Measuring First Graders’ Growth on Word Identification Fluency: A Prognosis for Later Performance?

Abstract

Background and Objective: Research has repeatedly found that students who have not developed sufficient reading skills by the third grade have lower rates of high school graduation and post high school outcomes. A growing body of literature has examined how the use of curriculum-based measurements (CBMs) can be used to identify these struggling readers earlier, with the hope of providing earlier interventions. The current study aims to build upon this past research, and investigate whether a student’s growth on an early reading skill CBM helps to predict third grade reading performance. Methods: The study utilized an existing data set from a southern Arizona school district that employs CBMs in school wide screenings three times a year. The sample consisted of 1304 students who were enrolled in first grade between the 2009-2010 and the 2011-2012 school year. Participants were 51.5% male and 48.5% female, with 64.6% of the sample identified as White, non-Hispanic, 25.8% Hispanic, 4.6% African American, 4.3% Asian, and 0.7% Native American. The dependent variables used were third grade Oral Reading Fluency (ORF) and the Reading Arizona Instrument to Measure Standards (Reading AIMS). The main independent variables were end of first grade Word Identification Fluency (WIF) scores, and a simple computation of WIF growth during first grade. Demographic variables including the percentage of students on free and reduced lunch at the student’s school, ethnicity, and gender were utilized in the analyses. Results: Analyses showed significant correlations between end of first grade WIF, end of third grade ORF, and third grade Reading AIMS. Significant correlations were also found between growth in WIF, end of first grade WIF, and end of third grade ORF. WIF growth did not show a significant correlation to Reading AIMS. Based on the results from a hierarchical multiple regression, the addition of end of year WIF performance to the regression model with only demographic information resulted in a statistically significant increase in the proportion of variance explained with an R2 of 51.1% for the model, F(1, 1400) = 1473.830, p <.001. Additionally, the simple measure of WIF growth significantly added to the prediction of end of third grade ORF performance, especially for those students in the “frustrational range” at the winter WIF screening. A separate hierarchical multiple regression was used to assess the value of end of year WIF and WIF growth when predicting third grade Reading AIMS performance. Results indicated that gender and ethnicity accounted for about 2% of the variation in AIMS scores, but adding end of year WIF scores into the model resulted in an increase in the proportion of variance explained with ΔR2 of 18.9% for the model, F(1, 1197) = 288.816, p <.001. The addition of growth in WIF provided less than a one percent improvement in variance explained. Conclusion: Students’ performance on WIF in first grade was related to both their performance on ORF in third grade, and on third grade reading AIMS. The results also suggest that students’ performance on third grade reading assessments is also related to demographic variables including ethnicity and gender. The results of the current study demonstrated that student’s growth on WIF during first grade is related to the student’s ORF performance in third grade; additionally, the growth in WIF was predictive of third grade ORF, especially for those students who were classified in the “frustrational range” at the winter WIF screening in first grade. However, the addition of this measure did not significantly improve the explained variance in third grade Reading AIMS.