Ex post facto designs

A procedure used in correlational studies to reduce the third-variable problem. Groups of subjects are carefully matched to eliminate variables likely to confound research results.