A Record Linking System for Research on School Improvement

Abstract

A computer algorithm which links student test records using information available on two files is described and data on its accuracy presented. During the past four years it has been used to match over 1.6 million student records for use in a statewide school reward program. The results indicate that after minimal editing about .18 percent of the computer matches were incorrect and .44 percent of the potential valid matches were missed. By comparison, matches based on unedited student ID numbers produced almost no mismatches (.06%) but missed about 6 percent of the possible matches. For many applications the computer error rates should be tolerable and, since the system utilizes extant data, it is clearly cost-effective. A number of additional areas of application are described.

Get full access to this article

View all access options for this article.

References

Phillips

Bahn

A. K.

, and Miyasaki

, Person-Matching by Electronic Methods, Communications of the ACM, 5: 7, Baltimore, Maryland, pp. 404–407, 1962.

Cooley

and Cox

, An Automated Procedure for Matching Record Check and Household-Reported Health Care Data, Proceedings of the Survey Research Methods Section, American Statistical Association, Washington, D.C., pp. 418–423, 1981.

Freeman

R. W.

, 1981 to 1976 Census of Agriculture Record Linkage, Proceedings of the Business and Economic Statistics Section, American Statistical Association, Washington, D.C., pp. 91–99, 1983.

Crane

J. L.

and Kleweno

D. G.

, Project LINK-LINK: An Interactive Database of Administrative Record Linkage Studies, Proceedings of the Survey Research Methods Section, American Statistical Association, Washington, D.C., pp. 517–521, 1985.

Jaro

, Record Linkage Research and the Calibration of Record Linkage Algorithms, Proceedings of the Social Statistics Section, American Statistical Association, Washington, D.C., pp. 599–601, 1984.

Kelley

R. P.

, Blocking Considerations for Record Linkage Under Conditions of Uncertainty, Proceedings of the Social Statistics Section, American Statistical Association, Washington, D.C., pp. 602–605, 1984.

SAS User's Guide: Basics, (Version 5 Edition), The SAS Institute, Inc., Cary, North Carolina, 1985.

Mandeville

G. K.

and Rivers

J. L.

, Effects of South Carolina's Hunter-Based Program, Educational Leadership, 46: 4, Alexandria, Virginia, pp. 63–66, 1988/1989.

Mandeville

G. K.

and Rivers

J. L.

, Is the Hunter Model a Recipe for Supervision?, Educational Leadership, 46: 8, Alexandria, Virginia, pp. 39–43, 1989.

10.

Clark-Stewart

K. A.

and Fein

G. C.

, Early Childhood Programs, in Infancy and Developmental Psychology, Haith

M. M.

and Campos

J. J.

(eds.), Wiley, New York, pp. 917–1000, 1983.