading comprehension, the Diagnostic Assessment of Reading Comprehension (DARC), designed to reflect central comprehension processes while minimizing decoding and language demands, was pilot tested. We conducted three pilot studies to assess the DARC’s feasibility, reliability, comparability across Spanish and English, developmental sensitivity, and relation to standardized measures. The first study, carried out with 16 second‐through sixth‐grade English language learners, showed that the DARC items were at the appropriate reading level. The second pilot study, with 28 native Spanish‐speaking fourth graders who had scored poorly on the Woodcock‐Johnson Language Proficiency Reading Passages subtest, revealed a range of scores on the DARC, that yes‐no answers were valid indicators of respondents’ thinking, and that the Spanish and English versions of the DARC were comparable. The third study, carried out with 521 Spanish‐speaking students in kindergarten through grade 3, confirmed that different comprehension processes assessed by the DARC (text memory, text inferencing, background knowledge, and knowledge integration) could be measured independently, and that DARC scores were less strongly related to word reading than Woodcock‐Johnson comprehension scores. By minimizing the need for high levels of English oral proficiency or decoding ability, the DARC has the potential to reflect the central comprehension processes of second‐language readers of English more effectively than other measures.

Highlights

Knowledge IntegrationStandford-9 Spring 2001 (NCE) Total ReadingReading Comprehension na Maximum score: Story 1Note.—NCE ‫ ס‬normal curve equivalent.standardized reading comprehension assessment.Examination of responses to the clinical interview suggested that some sentence constructions used in the statements to be judged true or false confused the participants
The three pilot studies indicated that it is possible to use the Diagnostic Assessment of Reading Comprehension (DARC) with students as young as kindergarten, to develop texts that are decoded but yet place demands on students’ abilities to form inferences and integrate information across propositions and with background knowledge, that the yes/no scores accurately reflect students’ elaborated responses, and that the test can be used to differentiate inferencing from text memory and background knowledge
We have presented findings from three pilot studies that demonstrate the potential value, usability, and discriminative capacity of a new diagnostic assessment of reading comprehension

Summary

Introduction

Knowledge IntegrationStandford-9 Spring 2001 (NCE) Total ReadingReading Comprehension na Maximum score: Story 1Note.—NCE ‫ ס‬normal curve equivalent.standardized reading comprehension assessment.Examination of responses to the clinical interview suggested that some sentence constructions used in the statements to be judged true or false confused the participants. We wanted to ascertain the extent to which students had correctly reasoned about the relations among entities in the test, and the extent to which students’ dichotomized responses reflected their underlying thinking about the text and the relation among entities. To address this question, we examined students’ responses to the follow-up questions and scored these as correct or incorrect. We examined students’ responses to the follow-up questions and scored these as correct or incorrect We refer to these scores as justified scores because students had to justify their answer after giving it. These justified scores were compared to students’ original answers, which we refer to below as simple scores

Objectives

Methods

Results

Conclusion