Mechanisms and Methods for Enhancing the Effectiveness of Reading Comprehension Interventions

Mechanisms and Methods for Enhancing the Effectiveness of Reading Comprehension Interventions

First Author: Dr. AMY BARTH -- Buena Vista University
Keywords: Comprehension, Intervention, Measurement
Abstract / Summary: 

Paper 1-Barth examines the effects of a text-processing reading comprehension intervention that targets linguistic comprehension through text-based discussions of grade-level, informational texts among middle grade struggling readers. Paper 2-Roberts determines the efficacy of a 4th grade reading intervention conducted with students who are very low reading comprehenders, on their performance compared with a BAU based on 1 year treatment or 2 year treatment. Paper 3- Lovett examines the impact of grade-at-intervention for young children with reading disabilities (RD) and variables that predict individual differences in immediate and longer-term reading outcomes. Paper 4-Piasta examines the extent to which targeting lower- and higher-level language skills affects language and comprehension outcomes in prekindergarten through grade 3. Paper 5-Kulesz explores different statistical models that can be formulated to determine whether the intervention impacts test-specific components of variance in comprehension.

Symposium Papers: 

The Effects of a Text-processing Comprehension Intervention on Struggling Middle School Readers

First Author/Chair:AMY BARTH -- UNIVERSITY OF MISSOURI-COLUMBIA
Additional authors/chairs: 
SHARON VAUGHN; PHIL CAPIN; EUNSOO CHO; AMORY CABLE

Purpose: We examined the effects of a text-processing reading comprehension intervention emphasizing listening comprehension and expressive language practices with middle grade struggling readers.
Method: A total of 134 struggling readers in grades 6-8 were randomly assigned to treatment (n = 83) and control conditions (n = 51). Students in the treatment condition received forty minutes of daily instruction in small groups of 4-6 students for approximately 17 hours.
Results: One-way analysis of covariance (ANCOVA) models on outcome measures with the respective pretest scores as a covariate revealed significant gains on proximal measures of vocabulary and key word and main idea formulation. No significant differences were found on standardized measures of listening and reading comprehension.
Conclusions: Results provide preliminary support for integrating listening comprehension and expressive language practices within a text-processing reading comprehension intervention framework for middle grade struggling.

The Effects of a One or Two-Year Reading Comprehension Intervention on Struggling Readers in Grades 4 to 5.

First Author/Chair:GARRETT ROBERTS -- UNIVERSITY OF TEXAS-AUSTIN
Additional authors/chairs: 
SHARON VAUGHN

Purpose: We examined the effects of a multi-component reading comprehension intervention on the word reading, fluency, and comprehension of 4th to 5th graders with significant reading comprehension problems.
Method: A total of 424 struggling readers in grade 4 were randomly assigned to 1 year of treatment (n = 133), 2 years of treatment (n=134), or control condition (n = 141) in the fall of their 4th grade year. Intervention students received thirty minutes of daily instruction in small groups of 4-6 students for approximately 7 months for either one or two years.
Results: No significant differences were found on standardized measures of spelling, sentence reading fluency and reading comprehension across groups, and a marginally significant difference was found on a standardized measure of word reading (d = .27, 95% CI [.00, .54]). The two year treatment group made significant gains in word reading (d = .43, 95% CI [.17, .69]) and sentence reading fluency (d = .39, 95% CI [.13, .65]) whereas the control group only made gains on sentence reading fluency (d = .31, 95% CI [.04, .58]).
Conclusions: Findings from this study suggest that when effects sizes are used as an indicator of impact, students who received treatment generally outperformed students in the control condition. Many gains were substantial when compared to previous interventions in the upper elementary grades. Implications include: (a) it may be necessary to provide even more intensive interventions for some students; and (b) additional individualization may be needed beyond the intervention setting.

Age and individual differences in intervention response in school-age children with reading disabilities.

