Open AccessEVALUATION OF A CONDITION‐ADAPTIVE TEST OF READING COMPREHENSION FOR STUDENTS WITH READING‐BASED LEARNING DISABILITIES
Author(s) -
Stone Elizabeth,
Cook Linda,
Laitusis Cara
Publication year - 2013
Publication title -
ets research report series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.235
H-Index - 5
ISSN - 2330-8516
DOI - 10.1002/j.2333-8504.2013.tb02327.x
Subject(s) - fluency , test (biology) , reading (process) , mathematics education , psychology , reading comprehension , computer science , computerized adaptive testing , accommodation , developmental psychology , psychometrics , linguistics , paleontology , philosophy , neuroscience , biology
This study presents secondary analyses on a 2–stage test of reading comprehension for students with reading‐based learning disabilities (RLD). The present paper describes student perceptions of the test and its features as well as analyses focused on the routing test and associated cut score. The routing test contained typical state assessment content and was designed to route the RLD participants into 1 of 2 accessible 2nd‐stage tests, 1 accommodated and 1 not accommodated, and we refer to this structure as condition‐adaptive . The accommodated test was presented with a read‐aloud accommodation, and an oral reading fluency subtest was also administered as part of that test to evaluate decoding skill that may have been masked by the read‐aloud accommodation. This design allowed for componentwise measurement of reading proficiency at the 2nd stage with the goal of creating an accessible tailored test that would provide more comparable accommodated and nonaccommodated test scores for possible use for accountability. Overall, student perceptions differed depending on which 2nd‐stage test was taken, but most students indicated that they tried as hard on this field test as they did on other tests, and many expressed a preference for the read‐aloud accommodation. We found that routing test length was more strongly correlated with 2nd‐stage test performance for the higher performing RLD students, that relatively similar routing decisions could be made using a shorter routing test made up of a smaller subset of the available passages, and that the percentage of students routed into each 2nd‐stage test was sensitive to perturbations of the cut score.