Item position effects in a reading comprehension test: An IRT study of individual differences and individual correlates

Item position (IP) effects typically indicate that items become more difficult towards the end of a test. Such effects are thought to reflect the persistence with which test takers invest effort and work precisely on the test. As such, IP effects may be related to cognitive and motivational variables that are relevant for maintaining a high level of effort and precision. In this article, we analyzed IP effects in a reading comprehension test. We propose an IRT model that includes random IP effects affecting item difficulties and fixed IP effects affecting item discriminations. We found evidence for gradually increasing item difficulties and decreasing discriminations. Variation in IP effects on the item difficulties was systematically related to students’ decoding speed and reading enjoyment. The results demonstrate that the relationship between the overall scores and other variables is affected by respondents’ test-

taking behavior, which is reflected in the random IP effect.