First Author/Chair:MAUREEN LOVETT -- The Hospital for Sick Children and University of Toronto
Additional authors/chairs: 
Jan C. Frijters ; Karen A. Steinbach ; Maryanne Wolf ; Rose A. Sevcik ; Robin D. Morris

Purpose: This research addresses two questions: What is the impact of grade-at-intervention for young children with reading disabilities (RD)? What predicts individual differences in immediate and longer-term outcomes in decoding and reading comprehension?
Method: A multiple-component reading intervention was offered to small groups of four children with RD in 1st, 2nd, or 3rd grade. A total of 161 children received intervention, and 47 were control participants. Change during intervention and one year later (six testing points), and the influence of individual differences in predicting outcomes, were assessed using a battery of reading and reading-related repeated measures.
Results: Intervention children out-performed control children at posttest on all outcomes, with effect sizes ranging from .63 to 2.08 (Mean Cohen’s d = 1.07). On foundational word reading skills, children who received intervention earlier, in 1st and 2nd grade, made gains relative to controls almost twice that of children receiving intervention in 3rd grade. At follow-up, the advantage of 1st grade intervention was even clearer: First graders from the intervention condition continued to grow at faster rates than 2nd graders on 50% of key reading outcomes. On some outcomes with metalinguistic demands beyond the phonological, a posttest advantage was revealed for 2nd grade and for 3rd grade participants relative to controls. Predictors of growth during intervention included estimated IQ and receptive vocabulary. Growth during follow-up was predicted by vocabulary and visual sequential memory.
Conclusions: These findings provide evidence on the importance of early intensive evidence-based intervention for reading problems in the primary grades.

Targeting Lower- and Higher-Level Language Skills to Support Comprehension: Effects of Let’s Know!

First Author/Chair:SHAYNE PIASTA -- OHIO STATE UNIVERSITY
Additional authors/chairs: 
Language and Reading Research Consortium

Purpose: We examined the effects of Let’s Know!, a multi-grade comprehension intervention targeting children’s lower- and higher-level language skills.
Method: Across two sequential cohorts, more than 1700 children across five grades (prekindergarten through grade 3) were randomly assigned to experience one of two instantiations of the language-focused intervention Let’s Know! or to a business-as-usual control. Children assigned to Let’s Know! conditions experienced whole-class lessons targeting word knowledge, grammar, text structure, and integration (e.g., inferencing, comprehension monitoring) as implemented by their teachers over one academic year. Children completed pretest and posttest measures plus ongoing curriculum-based assessments.
Results: Multilevel analyses indicated large, consistent, and statistically-significant effects of Let’s Know! on curriculum-based comprehension monitoring and vocabulary probes relative to control (ds > 0.80) but minimal effects on probes tapping understanding of narrative and expository text structure. Overall, the two Let’s Know! instantiations, which varied in the emphasis and amount of practice afforded on specific language skills, did not appear to be differentially effective. Multilevel structural equation model analyses examining impacts on listening and reading comprehension outcomes are currently underway and will be completed by Spring 2016.
Conclusions: Initial results provide preliminary evidence supporting Let’s Know! as impacting select lower- and higher-level language skills . Subsequent results will determine whether the intervention impacts children’s reading and listening comprehension as the main outcomes of interest. Together, findings will elucidate the extent to which lower- and higher-level language skills, as key components of models of skilled reading, affect children’s comprehension abilities.

Comparison of statistical models examining whether intervention impacts test-specific components of variance in comprehension, broad aspects of the comprehension construct, or specific dimensions of comprehension that are common across assessments

First Author/Chair:PAULINA KULESZ -- UNIVERSITY OF HOUSTON
Additional authors/chairs: 
David Francis; YUSRA AHMED; JACK FLETCHER; SHARON VAUGHN

Purpose: Interventions that obtain results on experimental measures but not on standardized reading assessments necessitate a careful examination of the assessments involved. Recently, there has been interest in reading comprehension interventions to improve students’ inferencing abilities. In our Reading for Understanding (RFU) program of research, we have found positive effects in favor of the interventions on experimenter-designed assessments, but effect sizes on standardized assessments have been more modest. It is possible that the intervention differentially influences the specific aspects of comprehension captured by individual assessments. Alternatively, the tests may assess common skills, but at different levels of difficulty. The current paper explores different statistical models that can be formulated to determine whether the intervention impacts test-specific components of variance in comprehension, broad aspects of the comprehension construct common across assessments, or specific dimensions of comprehension common across assessments, but are possibly assessed at different levels of difficulty and in differing proportions on different assessments.
Method/Results: We examine these possibilities using explanatory item response models and data from a large randomized control trial involving a sample of over 1,000 students in middle-school and high-school. Intervention effects enter the models in different ways in an effort to better understand the nature of intervention effects and the factors that might explain differential impact across measures.
Conclusions: The paper addresses these questions in the context of a specific intervention, but the issues and models presented have implications for the broader analysis of intervention effects in reading research and these broader implications will be discussed.