skip to main content skip to footer

Reliability of English Organization-Type Tests

Fan, Chung-Teh
Publication Year:
Report Number:
ETS Research Memorandum
Document Type:
Page Count:
Subject/Key Words:
English Composition Tests, Organization, Test Reliability, Writing Evaluation, Writing Processes


It was hypothesized that students taking no courses in mathematics during the experimental period would profit more by coaching than would students taking one or more courses in mathematics. Some of the correlations reported in RM-52-12 can be interpreted to support this hypothesis. It is reasonable that taking mathematics courses should correlate with mathematic scores. It is also reasonable that taking mathematics courses in the senior year should correlate more with a test taken in March than one taken in October, since the courses have had more effect by March. If coaching is able in part to replace mathematics courses, then the effect of coaching both those who take and those who do not take mathematics courses would be to obscure the difference between taking and not taking mathematics. This would hold down the correlation between number of courses being taken and test score received after coaching. It does.

Read More