What Can Standardized Reading Tests Tell Us? Question-answer Relationships and Students' Performance
Journal of College Reading and Learning
This study examined the comprehension subsection of Nelson-Denny Reading Test Form G (Brown, Fishco, & Hanna, 1993a) and some urban developmental students' performance on it. Three types of question-answer relations were identified using Pearson and Johnson's taxonomy. Students' performance was expressed in their scores on the three types of question-answer relationships so that comparison of their performance on the three types of questions could be analyzed. The largest portion of textually
... plicit (TE) questions reflected a stress on such basic reading skills as perceptual match and recall of details explicitly cued in the language of the text. Textually implicit (TI) questions and scriptally implicit (SI) questions are equally divided, which measure meaning construction and prior knowledge. The students scored highest on TE questions and considerably lower on TI and SI questions, indicating a general strength in locating explicitly cued text information and a common weakness in sense making and prior knowledge. Further analysis of each type of question yielded detailed information of some instructional value. The students' performance on TE questions indicated the effect of cued text expressed in different sentence structures and located in different positions relative to answers. Their performance on TI questions suggested limited vocabulary and inefficiency in understanding the author's message differently phrased. The students' performance on SI questions showed inability to read critically and a narrow knowledge base. The findings have instructional implications: these developmental students need explicit instruction in reading strategies, a broader knowledge base, and more sophisticated analytical skills.