x
Few studies have explored the relationship between student growth in math and reading grades across different time points. Different measurement methods may affect observed growth patterns. Additionally, multiple student characteristics have not been explored within the same analytical model, and may affect growth even when held constant across grades. In this article, we discuss the differences between growth in math and reading across time points and across SES.
Domain-specific connections between math and reading
The interplay between mathematics and reading grade development may have some underlying reasons. In this article, Marcia Barnes summarizes the evidence from four empirical studies to examine possible explanations for the comorbidity of reading and math difficulties. She also suggests possible intervention strategies. In the final section, she provides policy and practice recommendations to support both domains of learning.
The results showed that there is substantial overlap between math and reading difficulties at both grade levels. For both skills, the observed/expected ratio of comorbid difficulties ranged from 2.1 to 2.6 using a lenient criterion, to 3.9 to 5.1 using a stricter criterion. 151Please respect copyright.ＰＥＮＡＮＡmzVErzawwP
Furthermore, the percentage of children below the 7th percentile in both math and reading at Grade 3 was between 45 and 58 percent.
Effect sizes of growth
The growth rates of math and reading in a given cohort were compared using effect sizes. These results showed that the growth rates in math were larger than those in reading, and were more consistent. 151Please respect copyright.ＰＥＮＡＮＡcFNqT99JJd
However, the difference between math and reading growth was larger in students with lower starting achievement.
Effect sizes were also higher for math than for reading. The slope of math was smaller than the reading slope, and the coefficients were larger. There was a small effect of gender, favoring boys in all but one cohort. In addition, the percentage of variance that was accounted for by the math slope was greater than that in reading.
Although this paper helper study does not investigate the reasons for this disparity, it has provided important information to educators. This research can be used to understand how student growth varies across different demographic groups. It can also help policymakers in their work on reducing achievement gaps.
Effect sizes of growth on achievement gaps
A growing body of research has demonstrated that students' growth patterns change as they advance through elementary school. Previous research has found that annual growth rates for students in math and reading decrease as they progress through the elementary school years. 151Please respect copyright.ＰＥＮＡＮＡTqZEF2lmuI
This study looked at longitudinal growth patterns for 5,900 students who were tested at six different points in math and reading over two years. The authors calculated effect sizes of growth by using cohort-sequential latent growth curve models. Students who were less proficient at the start of school grew faster than students who were more proficient in their subjects.
For all students in the elementary school cohort, effect sizes were calculated across each grade. As students progressed through elementary school, the effect size in math and reading decreased, from g = 1.37 in Grade 1 to g = 0.50 in Grade 5. 151Please respect copyright.ＰＥＮＡＮＡwb5NLqhssl
However, growth across grades was still significant, as observed by the effect sizes for within-year growth. The authors found that students' growth was greater when they were tested at the start of the year than when they were assessed in mid-year.
Effect sizes of growth on SES
Previous studies have documented that student growth in reading and math decreases with SES. However, these studies focused on growth in one grade rather than across grades. 151Please respect copyright.ＰＥＮＡＮＡ0xt16KFTHO
In this study, researchers examined student growth across grades using normative samples from standardized achievement tests. The authors found that students with low math proficiency in kindergarten showed the slowest growth, compared with those who reached average math proficiency at the beginning of the grade.
The effect size is influenced by the range of achievement in the population. For example, a five-point increase in math grade on a scale with a mean standard deviation of 50 would have an effect size of 0.20. In contrast, a five-point increase would result in a larger effect size if the subsample was smaller.
Effect sizes of growth on race or ethnicity
Effect sizes should be interpreted using relevant empirical benchmarks. They should not be interpreted as desired policy effects. Instead, they provide snapshots of the effects that have been found in previous studies. 151Please respect copyright.ＰＥＮＡＮＡyu8UGbdXCL
Moreover, effect sizes may differ across studies due to different ways of measuring the same outcome construct. For this reason, empirical benchmarks are only helpful when the studies used are high-quality, and they measure similar outcomes.
Although effect sizes may differ across race and ethnicity, they are generally smaller for differences between Blacks and Whites. The gap between Black and Latino youth is generally higher than for Whites, and is greater when comparing high school GPA and math course-taking. 151Please respect copyright.ＰＥＮＡＮＡj05B75DFNk
However, the gap in academic preparation across STEM and business majors is smaller than among students in the humanities.
ns3.16.130.230da2
