Improving Student Achievement

Improving Student Achievement: What State NAEP Test Scores Tell Us

DAVID GRISSMER
ANN FLANAGAN
JENNIFER KAWATA
STEPHANIE WILLIAMSON
Copyright Date: 2000
Edition: 1
Published by: RAND Corporation
Pages: 312
https://www.jstor.org/stable/10.7249/mr924edu
  • Cite this Item
  • Book Info
    Improving Student Achievement
    Book Description:

    Why do students have different achievement levels across states? Is math achievement improving across states? Differences in average achievement levels across states are mainly traceable to differing family characteristics. However, students from similar families also score differently across states. These differences are related to differences in resource levels and in how resources are spent. States with high spending per pupil, lower pupil-teacher ratios, higher participation in public prekindergarten and higher reported teacher resources have higher achievement. Disadvantaged children are the most sensitive to low resource, and additional resources could substantially their scores. Between-state, rather than within-state, differences in resources appear to be the main reason for inequitable resource levels for students of lower socioeconomic status. The conclusion is that significant math gains are occurring across most states that cannot be traced to resource changes, that the rate of gain varies significantly by state, and that reform efforts are the likely cause of these gains. The results certainly challenge the traditional view of public education as unreformable.

    eISBN: 978-0-8330-4327-6
    Subjects: Education, Sociology

Table of Contents

  1. Front Matter
    (pp. i-ii)
  2. PREFACE
    (pp. iii-vi)
  3. Table of Contents
    (pp. vii-x)
  4. FIGURES
    (pp. xi-xii)
  5. TABLES
    (pp. xiii-xvi)
  6. SUMMARY
    (pp. xvii-xxxvi)
  7. ACKNOWLEDGMENTS
    (pp. xxxvii-xxxviii)
  8. ABBREVIATIONS
    (pp. xxxix-xl)
  9. Chapter One INTRODUCTION
    (pp. 1-10)

    States are the primary policymakers in several important areas of K–12 education. They are instrumental in determining how much is spent on education and how that money is used to reduce inequity in funding among school districts. States’ policies include setting teacher certification standards; establishing maximum class sizes and minimum graduation requirements; setting educational standards in subjects; and establishing methods of assessing student performance and methods of accountability for teachers, schools, and school districts.

    States have been the primary initiators of the latest wave of educational reform, starting in the mid-1980s, and a broad and diverse range of new...

  10. Chapter Two THE STATE NAEP ACHIEVEMENT RESULTS AND STATE FAMILY AND EDUCATIONAL CHARACTERISTICS
    (pp. 11-22)

    This chapter first describes the NAEP achievement tests and their results by state. Achievement scores partially reflect the characteristics of families in states, and we discuss the two components of family influence: family and social capital. We also contrast the differing family characteristics by state. Achievement scores also reflect the characteristics of educational systems, and we contrast the differing educational system characteristics by state.

    The NAEP tests are the only tests using nationally representative samples of U.S. students that can be used to track long-term national trends and accurately measure differences between states. The tests have been given to national...

  11. Chapter Three REVIEW OF THE LITERATURE
    (pp. 23-42)

    Our literature review traces the changing debate about the effectiveness of additional resources in education and the effects of specific uses of resources on educational outcomes. Two-thirds of the total variation in per-pupil spending is between states, with only one-third within states. Besides the wide variation in average spending by state, the states also differ markedly in how the resources are spent. Surprisingly, almost all of the variance in the expenditure patterns between states can be captured by a few variables. Educational systems can hire more teachers to lower class size or support instruction, provide teachers higher salaries, provide teachers...

  12. Chapter Four METHODOLOGY
    (pp. 43-54)

    The above evidence suggests that specifying aggregate models of state scores that take account of all educational characteristics from previous years is consistent with the Tennessee experiment. In fact, such aggregate analysis may be preferable to using the NAEP data at the individual or school level, since no previous-year data are available either for students or schools and since the quality of family data is decidedly inferior to that available at the state level.

    The previous studies also suggest that the estimated model should compare its prediction of pupil-teacher ratio with the results from Tennessee using population parameters similar to...

  13. Chapter Five TRENDS IN STATE SCORES
    (pp. 55-64)

    The current wave of educational reform started in the late 1980s, and changes continue. It is too early to expect to see the full effects of such reforms but not too early to expect some effects. Some educational reforms require significant changes in the behavior of organizations and large groups of individuals—a process that requires years, not months. Other reforms operate gradually because they effectively “grandfather” current students and teachers. For instance, changes in entrance standards for teachers will take 10 years or more to affect a sizable portion of teachers, and new graduation requirements for students are usually...

