Livingston's reliability coefficients and Harris' indices of efficiency were computed along with the classical internal consistency coefficients, KR-20's (Kuder-Richardson internal consistency coefficient), for 678 criterion-referenced tests in the A through E levels of an individualized mathematics program. The coefficients were carefully studied and compared with each other in relation to the number of students, the number of items, the percentage points of the mastery criterion score and the mean, the absolute value of difference of the mean from the mastery criterion score expressed both as a percentage and in a standard score form, the standard deviation, the proportion of mastery students, the shape of the score distribution, and the mastery status indices derived from the cross-tabulated tables of students' performance on the pretest and the Curriculum Embedded Test (CET), the pretest and the posttest, and the CET and the posttest. (Author/RC)