The QUANTSELECT Procedure

The backward elimination technique starts from the full model, which includes all independent effects. Then effects are deleted one by one until a stopping condition is satisfied. At each step, the effect that shows the smallest contribution to the model is deleted. You request this method by specifying SELECTION=BACKWARD in the MODEL statement.

Suppose you specify the SELECT=SL method-option and the TEST=LR1 method-option to gauge improvement in quantile regression fit. At any step, the predictor that produces the least significant LR1 statistic is dropped and the process continues until all effects that remain in the model have LR1 statistics that are significant at the stay significance level (which is specified in the SLS= option).