• Version
• Download 670
• File Size 622.31 KB
• File Count 1
• Create Date October 28, 2019
• Last Updated October 28, 2019

Developing items for a 1st grade screening test in reading: testing some basic principles

A strong body of evidence has emerged supporting the validity of multiple-choice items in different contexts and providing insights into good practice when it comes to developing and validating items (See e.g. Haladyna, 2004). However, both item content and item design must be adapted to the purpose of the testing. General guidelines derived from theory on mainstream reading development apply optimally to tests for normally distributed populations, as they focus on how the full range of a skill can be measured. For tests with a more specific purpose, such as identifying those who are at risk for reading and writing difficulties, it may be advisable to violate some generally recognized guidelines and to formulate new ones.
In this study we take the first steps towards formulating tentative guidelines on the design of multiple-choice items when the purpose is to identify struggling readers through group administered digital tests. The empirical grounds for the study come from three sources: a) A list of criteria describing our thinking during item development for a large-scale screening test is the deductive part of our data material. We confront these theoretical assumptions with two empirical sources: b) think-alouds with at-risk pupils during informal testing and c) a quantitative data material, namely the statistics from a large pilot study of the test items.
The approach taken emphasizes the application of knowledge on the unique considerations of those students who are likely to fall behind. The assumption is that this systematic focus will strengthen the predictive validity if not also concurrent validity. As such, the present study will contribute to our knowledge some tentative guidelines for item development for screening tests in reading.

Attached Files