Search

The purpose of this report is to discuss the continuing difficulties school leaders and students face as they grapple with the "achievement gap" among students of different ethnic and socio-economic backgrounds. Several years ago, researchers at the Washington School Research Center sought to evaluate the "unique contribution of low income and ethnicity to academic achievement" (Abbott & Joireman, 2001) in Washington State. This current report is a replication of the 2001 study, which used data from the 2000 school year. The authors analyzed 2008 data to note any changes in the relationships among ethnicity, low income and achievement Washington. The 2001 report concluded that the percentage of low-income students by schools was much more predictive of achievement than school-level ethnic proportion. In the present study, the authors included school-level reading and math achievement as the outcome measures to parallel the 2001 report. Their analyses focused on the impact of ethnicity and low income on achievement, as well as the relationship all study variables. The analyses they present in this study are based on aggregated school-level data in 2008 available from the Office of the Superintendent of Public Instruction in Washington State. The data from 2008 grades 3, 4, 5, 6, 7, 8, and 10 support the conclusion that low income explains a much larger percentage of the variance in academic achievement than ethnicity. As noted in the 2001 report, the authors do not conclude that ethnicity is unrelated to academic achievement; rather it is a much less powerful predictor of achievement when low income is included in the analyses. Specifically, they affirm the 2001 finding that "the relationship between ethnicity and academic achievement is mostly indirect: ethnicity relates to low income, and low income in turn relates to academic achievement" (Abbott & Joireman, 2001). Another 2001 finding supported by their analyses is that, "while low income and ethnicity together explained a relatively high percentage of the variance in most of the outcome measures, a sizable percentage of the variance in achievement scores could not be accounted for by these variables" (Abbott & Joireman, 2001). In the present study, 20% to 47% of the variance in reading and math achievement across all the study grades was unexplained by low income and ethnicity as a set. This calls for subsequent analyses that might identify other influences on achievement that might clarify the low income--ethnicity relationships. Using "Percent White" as Ethnicity is appended. (Contains 3 tables, 4 figures and 4 footnotes.) [For the previous report, "The Relationships among Achievement, Low Income, and Ethnicity across Six Groups of Washington State Students. Technical Report #1," see ED454356.]

In April 2002, Jeffrey Fouts (2002) presented a longitudinal study of student performance in Washington using Washington Assessment of Student Learning (WASL) results from 1998 to 2001. Although constrained by the lack of individual student identification numbers at that time, he concluded that success on the WASL in the 4th grade was a strong predictor of achieving success in later grades. The opposite was also true; starting at the lowest levels of the WASL in the 4th grade strongly predicted less success in meeting the standards in later testing. According to Fouts, "A 4th grade Level 4 reading student was 28 times more likely to have met the reading standard three years later than was a 4th grade Level 1 reading student" (p. 20). Results were similar for math testing. These dramatic findings highlighted the need for assisting students early in their academic experience. However, according to Fouts the current education system did not "appear to be serving these students adequately" (p. 21). This study is an attempt to replicate and extend Fouts' 2002 study using the most recent data. Whereas Fouts used data from 4th and 7th grades to predict 10th grade achievement, the current study follows student progress from 4th through 10th grades. The following questions encompass Fouts' questions, but extend the findings through the more recent data: (1) How did students who took the 2001 7th grade reading and math WASL perform three years later on the 2004 10th grade WASL?; (2) How did students in different levels on the WASL in 1998 4th grade perform in 2001 7th and 2004 10th grades?; (3) What percentage of students scoring at various levels on the 4th grade WASL in 1998 met the WASL standard in the 10th grade in 2004?; and (4) Are there student factors that are related to student performance over time? A total of 8,463 students were included in the final database. The students in this study represented 33 districts and 318 schools around the state. The findings presented in this report extend Fouts (2002) study and are consistent with his report. The examination into student achievement over time revealed that past achievement was the best predictor of future achievement. One variable that is often cited as being influential in student achievement is ethnicity. The current study found that the likelihood of moving out of Level 1 reading was the greatest for Asian/Pacific Islander students and the least for Black/African American students. For math, Asian/Pacific Islander and White students were more likely to move out of Level 1 than Hispanic and Black/African American students. Hierarchical regressions revealed that mothers' education, amount of time per week doing homework, and three interaction terms predicted 16% of the variance in 10th grade reading achievement. Taken together, mothers' education, amount of time per week doing homework, and two interaction terms accounted for 23% of the variance in math achievement. The authors' study affirmed Fouts' (2002) conclusion that early success on the WASL was a strong predictor of later success. Although there were some differences between his predicted 10th grade scores and their actual scores, the dynamics of achievement progression over the grades was the same. Their last finding implies that there are some student practices that may be potentially helpful in partially mitigating the lack of early success. This is an area that deserves further investigation. This may assist leaders and practitioners in their quest for improving school success. (Contains 12 tables, 2 figures and 3 footnotes.) [For related report, "The Power of Early Success: A Longitudinal Study of Student Performance on the Washington Assessment of Student Learning, 1998-2001. Research Report," see ED479541.]

