Date of Award

10-26-2009

Degree Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Middle-Secondary Education and Instructional Technology

First Advisor

Gertrude Tinker Sachs, Ph.D. - Committee Chair

Second Advisor

Brendan Calandra, Ph.D. - Committee Member

Third Advisor

Joyce E. Many, Ph.D. - Committee Member

Fourth Advisor

Rihana S. Williams, Ph.D. - Committee Member

Abstract

Fluency instruction has had limited effects on reading comprehension relative to reading rate and prosodic reading (Dowhower, 1987; Herman, 1985; National Institute of Child Health and Human Development, 2000a). More specific components (i.e., error detection) of comprehension may yield larger effects through exposure to a wider range of materials than repeated readings (Kuhn, 2005b). Thirty-three students reading below college level were randomly assigned to a Repeated Readings (RR), a Wide Reading (WR), or a Vocabulary Study (VS) condition and received training in 9 sessions of 30 minutes in a Southeast community college. RR students read an instructional-level text consecutively four times before answering comprehension questions about it; WR students read four instructional-level texts each once and answered questions while the VS group studied and took a quiz on academic vocabulary. An additional 13 students reading at college level provided comparison data. At pretest, all participants completed the Nelson Denny Reading Test, Test of Word Reading Efficiency, Error Detection task (Albrecht & O'Brien, 1993), working memory test, Metacognitive Awareness of Reading Strategies Inventory (MARSI; Mokhtari & Reichard, 2002), a maze test, Author Recognition Test (ART), and reading survey. All pretest measures except for the ART and reading surveys were re-administered at posttest to training groups. Paired-samples t-test analyses revealed (a) significant gains for the WR condition in vocabulary (p = .043), silent reading rate (p < .05), maze (p < .05) and working memory (p < .05) (b) significant gains for the RR students in silent reading rate (p = .05) and maze (p = .006) and (c) significant increases on vocabulary (p < .05), maze (p = .005), and MARSI (p < .005) for the VS group at posttest. Unreliable patterns of error detection were observed for all groups at pretest and post-test. Results suggest that effects of fluency instruction be sought at the local level processes of reading using the maze test, which reliably detected reading improvements from fluency instruction (RR, WR) and vocabulary study (VS) in only 9 sessions. With significant gains on more reading measures, the WR condition appears superior to the RR condition as a fluency program for struggling college readers. Combining the WR condition with vocabulary study may augment students’ gains.

Included in

Education Commons

Share

COinS