Finch, W. H., & Finch, M. E. H. (2013). Investigation of specific learning disability and testing accommodations based differential item functioning using a multilevel multidimensional mixture item response theory model . Educational and Psychological Measurement , 73 (6), 973–993. https://doi.org/10.1177/0013164413494776
Finch, W. H., & Finch, M. E. H. (2013). Investigation of specific learning disability and testing accommodations based differential item functioning using a multilevel multidimensional mixture item response theory model. Educational and Psychological Measurement, 73(6), 973–993. https://doi.org/10.1177/0013164413494776
This extant data analysis for demonstrating the multilevel multidimensional mixture item response theory model examined the interaction of various accommodations—as provided to students per their IEPs—with item performance. Accommodations especially noted were extended time, test directions read aloud, test items read aloud, alternate test setting, calculator use (for some math items), and scribe recording student responses.
Language and math performance data were sampled from a national grade 3 data set for representativeness in gender and ethnicity, as well as geographic region. Participant data were also coded for accommodation status, and disability status and category. The total sample was about 2,553 students, of whom 265 had disabilities, and 209 received accommodations.
Achievement scores were from a national examination data set, using 20 language items and 30 mathematics items. The data were analyzed separately at the individual test-taker level and at the school level.
The results pertaining most directly to associations between accommodations provision and performance were gained from covariate analysis in the MMMixRM, and were emphasized in this summary. The examinee-level latent classes included students with high mean performance on both math and reading; students with low mean performance on both math and reading, and who have higher incidence of math and reading disabilities and of receiving accommodations; and students with high math and low reading performance, who had higher incidence of reading disabilities than the first group. Several items in both assessments exhibited differential item functioning (DIF) for at least two latent classes, suggesting validity concerns. The researchers concluded that the model used in this analysis, as applied to the examinee-level data, permitted simultaneous analyses of different academic constructs which permitted a clearer demonstration of the complex picture of students with disabilities, their performance, and the connections with access to specific accommodations. Limitations of the study were reported, and future research directions were suggested.