Abstract
Answer sheets from an administration of a nationally standardized, reading comprehension test to 39 seventh- and 35 eighth-grade students were analyzed to determine the effects of response changing on a number of test characteristics and outcomes. Comparisons were made between characteristics of the test for these groups of students and those reported for the national norm groups. Parallels and distinctions between results of the current study and previous research are included. In general, previous findings regarding improved scores, ratios of gains to losses as a result of changed responses, and the percents of students changing answers or of answers being changed for the junior high school population were supported. Earlier contentions that reliability is decreased when students fail to reevaluate their responses, and that changed responses yield overall higher discrimination for tests were not supported.

This publication has 7 references indexed in Scilit: