Boscardin, C. K., Muthén, B., Francis, D. J., & Baker, E. L. (2008). Early identification of reading difficulties using heterogeneous developmental trajectories. Journal of Educational Psychology, 100, 192–208.
Students who struggle to read in early grades are most likely to continue as poor readers; however, when identified early (before third grade) and provided with appropriate instructional support, many of these students can learn to read. Potential reading difficulties can be prevented and developmental trajectories can be changed. For this reason, it is important to identify students at risk for developing reading difficulties as early as possible, before the onset of severe reading disability (RD), when possible.
The historical standard for identifying students with RD has been the presence of a discrepancy between the student's aptitude (usually measured as IQ) and his or her achievement. Often, at-risk students are identified as RD in third or fourth grade, when enough time has lapsed for a low achievement pattern to become evident. However, this method has not shown strong validity (Stuebing et al., 2002). Its reliance on measurement at a single time point and arbitrary cut points on continuous distributions make it unreliable as a means of identifying students who may benefit from more intensive intervention because of the measurement errors of the test and the absence of qualitative cut points (Francis et al., 2005). Waiting for students to fail before intervening is unfair to the students involved and ultimately counterproductive to the aim of helping all students achieve at the highest of possible levels.
Response to intervention (RTI) is a service delivery model focused on early intervention that can contribute data to identification of RD. However, in an RTI model, identification focuses on students who are hard to teach and the wait-to-fail component is eliminated because students immediately enter intervention. With RTI, student progress is monitored early and often on key reading skills that are known to predict later reading success. Students who struggle with these early skills receive more intensive levels of instructional intervention as a means of promoting their mastery of the skills that are essential to reading. Students continue to participate in increasingly intensive interventions until they are "back on track." Students who do not respond at a pace necessary to get back on track may have a reading disability, and the information collected during the "early and often" progress monitoring provides a primary data source for deciding about a given student's status. Identification as RD completes the RTI process; students who respond inadequately to increasingly large doses of quality intervention may benefit from identification and the special education services that entails. Measurement error and unreliability are reduced because of the use of multiple time points and a potentially self-correcting system that reduces errors in identification because of ongoing monitoring.
RTI assumes that a variety of distinct growth patterns characterize the group of early readers. Growth mixture modeling (GMM) makes it possible to test these assumptions in a statistical framework. Using student progress data from the measures of key reading skills, GMM can model student growth over time and also detect the underlying group structure that may characterize a sample. This question was explicitly addressed by Boscardin et al. (2008), who used GMM to identify subgroups of students who are most at risk for RD based on the distinct developmental profiles in early reading skills.
The data for this study were collected as part of the Early Assessment of Reading Skills project. A total of 411 students not selected for reading difficulties and who were not part of an intervention study were first assessed in kindergarten and followed through grade 2. Phonological awareness was evaluated on four occasions during kindergarten. Word reading skill was tested four times per year in grades 1 and 2. These early skills are known predictors of later reading success. In addition, rapid naming was measured once at the end of kindergarten.
The authors addressed the following questions by using GMM:
Can we identify a subgroup of students who are most at risk for developing RD based on the patterns of growth in early reading skills (phonological awareness and word reading)?
The researchers identified 10 subgroups of students with different growth trajectories in early reading development. Among these subgroups, 59% demonstrated changes in growth trajectories, whereas others did not. Of particular interest is the subgroup with the lowest profile. Students in this subgroup were very homogeneous and stable across time: They initially had poor phonological awareness skills, made no improvement during kindergarten, and subsequently had poor word reading skill with nearly flat trajectories during grades 1 and 2. Given the lack of growth, this subgroup of students is considered most at risk for developing RD.
What is the relation between growth in phonological awareness skills in kindergarten and subsequent word reading development in grades 1 and 2?
Students' phonological awareness skills at the end of kindergarten were predictive of word reading growth rate in grades 1 and 2. This finding means that students who have poor phonological awareness skills at the beginning of kindergarten are not likely to develop RD if the deficits are remediated by the end of kindergarten. However, if students' phonological awareness skill deficits are not addressed, they are very likely to show no growth in word reading development.
The GMM results indicate that students at high risk for developing RD can be identified as early as kindergarten. This study underscores the importance of early identification and prevention. If not identified early, at-risk students are more likely to develop RD and become increasingly less likely to read at grade level. Later reading difficulties are indicated by early deficits in the key skill areas addressed in this study (and summarized by prior research), such as phonological awareness and rapid naming skills. If students have poor phonological awareness skills and/or slow naming speed, they are most at risk for developing RD and may benefit from intensive targeted instruction in these skill areas. The key recommendation is to screen all students learning to read with brief measures of phonological awareness, word reading, and rapid naming on at least three occasions during the school year. Also, for students who are participating in more intensive interventions, it is important to regularly monitor their progress in these same skill areas as a means of making reliable decisions about discontinuing intensive intervention, increasing intervention's intensity, or identifying a student as having RD.
Francis, D. J., Fletcher, J. M., Stuebing, K. K., Lyon, G. R., Shaywitz, B. A., & Shaywitz, S. E. (2005). Psychometric approaches to the identification of LD: IQ and achievement scores are not sufficient. Journal of Learning Disabilities, 38, 98–108.
Stuebing, K .K., Fletcher, J. M., LeDoux, J. M., Lyon, G. R., Shaywitz, S. E., & Shaywitz, B. A. (2002). Validity of IQ-discrepancy classifications of reading disabilities: A meta-analysis. American Educational Research Journal, 39, 469–518.