Contextualizing the science of reading I—Vulnerability as a framework to understand and address reading

Contextualizing the science of reading I—Vulnerability as a framework to understand and address reading

First Author: Nicole Patton-Terry -- Florida State University
Keywords: At Risk Students, Assessment, Intervention studies, Academic success, Learning disability or difficulty
Abstract / Summary: 

The purpose of this two-part symposium is to contextualize the science of reading to better understand and address reading difficulty and disability among students who are vulnerable to experiencing difficulty in school.  Models regarding vulnerable populations have been applied in the social sciences and public health to understand disparities in various outcomes and to design interventions to improve those outcomes.  Importantly, vulnerability is governed by context; consequently, the causes of disparity are not always easily understood, in part because vulnerability is multifaceted. Applied to the field of reading, research suggests that both individual differences in learners and conditions within and outside of schools can increase vulnerability for reading difficulty.  Across two symposia, we will discuss findings from research teams that are considering multiple factors associated with the vulnerability of specific student populations to address reading assessment and development (symposium 1) and reading instruction and intervention (symposium 2) in school-based settings

Symposium Papers: 

Including biopsychosocial measurement into screening models for reading risk

First Author/Chair:Yaacov Petscher -- Florida State University
Additional authors/chairs: 
Hugh Catts; Jaime Quinn; Sarah O'Dell

Purpose: The purpose of this study was to test the extent to which classification accuracy for risk of word reading or language disabilities can be improved beyond using literacy screeners by including family history of reading problems, teacher ratings of student behavior, measures of growth mindset, and family caregivers’ adverse childhood experiences.

Method:  Using a three-form, planned missing data design, approximately 1,000 kindergarten students were administered a set of reading (print knowledge, phonological awareness, and word reading) and academic language (expressive and receptive vocabulary, listening comprehension) assessments in the fall and spring of the school year, and were asked about their growth mindset. Teachers rated each participant on selected measures on the BASC, and parents were administered the Adverse Childhood Experiences (ACES) survey.

Results: ROC curve analysis with one measure of letter sounds predicting the 16th percentile of the KTEA word reading outcomes showed an area under the curve (AUC) of .77. Including multiple measures of literacy skills improved the AUC to .83. The addition of growth mindset, ACES, and behavior ratings improved the AUC to .90. Finally, including selected interaction effects increased the AUC to .96.

Conclusions: Including biopsychosocial measures that reflect a whole person approach to measurement may hold value in screening for word reading and language disabilities. The data will be discussed in light of current and emerging research and practice. 

School-based, classroom-wide screenings for dyslexia and developmental language disorders

First Author/Chair:Tiffany Hogan -- MGH Institute of Health Professions
Additional authors/chairs: 
Julie Wolter; Jessie Ricketts; Rouzana Komesidou; Crystle Alonzo

Purpose: The purpose of this study was to determine the effectiveness of two classroom-wide screeners for determining risk for dyslexia or developmental language disorder (DLD) within an implementation science framework. Dyslexia and DLD are separate disorders that co-occur at 50%. Dyslexia is a word-reading disability and DLD involves poor language comprehension. 

Method:  We met with school administrators in two districts to develop a shared goal: to identify kindergarteners at risk for dyslexia and DLD. Before the administration of the screeners, we ensured that appropriate steps were taken (i.e., creating a data management system and training kindergarten teachers to administer the screeners). The language screener consisted of 20 complex sentences read aloud by the classroom teacher to student who individually circled a picture to represent each sentence. In the spring, a similar set-up will be used to administer a 20-item classroom-wide screener for dyslexia. We also collected data on feasibility, validity, & reliability, as well as qualitative data from teachers to identify areas of improvement.

Results: Initial results from 1939 children show that the DLD classroom screener was administered with good fidelity (.81) and reliability (split half = .81). The predictive power is unknown until the spring when confirmatory tests are administered, but promising prediction was show in the pilot study last year (sensitivity was .81 and the specificity was .68; 300 children). 

Conclusions: This study involves a unique researcher-school collaboration, to screen for dyslexia and DLD. The data will be discussed including barriers and facilitators for bridging research and practice. 


Comparing the effects of explicit and implicit alphabetic coding instruction on the application of orthographic patterns in reading and spelling

First Author/Chair:Alison Arrow -- University of Canterbury, New Zealand
Additional authors/chairs: 
Gail Gillon; Brigid McNeill; Amanda Denston; Amy Scott

Purpose: Without explicit instruction in orthography many beginning readers have difficulty using orthographic patterns in reading and spelling, with children with lower language abilities having more difficulty. This study examines the effects of a Tier 1 literacy intervention programme, for children in their first year of formal instruction. The programme provided intense and explicit understanding of how print works, through large-group vocabulary and phonemic awareness component and small-group structured and sequenced decoding instruction. 

Method: Thirteen Schools were randomly allocated to Strand A (experimental; n=165) or Strand B (control; n=121). All children who had been at school for 22 weeks or less (m = 63 months) were invited to participate in the research. The 10-week intervention was delivered by the classroom teacher with professional development and coaching. Control teachers provided their business as usual instruction which relied on implicit acquisition of orthographic patterns. Children were assessed on literacy and language measures pre-intervention and post-intervention. Literacy measures included sound-letter knowledge, pseudoword reading and spelling. Language measures included expressive vocabulary and story retell performance.

Results: Data collection will conclude in November 2019 and presented results will focus on the differences between the experimental and control children who presented with low language abilities, in their application of orthographic patterns to pseudoword reading and spelling compared to children who had implicit exposure to orthographic patterns.

Conclusions: This study contributes towards the research on the efficacy of full programme approaches that emphasise both vocabulary and explicit orthographic pattern use for students at-risk of developing reading difficulties.

Persistence and fadeout: Investigating long-term effects of a language intervention for children at risk

First Author/Chair:Åste Mjelve Hagen -- University of Oslo
Additional authors/chairs: 
Monica Melby-Lervåg; Arne Lervåg

Purpose: The purpose of this study was to examine longer-term effects of a large scale oral language intervention for preschoolers with weaker vocabulary skills.

Method: We conducted a randomized controlled trial in 150 classrooms in 77 preschools. Using a vocabulary screener, we identified the lowest 35 percentile (n = 301, 49.4% girls) of a cohort of 4-year-olds for program participation (M age = 57.8 months, SD = 3.4). We randomly assigned children at the classroom level to receive the oral language program or business-as-usually instruction. The children were assessed on a range of language measures pre-intervention, immediately after the intervention ended and at a seven-month follow-up. From 1st, 2nd and 3rd grade, we have student’s scores on national assessments of reading and math. In 4th grade, we tested student’s reading comprehension, listening comprehension as well as their word definition skills. 

Results: There were significant and moderate immediate effects across measurement types including taught vocabulary (d = .828), the intermediate language factor (d = .662) and the distal language factor (d = .563). On the seven-month follow-up, we did not see fadeout effects on the distal measures (d = .340), however the near and intermediate effects were weaker (d = .426 and d = .482, respectively) than at the immediate posttest. We will present analyses of the reading, math and language tests in school to examine fade- out. 

Conclusions:  This study is unique and will contribute much needed knowledge about fadeout and persistence of effects of early childhood interventions.  


First Author/Chair:DISCUSSANT Nicole Patton Terry -- Florida State University