  14. Chapter Six ESTIMATING SCORES ACROSS STATES FOR STUDENTS FROM SIMILAR FAMILIES
    (pp. 65-74)

    The family variables in our equations account for about 75 percent of the variance in average state achievement scores.¹ So, the raw NAEP scores mainly reflect the family characteristics of students in the state and not the particular policies and characteristics of the educational systems. Raw scores are perhaps the important measure for parents and students, since they will partially determine future admission to colleges and labor force opportunities. However, the use of such raw scores in evaluating school systems or teachers has been heavily criticized, and the development of “value-added” indicators has been recommended (Meyer, 1996; Raudenbush and Wilms,...

  15. Chapter Seven EFFECTS OF STATE EDUCATIONAL POLICIES AND CHARACTERISTICS
    (pp. 75-84)

    We report the results using three different sets of resource variables. The first results include the most aggregate measure of educational resources only: per-pupil expenditure. The second set of results breaks per-pupil expenditure into four component parts: pupil-teacher ratio, teacher salary, teacher-reported resources, and public prekindergarten participation. These four categories of expenditures (together with a per-pupil transportation, special education, and LEP measure) account for 95 percent of the variance in per-pupil expenditures across states (see Chapter Eight). The third model replaces teacher salary with two categories that explain most of the variance in teacher salaries: teacher experience and teacher education....

  16. Chapter Eight ASSESSING THE COST-EFFECTIVENESS OF DIFFERENT RESOURCE UTILIZATIONS
    (pp. 85-94)

    Little research on the cost-effectiveness of educational programs, as opposed to their effectiveness, has been done. Levin (1983) stated the rationale and need for such analysis. Other studies emphasized the importance of focusing on productivity in education that analyzes both outcomes and costs (Hanushek, 1986; Monk, 1990, 1992; Hanushek, 1994; Consortium on Productivity in the Schools, 1995; Grissmer, 1997).

    One previous study made estimates of the cost-effectiveness of several major educational policy variables on student achievement: per-pupil expenditure, pupil-teacher ratio, teacher education, teacher experience, and teacher salary (Greenwald et al., 1996). The authors took their estimates of effect sizes on...

  17. Chapter Nine CONCLUSIONS
    (pp. 95-114)

    The state NAEP scores are the first achievement scores with representative sampling across states that allow comparison of performance across states. Since states are the leading initiators of educational reform, these scores may be the principle source for evaluating the effectiveness and cost-effectiveness of reform initiatives. These scores are also important because across-state variation accounts for two-thirds of the total variance in per-pupil expenditures among school districts. Because the reasons for variations in expenditures across states may differ from the reasons for variations in expenditures within a state, measurements of resource effects across states are of interest.

    However, a number...

  18. Appendix A STATE NAEP TEST SCORES AND STATE FAMILY AND EDUCATIONAL SYSTEM CHARACTERISTICS
    (pp. 115-140)
  19. Appendix B NAEP EXCLUSION AND PARTICIPATION RATES
    (pp. 141-152)
  20. Appendix C SOURCES OF BIAS
    (pp. 153-164)
  21. Appendix D THE TENNESSEE EXPERIMENT
    (pp. 165-174)
  22. Appendix E FAMILY VARIABLE DEVELOPMENT
    (pp. 175-186)
  23. Appendix F VARIABLE DEFINITIONS
    (pp. 187-196)
  24. Appendix G STATISTICAL RESULTS FOR ESTIMATING STATE TRENDS
    (pp. 197-224)
  25. Appendix H STATISTICAL RESULTS FOR ESTIMATING SCORE DIFFERENCES FOR STUDENTS FROM SIMILAR FAMILIES ACROSS STATES
    (pp. 225-230)
  26. Appendix I STATISTICAL RESULTS FOR ESTIMATING EFFECTS OF STATE POLICY AND EDUCATIONAL CHARACTERISTICS
    (pp. 231-248)
  27. Appendix J ROBUST REGRESSION RESULTS
    (pp. 249-252)
  28. Appendix K MAKING COST-EFFECTIVENESS ESTIMATES FROM THE TENNESSEE CLASS-SIZE EXPERIMENT
    (pp. 253-254)
  29. Appendix L REGRESSION COST ESTIMATES
    (pp. 255-256)
  30. BIBLIOGRAPHY
    (pp. 257-272)