Regional Correlates of Educational Performance In West Virginia Middle and High Schools Senator Bob

1
Regional Correlates of Educational Performance In
West Virginia Middle and High Schools Senator
Bob PlymaleChair, Senate Education Committee
Michael J. Hicks, Ph.D. Viktoriya Rusalkina,
MBA www.marshall.edu/cber
2
Study Design

• Technical Study of inputs/outputs in States High
  Schools and Middle Schools
• Set-Up analysis for school financing research
• Review existing research for gaps and
  misinterpretation
• Perform WV specific analysis
• Rank schools by Adjusted Performance Measures
  to provide for a double blind study of
  qualitative variables influencing educational
  outcome

3
Literature Review

• Extensive economic literature on school
  performance with large samples
• Total agreement that parental and community
  factors dominate outcomes
• Much dissonance regarding link between school
  inputs and outcomes
• Much research evidence erroneously applied in WV

4
This Analysis

• A Study of 6-12th grade students in all of West
  Virginia public Schools (2000-2001 school year)
• Model the relationship Y f(X, M, C)
• Y is school performance (test scores, AP tests,
  test taking, etc.)
• X is teacher quality (years and types of
  education, tenure, etc.)
• M is school attributes (age, administrative
  profiles, class size, overall size, etc.)
• C is control measures of regional characteristics
  (e.g school district educational attainment,
  incomes, poverty, etc.)
• Run models on roughly 250 variables to find mix
  that best explains school performance (using
  technical measures of statistical relationship).
  Model explains 50 to 57 of variation in school
  performance in main results.

5
What Effects School Performance

• Educational achievement of parents, regional per
  capita income and poverty are largest
  determinants of school performance (however we
  measured it). The effects mattered most for HS.
• Teacher education (average years of college) were
  very significant (both statistically and policy
  wise) for 10th and 11th grade students.
• Smaller classes, lower drop out rates and higher
  attendance at schools also are correlated with
  better test scores
• More densely populated school districts enjoy
  better school performance

6
What Does Not Effects School Performance

• Age of school
• Presence of alternative education (home school
  and private school)
• Attributes of administrators or their numbers
• School size (though bigger schools enjoy better
  SAT and AP test scores)
• Any visible 5 year trends in these variables

7
What We Dont Yet Know About

• How these variables affect performance in
  elementary schools
• How efficiently schools are using resources
• What are the scale economies associated with
  schools
• How do non-standard test takers affect
  performance
• How many of these mechanisms influence individual
  students

8
Additional Research Options

• Extend research to elementary schools
• Link financing to performance (efficiency and
  economies of scale)
• Perform double blind study of Adjusted
  Performance Measures to further isolate school
  factors explaining performance