Tid: 29 mars 2017, kl 13-14
Plats: B705

Abstract

In large-scale assessments one usually uses several test forms due to test security issues. Test equating methods make use of an appropriate transformation function to map the scores of one test form into the scale of another, so that scores are comparable and can be used interchangeably. There exist a large number of equating methods depending on how the test has been administered and which data has been collected. Some equating methods require that the same or equivalent test takers have taken the tests, and other equating methods use common items on the test forms. In some situations, one might not have access to common items and the test takers are neither the same nor can be viewed as equivalent. In such situations one can instead use information from covariates. The overall aim of this talk is to discuss different test equating methods when covariates are used to estimate the equating transformation. The different methods are illustrated and practical implications are discussed.