Joakim, S. E. (2015). Help me fail: A study on testing accommodations for students with disabilities in writing assessments (Publication No. 3730563) [Doctoral dissertation, Capella University]. ProQuest Dissertations and Theses Global. https://www.proquest.com/docview/1735801010

Capella University (Minneapolis, MN); ProQuest document ID: 1735801010

The impact of various accommodations were investigated, including breaks, checking for understanding directions, extended time, familiar examiner, individual testing, small-group setting, specialized setting, testing across different days, and varying time of day. The researcher noted that some of the accommodations that were attempted to be investigated had no data points or had too few data points to be satisfied that comparisons yielded valid differences: large-print test booklets and students relaying to test administrator their responses to multiple-choice items.

Student participants were grade 5 (n=84) and grade 8 (n=72) students with various disabilities who had completed the state writing assessment either with or without accommodations in the 2012 testing year. The researcher examined data of the total grade-level samples of students with disabilities, as well as for subsets of students, including those with the primary disability categories of autism (n=17), emotional behavioral disabilities (n=21), other health impairments (n=47), specific learning disability (n=58), speech-language impairments (n=8), and multiple disabilities (n=5). There were not enough students with speech-language impairments and multiple disabilities to perform comparisons for both grade 5 and grade 8 students. The participant sample was drawn from Maine's largest school district; the population's demographic composition was described.

The extant data set for this post-hoc study was composed of all test-takers with disabilities who completed the 2012 New England Compact Assessment Program (NECAP) assessment of writing at the grade 5 and grade 8 levels. The researcher compared writing score group means for students using and not using accommodations within each grade level, and analyzed data subsets for score comparisons by independent variables, including some disability categories and some specific accommodations.

The researcher reported the numbers of students using each of the accommodation types, including by grade level: presentation (n=131), response (n=7), setting (n=80), and timing (n=107); overall, 132 students used at least one accommodation, and 24 students used no accommodations. Mean scores of the groups of students taking the writing assessment with and without accommodations at each grade level were compared. Grade 5 students not using accommodations scored higher than those who used accommodations, at a statistically significant level. Grade 8 students using and not using accommodations did not score significantly differently. When making comparisons for specific accommodations, the researcher found that there were no statistically significant differences for "administrator verification of student understanding following the reading of test directions," breaks, and extended time—at either grade 5 or grade 8 level. Grade 5 students not using read aloud, familiar examiner, or small group scored significantly higher than those using these accommodations, and grade 8 students not using individual testing scored significantly higher than those using individual testing. No student group means were found to be significantly higher when using specific accommodations. When making comparisons by disability categories, the researcher indicated that there were no statistically significant differences when using or not using accommodations for grade 5 students with autism and for grade 8 students with emotional behavioral disabilities. Students not using accommodations in grade 5 who had learning disabilities, and in grade 8 who had other health impairments, scored significantly higher than those not using accommodations. The researcher noted that some disability category subgroups had too few data points in one of the grade levels for comparison purposes. This observation applied to emotional-behavioral disabilities and other health impairments in grade 5, and autism and learning disabilities in grade 8 In addition, some subsets of students using or not using specific accommodations had too few data points in one of the grade levels for comparison purposes. This was the case for grade 8 students using the time of day accommodation. The researcher noted that these comparison results cannot be inferred to indicate a causal relationship, stating "students with approved accommodations may perform more poorly than students without approved accommodations to start with, and this is why they are allowed accommodations in the first place" (p. 145). Other limitations of the study were reported, and future research directions were suggested.