Date of Award
Master of Arts
Speech, Language and Hearing Sciences
Dr. Nickola W. Nelson
Dr. Amy B. Curtis
Dr. Yvette D. Hyter
Masters Thesis-Campus Only
This research was designed to identify measures that would best differentiate students with typical language (TL) from students with language-learning disabilities (LLD) on a formal sentence-combining writing assessment task. Students from four age/grade levels (grades 1-2,3-5,6-8, and 9-12) (n= 51 TL and 51 LLD) completed the writing task and were assessed based on three candidate measures from word, sentence, and discourse levels. The candidate measures were proportion of word level errors, content units per T-unit, and proportion of content maintained from the original story. Trained research assistants coded the samples by hand and entered them into a computer program for these candidate measures. The sentence-level measure of content units per Tunits showed the best capacity to differentiate the two ability groups for the total sample and when divided by age/grade level The discourse and word-level measures also revealed statistically significantly difference between the two groups as a whole, with effect sizes of.32 and .21, but they did not work as well at every age level A total writing score measure that was a composite of the three candidate measures differentiated the two groups as well with an effect size of.48.
Magalski, "A Sentence Combining Task: Scoring Methods For Differentiating Abilities" (2010). Master's Theses. 321.