OF THE EIGHT INITIAL EXPERIMENTS CONDUCTED BY RESEARCH AND INSTRUCTION (R AND I) UNITS, ESTABLISHED IN WISCONSIN SCHOOLS TO DEVELOP AND EXECUTE EXEMPLARY INSTRUCTIONAL PROGRAMS AND CONTINUOUSLY IMPROVE THEM THROUGH A PROGRAM OF RESEARCH, THREE EXPERIMENTS ACHIEVED SIGNIFICANT RESULTS. IN ONE, 99 SIXTH-GRADERS RECEIVED THREE TYPES OF SPELLING INSTRUCTION--WORKBOOKS, AN INTEGRATED APPROACH IN WHICH STUDENTS WERE TAUGHT A SPELLING METHOD AND WERE LED TO APPLY THIS IN WRITING NEW WORDS, AND INDIVIDUALIZED INSTRUCTION. ON RECALL AND RECOGNITION TESTS AND A PROOFREADING TASK, PREVIOUS ACHIEVEMENT LEVELS WERE MAINTAINED REGARDLESS OF THE METHOD OF INSTRUCTION USED, AND FEMALES SCORED HIGHER THAN MALES. SPELLING ERRORS ON THEME WRITING REVEALED THAT WORKBOOK INSTRUCTION SHOWED THE LEAST TRANSFER TO A WRITING SITUATION. HOWEVER, FREQUENCY OF DICTIONARY USAGE WAS MARKEDLY HIGHEST UNDER THE SPELLING WORKBOOK APPROACH. TWO OTHER EXPERIMENTS, INVOLVING NINE-WEEK STUDIES IN HANDWRITING INSTRUCTION, WERE CONDUCTED BY THIRD- AND FOURTH-GRADE R AND I UNITS. LEGIBILITY RATINGS OF HANDWRITING SAMPLES UNDER NORMAL, FAST, AND BEST CONDITIONS INDICATED THAT ON THE THIRD-GRADE LEVEL THE TRADITIONAL GROUP-METHOD OF INSTRUCTION WAS MUCH LESS EFFECTIVE THAN THE DIAGNOSTIC INDIVIDUALIZED APPROACH. IN THE FOURTH-GRADE LEVEL EXPERIMENT, NO SIGNIFICANT DIFFERENCES IN IMPROVEMENT WERE FOUND BETWEEN THESE METHODS OF INSTRUCTION, BUT IMPROVEMENT IN LEGIBILITY WAS GREATER IN GRADE 4, WITH EITHER TREATMENT, THAN IN GRADE 3. OTHER UNITS CONDUCTED FIVE ADDITIONAL SHORT EXPERIMENTS IN READING, BUT NO SIGNIFICANT RESULTS WERE OBTAINED. (DL)