This study built on a 2001-02 classroom observation study of Washington K-12 and technical schools that identified the extent of constructivist teaching activity. Results from classroom observations found that strong constructivist teaching was observable in 17 percent of the classroom lessons. The other 83 percent of the lessons observed may have contained some elements of constructivist teaching, but up to one-half had very little or no elements of constructivist teaching present. More constructivist teaching appeared to occur in alternative schools and integrated subject matter classes. There appeared to be no differences among elementary, middle/junior, and high schools in the degree to which constructivist practices were used. This study explored the relationship of this practice to student achievement, examining the percent of variance in student achievement accounted for by constructivist teaching beyond that contributed by low-income. Data came from the original observation study and from school-level standardized test scores of 4th, 7th, and 10th graders. Results found large correlations between study variables (a negative correlation between school-level family income and student achievement, large positive correlations between constructivist teaching and student achievement, and a negative correlation between constructivist teaching and school-level family income). (SM)

This study examined how constructivist teaching and the organization of the learning environment relate to student achievement. The researchers used structural equation modeling (SEM) to examine the relationships among school and teaching attributes and student achievement in reading, writing, and mathematics as measured by the Washington Assessment of Student Learning (WASL). Data came from the average school-level scale scores of 79,494 students and from questionnaires completed by 4,307 teachers from 239 elementary, middle/juniors, and high schools in Washington state. The analyses show a strong positive relationship between school attributes and constructivist teaching. Schools that emphasize parental and community involvement and that have teaching staff who model and expect responsible behavior and mutual respect are more likely to have staff who use constructivist teaching methods. Constructivist teaching appears to have a meaningful influence on student achievement as measured by the WASL. Findings also show structural relationships among many school and classroom factors and student achievement. Maximizing achievement appears to be the result of an environmental shift in the school, a reinvention of education. (Contains 4 figures and 19 references.) (SLD)

This study was a replication of the method used by R. Bickel and C. Howley (2000), applying the approach to Washington state academic performance of fourth and seventh graders. While Bickel and Howley focused on the 8th grade Iowa Test of Basic Skills and the 11th grade Georgia High School Graduate Test, this study examined performance on the Washington Assessment of Student Learning (WASL). Hierarchical linear modeling was used to attempt to specify the joint relationships and cross-level interactions of two structural levels (district and school) on academic performance. Data provided by the Washington State Office of the Superintendent of Public Instruction consisted of all fourth and seventh grade WASL scale scores in reading and mathematics, aggregated to the school level. Findings show that large school district size is detrimental to achievement in fourth and seventh grades in Washington in that it strengthens the negative relationship between school poverty and student achievement. This finding replicates that of Bickel and Howley. The findings did not replicate another of the Bickel and Howley findings, however, district affluence in this study did not have a significant impact over the school size-student achievement relationship. The nature and configuration of Washington schools and the nature of the WASL may explain the discrepancy between the two studies. Multilevel findings of this study argue against the simplistic conclusion that reducing school or district size will automatically improve student achievement or result in more educational equity. (Contains 4 tables, 4 figures, and 15 references.) (SLD)

This report examines the overlap between student test results on the Iowa Test of Basic Skills (ITBS) and the Washington Assessment of Student Learning (WASL). The two tests were compared and contrasted in terms of content and measurement philosophy, and analyses studied the statistical relationship between the ITBS and the WASL. The ITBS assesses a student's level of achievement within the four broad domains of reading, mathematics, language, and vocabulary. The WASL assesses a student's level of achievement within the four broad domains of reading, mathematics, writing, and listening. The statistical analyses, which considered whether the mathematics and reading subtests of both tests would form two distinct factors or a single underlying factor, were based on individual student-level data for 45,601 students who were given the ITBS in grade 3 in 1999 and grade 4 in 2000. The ITBS and WASL are similar yet different in content and measurement philosophy. In general, students who score high on one test will score high on the other. But the scores on the ITBS and the WASL are not perfectly related, and in some cases, are only weakly related. Results also suggest that the reading and mathematics subtests are more clearly distinguishable on the WASL than on the ITBS. It seems likely that a factor analysis on the individual items would yield a much cleaner distinction between reading and mathematics on both tests. (SLD)

This study examined the relationships between students' academic achievement, income level, and ethnicity using aggregate school 3rd and 6th grade Iowa Test of Basic Skills scores for 1999 and 2000, 4th grade Washington Assessment of Student Learning (WASL) scores for 1999 and 2000, and 7th grade WASL scores for 1999 for all schools in Washington state. Data analysis indicated that low income explained a much larger percentage of the variance in academic achievement than did ethnicity. Ethnicity explained between 0-6 percent of the variance in academic achievement, after controlling for the contribution of low income. Low income, by contrast, explained between 12-29 percent of the variance in academic achievement. Combined with the finding that ethnicity explains approximately 32.7 percent of the variance in low income, the results suggest that the relationship between ethnicity and academic achievement is mostly indirect. Ethnicity relates to low income, and low income in turn relates to academic achievement. (Contains 14 references.) (SM)

Examines the cultural and spiritual significance of Snoqualmie Falls to two Snoqualmie groups who live within the traditional valley of the Falls or elsewhere in Washington. Summarizes controversy over plans by Puget Sound Power and Light Company to increase water use, thereby diminishing the Falls. (KS)