Trends in the Seasonal Dynamics of Test Score Gaps: Comparing Results Across Two Early Childhood Longitudinal Program–Kindergarten Cohorts

• In Event: Summer Learning and Test Score Gap Trends: New Descriptive Research, Methodological Issues, and Policy Implications

Abstract

1. Objectives

Educators and social scientists have long been interested in the extent to which summer vacation negatively impacts student learning (Cooper, Nye, Charlton, Lindsay, & Greathouse, 1996). Early studies generally found that literacy test score gaps by socioeconomic status (SES) grew more over the summer than over the school year (Alexander, Entwisle, & Olson, 2001; Heyns, 1978), and such evidence was used to support the arguments that schools served as “equalizers” (Alexander et al., 2001; Downey, von Hippel, & Broh, 2004). An under-examined question in this line of research is whether, and how, these seasonal patterns may be changing over time. In this study, we use the two ECLS-K data sets to examine the question of whether seasonal trends in test score gaps by race/ethnicity and SES differ for the two cohort of students.

2. Theoretical Framework

Entwisle, Alexander, and Olson (2000) proposed the “faucet theory” to explain seasonal test score gap trends. The theory posits that the “resource faucet” is turned on for all students during the school year, but over the summer, students from disadvantaged backgrounds lose access to many resources that remain available to more advantaged students.

3. Methods

We fit multi-level models for each subject area (with test occasions nested within students) separately in the ECLS-K:99 and the ECLS-K:2011 in order to estimate the monthly rates at which (absolute and relative) racial/ethnic and SES gaps changed over kindergarten, post-kindergarten summer, and first grade. Following past researchers (Downey et al., 2004; Quinn, Cooc, McIntyre, & Gomez, 2016), our models adjust for the fact that students were not test on the first and last days of school. We then test whether these rates of test-score gap changes in each season differ across the two data sets.

4. Data

We use restricted-use data from the NCES’s ECLS-K:1999 as well as the ECLS-K:2011. Each of these studies consisted of a nationally-representative sample of students attending kindergarten over the 1998-1999 and 2010-2011 school years, respectively.

5. Results

We find that test score gaps in kindergarten are generally narrowing at a faster rate in recent years compared to the past. However, the SES math gap widened more over the summer for the recent cohort. Furthermore, the school-equalizing pattern is reversed in first grade, with changes in learning rates favoring higher-achieving groups in the ECLS-K:2011 survey compared to the ECLS-K:1998.

6. Scholarly Significance

Our results complicate the narrative that schools act as equalizers. Estimates for changes in reading gaps in kindergarten provide a more positive picture of educational inequality in early childhood than estimates for first grade reading gaps. While changes in kindergarten test score gap dynamics favor lower-achieving groups in more recent years, changes in first grade test score gap dynamics do not favor lower-achieving students in recent years.

Authors

• David Quinn, University of Southern California

• Q. Tien Le, Los Angeles Unified School District