A Comparison of Predictor Selection Techniques Using Monte Carlo Methods

Author(s):

Evans, Franklin R.; Linn, Robert L.; Patrick, Cathleen; Rock, Donald A.

Publication Year:

1969

Report Number:

RB-69-76

Source:

ETS Research Bulletin

Document Type:

Report

Page Count:

21

Subject/Key Words:

Comparative Analysis, Measurement Techniques, Monte Carlo Methods, Prediction, Predictor Variables, Sampling, Statistics

Abstract

Four methods of selecting a subset of predictors from a larger set were compared for 60 correlation matrices generated by Monte Carlo techniques. Two of the procedures were criterion-independent and two were criterion-dependent. The criterion-dependent procedures were found to be superior to the criterion-independent methods. The two criterion dependent procedures (forward and backward selection) worked about equally well except for the smallest sample size where the forward selection procedure was better.

Read More

• Request Copy (specify title and report number, if any)
• http://dx.doi.org/10.1002/j.2333-8504.1969.tb00755.x