Harvard Educational Review
  1. Fall 1975 Issue »

    Regression Analyses and Education Production Functions

    Can They Be Trusted?

    Daniel F. Luecke, Noel F. McGinn
    Many studies have purported to demonstrate that schooling has little independent impact on achievement and that administrators can do little to boost students' test scores. Daniel F. Luecke and Noel F. McGinn question such results and use variations of a computer simulation model to generate data sets similar to those collected by educational researchers. They subject the data generated to several kinds of aggregation procedures and regression analysis, and compare the statistics thus yielded with their knowledge of the causal relationships programmed into the data. They conclude that many "no significant effect" findings may be artifacts of statistical techniques used to analyze cross-sectional survey data.

    Click here to purchase this article.

  2. Share

    Fall 1975 Issue

    Abstracts

    Effects of High Schools on Their Students
    Christopher S. Jencks, Marsha D. Brown
    Regression Analyses and Education Production Functions
    Can They Be Trusted?
    Daniel F. Luecke, Noel F. McGinn
    Hierarchical Theories of Development in Education and Psychology
    D. C. Phillips, Mavis E. Kelly
    Call 1-800-513-0763 to order this issue.