Brown, D. W. (2007). The role of reading in science: Validating graphics in large-scale science assessment (Publication No. 3285596) [Doctoral dissertation, University of Oregon]. ProQuest Dissertations and Theses Global. https://www.proquest.com/docview/304834702
University of Oregon (Eugene, OR); ProQuest document ID: 304834702
The read-aloud accommodation was examined, administered via DVD compared with standard administration and compared with an accommodation in which the text responses—both key and distracter options—were replaced with pictures or graphics, called "graphically interpreted response options" (GIRO).
Participants (excluding the calibration study) were 868 grade 5 students from 55 classrooms throughout 17 schools in one district in the Pacific Northwest (U.S.). Students with reading disabilities numbered 54, and the remaining 817 participants had no disabilities. Demographic data including sex and ethnicity were also reported.
Scores on the researcher-constructed grade 5 science test of 30 items were analyzed. Students also were screened for their knowledge of science vocabulary used on the test; teachers scored students on reading and science content. The dependent variable comprised was an elementary-level science test. Students also completed a preference survey related to the three assessment conditions.
The read-aloud accommodation helped students with reading difficulties more than students whose reading skills were at or above teacher expectations. The GIRO format provided an additional boost, except for students with very high reading ability. The evidence gathered supported the conclusion that the scores taken from evaluations using standard science test administration may measure reading ability as well as science knowledge and skills. GIROs may have an important role in developing standardized science tests that reduce construct-irrelevant variance introduced by the dependence on student reading skills during testing. Students in special education expressed a preference for graphics (GIROs) at a higher rate, and text format at a lower rate, than students in general education. Limitations of the study were reported, and future research possibilities were suggested.