NULL HYPOTHESES WERE TESTED TO DETERMINE THE DIFFERENTIAL EFFECTS OF (1) EXPERIMENTAL ATMOSPHERE AND ABSENCE OF SAME, (2) NOTICE OF TEST (10 SCHOOL DAYS) AND NO NOTICE (1 SCHOOL DAY), (3) TEACHER ADMINISTRATION AND OUTSIDE ADMINISTRATION OF TESTS, AND (4) TEACHER SCORING AND OUTSIDE SCORING OF TESTS. SIXTH-GRADE CLASSES (N=64), EACH FROM A DIFFERENT SCHOOL IN A LARGE MIDWESTERN CITY, WERE RANKED AND GROUPED INTO FOUR STRATA ON THE BASIS OF ARITHMETIC ACHIEVEMENT. WITHIN EACH STRATA, CLASSES WERE RANDOMLY ASSIGNED TO 1 OF 16 EXPERIMENTAL TREATMENTS GENERATED BY A FACTORIAL DESIGN USING 4 INDEPENDENT VARIABLES. WRITTEN INSTRUCTIONS WERE USED. THE RESULTING CLASS MEANS FOR EACH OF 3 SUBTESTS IN THE EXAMINATION WERE SUBJECTED TO A 4X2 ANALYSIS OF VARIANCE. THE ERROR TERM WAS COMPOSED BY POOLING SELECTED HIGHER-ORDER INTERACTIONS. THE FOLLOWING CONCLUSIONS WERE REACHED--(1) ADVANCE NOTICE OF TEST DATE HAS A SIGNIFICANT EFFECT UPON PUPIL TEST PERFORMANCE WHEN TESTS INCLUDE NOVEL CONCEPTS WHICH ARE EASILY TAUGHT, (2) TEACHER ADMINISTRATION OF STANDARDIZED TESTS HAS A SIGNIFICANT EFFECT ON PUPIL TEST PERFORMANCE AS COMPARED WITH TEST ADMINISTRATION BY OUTSIDE PERSONNEL, (3) EXPERIMENTAL ATMOSPHERE, COMBINED WITH NOTICE OF TESTING, RESULTS IN SIGNIFICANTLY HIGHER PUPIL TEST PERFORMANCE WHEN TESTS INCLUDE NOVEL CONCEPTS WHICH ARE EASILY TAUGHT, (4) NO NOTICE OF TESTING, COMBINED WITH OUTSIDE SCORING, RESULTS IN SIGNIFICANTLY LOWER PUPIL TEST PERFORMANCE, AND (5) OUTSIDE SCORES PRODUCED HIGHER-GRADE PLACEMENTS THAN TEACHER SCORERS IN HIGH-ACHIEVING CLASSES RATHER THAN IN LOW-ACHIEVING CLASSES AND VICE VERSA. (HB)