Estimation of Test Length for Domain-Referenced Reading Comprehension Tests

Publisher Website
Google Scholar
Abstract
A sampling methodology is proposed for determining the lengths of tests designed to assess the comprehension or readability of written discourse. It is an extension of Bormuth’s (1) work on transformational analysis within a domain-referenced measurement framework. The sampling units specific to this type of analysis and alternative sampling models are examined. Guidelines are provided for computing sample size and selecting the sample of sentences to which the transformational rules can be applied. A table of sample sizes for a given set of conditions is presented to aid the practitioner in utilizing the sampling theory.