Anonymous Diploma Registration and research

(Play narration by clicking above.)

IDevice Icon Institutional accountability? Yes
  • Identifies the number of students who graduate from a single institution
  • Identifies the number of students who transfer and later graduate

IDevice Icon Internal program evaluation? Yes
  • Multiple methods use birthdates in two groups to estimate the overlap, i.e., the number of students who begin in a program and later graduate
    • Larsen (1994) on estimating a population from knowing the number of birthdates
    • Banks and Pandiani (1999) on estimating the overlap of two populations from the number of birthdates in each separate population (patented method)
    • Laska et al. (2003) on estimating the overlap of two popultions from knowing the exact birthdates in one or both separate populations
  • This class of methods was developed for evaluation of sensitive programs (mental health, HIV testing) and Banks and Pandiani's is recognized by the U.S. General Accounting Office (2001) as one innovative solution to record linkage problems when privacy is an issue.
  • Requires prior identification of groups (not an individual-level analysis technique), such as through propensity-score matching.

IDevice Icon Complex research? No, but there are better solutions
  • Longitudinal panel research remains the best tool for complex social-science analysis of educational experiences
    • Designed for sophisticated analysis and privacy
    • Quality control more efficient than administrative record-keeping (limited to the sample, not the whole population)
    • Examples: Adelman's The Toolbox Revisited (2006)
  • Examples
    • National Educational Longitudinal Study starting in 1988 (NELS)
    • Beginning Postsecondary Students Longitudinal Study, starting in 1990 (BPS)