Description:

Purpose: Children identified early for reading problems have a much better chance of reaching grade level reading achievement than do students identified in Grade 3 or later. Many assessment instruments exist to find and profile these children, but most require significant teacher time and training to be appropriately used. The purpose of this project is to develop a computer-delivered early reading assessment system, called Early ICARE, which children can complete with minimal attention required by teachers. Early ICARE will enhance and merge with ICARE, a computer-delivered assessment battery that is currently under development and is being used to identify older children with or at risk for problems in reading comprehension. The new system will be Early - aimed at finding problems in young children in grades K-3 who speak English as a first language and for English language learners who can follow basic directions in English. It will be Comprehensive in identifying weaknesses that underlie difficulties in word reading, language comprehension, and/or speed. Finally, it is an adaptive measure in which children do only needed measures and items. The aim of Early ICARE is to quickly separate children who are without risk of problems from those who are at risk. Additionally, it will identify the nature of problems using engaging and non-threatening assessments. Beginning in Year 4 of the project, Early ICARE will be translated to Spanish for independent use with English- or Spanish-speaking children, from kindergarten to Grade 3. The system will provide an instructional profile to guide initial intervention either within, or independent from, a related instructional system. Finally, the researchers are developing a subsystem called Independent dynamic early estimates of reading (I-deer), to measure growth in reading and to predict later growth.

Project Activities: To test both the concurrent and construct validity as well as the reliability of the Early ICARE screening system, 1,200 kindergarten to Grade 3 children who represent a broad range of reading abilities and diverse ethnic and family income backgrounds will be tested with Early ICARE and with reliable and widely accepted standardized measures. Those who score in the lower 40 percent on word reading or reading comprehension will take shorter follow-up assessments one and two years later. A number of analyses will be conducted to examine what (e.g., specific early reading skills) predicts later comprehension problems in poor readers.

Products: The expected outcomes from this project include the development and validation of Early ICARE, a computer-delivered tool that provides early screening, profiling, and dynamic periodic assessment of problems underlying poor reading comprehension in young children, plus published reports.

Structured Abstract

Purpose: The purpose of this project is to develop an early reading assessment system, delivered via computer, called Early ICARE, which can be used with minimal attention required by teachers. The aim of Early ICARE is to quickly separate children who are without risk of problems from those who are at risk. Early ICARE will provide an instructional profile to guide initial intervention either within, or independent from, a related instructional system.

Setting: This project takes place in two large urban school districts in Colorado.

Population: A minimum of 300 students will be tested for each grade level, kindergarten through grade 3, totaling 1200 students. Participating students include a diverse mix of ethnicities and socioeconomic status, as well as English language learners.

Research Design and Methods: Early ICARE will have two levels: a kindergarten level designed to flag children at risk for reading problems, and a first-third grade level designed to identify beginning readers who are at risk for developing reading problems over time. From the outset, the system will be built for readers who speak English as a first language and for English language learners who can follow basic directions in English. Adapting it for Spanish-speaking students will occur in Year 4. Early ICARE will have two forms per measure, to allow for mid-year and end-of-year testing when needed. Tests of language comprehension, reading comprehension, rate, and fluency will be developed using virtual tutor architecture developed for Interactive Books (an interactive computer system used with young readers to support their vocabulary and comprehension of stories they read). Most other measures build on the current ICARE's architecture. Early ICARE takes many ICARE measures, originally developed for older readers, to younger ages, making the goals of developing this system, and the dynamic measures, and the longitudinal testing and modeling feasible in the time span of the project award.

Key Measures: All children will be assessed on six measures: (a) letter knowledge, (b) rapid naming, (c) phonological awareness, (d) phonological memory/sensitivity, (e) language comprehension, and (f) expressive vocabulary. A standardized measure of each ability measured in Early ICARE will be used to validate the test and the system, and for analyses of latent variables that cohere in and that predict later reading comprehension among poor readers. Speech recognition technology will be used for measures of rate and pausing, and for measures of vocabulary and sentence memory, where errors should be phonologically distinct. Researchers will analyze how well speech recognition measures expressive vocabulary in comparison with human scoring. Early ICARE will have two forms for each measure to enable pre-, mid-, and post-intervention assessment. More frequent, related progress measures will be developed for Independent dynamic early estimates of reading (I-deer).

Data Analytic Strategy: Content validities will be evaluated via expert rating. Construct validity will be analyzed with item/total and subtest/total correlations and exploratory and confirmatory factor analyses. To determine concurrent validity, the research team will correlate students' scores on Early ICARE and I-deer with scores on corresponding measures. Evaluative or diagnostic validity of Early ICARE will be examined with a discriminant function analysis to determine whether it correctly classifies students into groups with normal reading achievement vs. documented reading disabilities (as determined by Colorado state standards). Predictive validity will be determined. The latent structure of component processes of reading comprehension at the different ages will be examined via structural equation modeling. Structural equation modeling will also be used to evaluate the contributions of latent variables to later reading comprehension among poor readers, from kindergarten to first and second grade, and from first to second and third grades.

Project website: http://carla.colorado.edu/#

Related IES Projects: ICARE: Independent Comprehensive Adaptive Reading Evaluation System (R305G040097)

Products and Publications

Book chapter

Wise, B., Rogan, L., and Sessions, L. (2009). Sharing Research Knowledge With Teachers: The Story of Linguistic Remedies. In S. Rosenfield, anc V. Berninger (Ed.), Implementing Evidence-Based Academic Interventions in School Settings (pp. 443–477). New York: Oxford University Press.

Journal article, monograph, or newsletter

Wise, B., and Raskind, M. (2007). Technology and Reading Difficulties: Introduction to the special issue. Perspectives, 3: 7–8.

Wise, B., and Van Vuuren, S. (2007). Choosing Software Gems to Improve Your Child's Reading. Perspectives on Language and Literacy, 33(3): 34–38.

Proceeding

Bolaños, D., Ward, W.H., Wise, B., and Van Vuuren, S. (2008). Pronunciation Error Detection Techniques for Children's Speech. In Proceedings of Interspeech 2008 (pp. 1725–1728). Brisbane, Australia: International Speech Communication Association.

Brojde, C., and Wise, B. (2008). An Evaluation of the Testing Effect With Third Grade Students. In Proceedings of the 30th Annual Conference of the Cognitive Science Society (pp. 1362–1367). Washington, DC: Cognitive Science Society.