Using a regression discontinuity design to determine the effectiveness of a vocabulary intervention for at-risk first grade readers

Author:
Ashworth, Kristen E., Curry School of Education, University of Virginia
Advisors:
Hallahan, Daniel, Curry School of Education, University of Virginia
Pullen, Paige, Curry School of Education, University of Virginia
Invernizzi, Marcia, Curry School of Education, University of Virginia
Kneedler, Rebecca, Curry School of Education, University of Virginia
Abstract:

The purposes of this study were to determine the effectiveness of a vocabulary intervention for first-grade students at risk for reading and language difficulties and to compare the results of a regression discontinuity design to those of an experimental design. The specific research questions were: (1) Do first-graders who are at risk of reading failure have significantly improved vocabulary learning outcomes (on measures of receptive, contextual, and expressive vocabulary) after receiving a second, more intensive tier of vocabulary instruction, than would be predicted if they were to receive only one tier of instruction? and (2) Are results from a regression discontinuity design of a first grade, tiered vocabulary intervention comparable to those from a randomized quasi-experimental design with an additional control group of at-risk children who did not receive treatment? Previous studies suggested that tiered vocabulary instruction provided in a shared storybook format improved word learning outcomes for students at risk for reading and language failure ( e.g., Pullen, Tuckwiller, Konold, Maynard, & Coyne, 2010; Tuckwiller, Pullen, & Coyne, 2010). With regards to the second research question, prior research suggested that regression discontinuity designs have produced similar results to those of randomized controlled trials, although some limitations were obvious (e.g., Aiken, West, Schwalm, Carroll, & Hsuing, 1998; Berk, Barnes, Ahlman, & Kurtz; 2010; Shadish, Galindo, Wong, Steiner, & Cook, 2011). The present study provided results through an analysis of archival data from a previous study (i.e., Pullen et al., 2010). The removal of the control group allowed the use of regression discontinuity analyses. To address the second question, I compared results from the original and present studies. Results from the analysis for the first research question provided evidence that rich, robust vocabulary instruction in a tiered format is beneficial to young students who are at risk of reading and language failure for receptive and expressive levels of word knowledge. Students who received Tier 2 instruction in addition to Tier 1 achieved higher scores on measures of receptive and expressive levels of target word knowledge, as well as total posttest scores which included a level of contextual knowledge, than would be predicted if they had not received the second tier of instruction. In comparing findings from the present study to those of Pullen and colleagues' study, I found that results were comparable for the receptive outcome scores in direction, magnitude, and significance, although the magnitude of the treatment effect for the RD design was slightly higher. Results were also comparable for contextual outcome scores with regards to direction and magnitude. I presented possible reasons for the different results between studies. Limitations and directions for future research are discussed.

Degree:
PHD (Doctor of Philosophy)
Language:
English
Rights:
All rights reserved (no additional license for public reuse)
Issued Date:
2012