A STUDY WAS CONDUCTED TO INVESTIGATE FEASIBLE CONDITIONS FOR ADAPTING SELF-INSTRUCTIONAL TECHNIQUES TO GROUP SETTINGS. THREE HYPOTHESES WERE TESTED--(1) THE TIME REQUIRED TO COMPLETE A PROGRAM BY A HETEROGENEOUS GROUP-PACED GROUP IS SIGNIFICANTLY GREATER THAN THAT OF A HETEROGENEOUS INDIVIDUALLY-PACED GROUP, (2) THE TIME REQUIRED TO COMPLETE A PROGRAM BY A HOMOGENEOUS GROUP-PACED GROUP DOES NOT DIFFER SIGNIFICANTLY FROM THAT REQUIRED BY A HOMOGENEOUS INDIVIDUALLY-PACED GROUP, (3) THE TIME REQUIRED TO COMPLETE A PROGRAM BY A HETEROGENEOUS GROUP-PACED GROUP IS SIGNIFICANTLY GREATER THAN THAT REQUIRED BY A HOMOGENEOUS GROUP-PACED GROUP. HIGH SCHOOL FRESHMEN PROVIDED THE SAMPLE. PRETESTING MADE UP THE HOMOGENEOUS AND HETEROGENEOUS GROUPS. TWO SECTIONS OF A LINEAR, SKINNERIAN-TYPE PROGRAM IN MODERN MATHEMATICS WERE USED FOR CONDUCTING THE EXPERIMENT. NONPARAMETRIC STATISTICAL PROCEDURES WERE USED FOR ANALYSIS OF DATA. THE "WILCOXON SIGNED-RANK TEST" WAS USED TO TEST HYPOTHESES ONE AND TWO. THE "WILCOXON COMPOSITE-RANK TEST" WAS USED TO TEST HYPOTHESIS THREE. ALL THREE HYPOTHESES WERE SUBSTANTIATED. (HB)