Inferencing in a Standardized Reading Comprehension Test.

This paper reports an investigation of inferencing in Form Five of the Reading Comprehension component of the Iowa Tests of Basic Skills. The first step in the investigation consisted of an analysis of the language of the test to determine how the inferencing strategies employed might serve to mask answers. This preliminary analysis suggested that conflicting strategies available for answering a given question might make it difficult for students to determine the strategy for which the test would credit them. A second stage of the study focussed on a comparison between the observations made in step one and students' perceptions of ambiguities in the test. Six children each in grades three, five and seven representing a spread of reading ability and verbal skills were interviewed within a month following administration of the test. While their ability to talk about the ambiguities varied, they did detect them. This investigation led to the conclusion that while the test necessitates reading comprehension, it really measures the particular language skills possessed by test-wise children. (Author/AMH)