The Development and Validation of a Formula for Measuring Single‐Sentence Test Item Readability

Abstract

This study describes the development and validation of the Homan-Hewitt Readability Formul~ This formula estimates the readability level of single-sentence test items. Its initial development is based on the assumption that differences in readability level will affect item difficulty. The validation of the formula is achieved by (a) estimating the readability levels of sets of test items predicted to be written at 2nd-through 8th-grade levels; (b) administering the tests to 782 students in grades 2 through 5; (3) using the class means as the unit of analyses and subjecting the data to a two-factor repeated measures ANOVA. Significant differences were found on class mean performance scores across the levels of readability. These results indicated that a relationship exists between students' reading grade levels and their responses to test items written at higher readability levels.

Student and school predictors of high‐stakes assessment in science

“…When a student responds incorrectly, it is not known whether it is due to lack of knowledge or inability to comprehend the test items (Homan, Hewitt, & Linder, 1994). If a test is to measure what it is intended to, there must be a distinction between a student's knowledge and that student's ability to respond to the test (Grise, Beattie, & Algozzine, 1982).…”

Exploring individual and item factors that affect assessment validity for diverse learners: Results from a large-scale cognitive lab

“…Access is also an issue for native English speakers who are poor readers (Clements, 1980;Helwig et al, 1999;Newman, 1977) and those who have learning disabilities (Tindal, Heath, Hollenbeck, Almond, & Harniss, 1998). Research to date has tended to focus on broad grouplevel differences and on the effectiveness of various test modifications, including linguistic simplification and read-aloud options, for group members (Abedi, Lord, Hoffstetter, & Baker, 2000;Helwig et al, 1999;Homan, Hewitt, & Linder, 1994). What is needed now for the improvement of test development and administration and test score interpretation is a more finely grained analysis of the interaction between individual and item factors.…”

Item readability and science achievement in TIMSS 2003 in South Africa

“…The method used in this analysis follows the method for measuring readability of singlesentence test items developed by Homan, Hewitt, and Linder (1994). They found that three predictor variables contribute the most in accounting for variation in the readability of sentences:…”