MOST PROGRAM EVALUATIONS DO SUMMATIVE EVALUATIONS OF FIXED PROGRAMS, BUT EVERY FIXED PROGRAM HAS A HISTORY AND DEVELOPMENT PROCESS, DURING WHICH FORMATIVE EVALUATION MAY BE APPROPRIATE. CONVENTIONAL FIXED-CONDITION EXPERIMENTS ARE PROBABLY INCAPABLE OF PRODUCING CONCLUSIONS THAT REPRESENT REALITY. A MODEL IS HERE PROPOSED THAT SUBSUMES BOTH FORMATIVE AND SUMMATIVE PROCESSES AND IS APPLIED TO THE IDEALIZED RESEARCH AND DEVELOPMENT PROCESS, AS IT MOVES FROM INITIAL STAGES THROUGH TO A STABLE (CLIMAX) CONDITION OF PROGRAM OPERATION. THE KAMEHAMEHA EARLY EDUCATION PROJECT (KEEP), WHICH SOUGHT TO MAKE THE LEAST CHANGES NECESSARY TO ACHIEVE AVERAGE OR BETTER EDUCATIONAL ACHIEVEMENT, EXEMPLIFIES A RESEARCH AND DEVELOPMENT CLIMAX PROGRAM. NECESSARY EXTERNAL CONDITIONS FOR A CLIMAX PROGRAM INCLUDE LONGEVITY, STABILITY OF VALUES AND GOALS, FUNDING STABILITY, AND EVALUATOR'S POWER TO EXERT SELECTION PRESSURE OVER PROGRAM ELEMENTS. IN THE PROGRAM'S STAGES OF DEVELOPMENT, WAYS OF KNOWING OR EVALUATING INCLUDE EXPERIMENTATION, PROGRAM EVALUATION, DATA GUIDANCE, AND QUALITATIVE METHODS AND PERSONAL KNOWING. USES AND LIMITATIONS OF EACH WAY OF KNOWING AND SERAL STAGES, OR TRANSITORY ASSOCIATIONS OF ELEMENTS, ARE DESCRIBED IN RELATION TO KEEP. THE EVALUATION MODEL PROPOSED IS AN IDEAL, RATHER THAN A REFLECTION OF REALITY, E.G., KEEP SUBJECTS WERE NOT RANDOMLY ASSIGNED. DIFFERENT ASPECTS OF EVALUATION REQUIRE VARYING PRIORITIES FOR DIFFERENT TYPES OF VALIDITY (EXTERNAL, INTERNAL, CONSTRUCT, AND STATISTICAL). IT IS NOT YET KNOWN WHETHER KEEP IS IN A CLIMAX CONDITION. A PROGRAM CAN BE KNOWN TO BE IN CLIMAX ONLY WHEN EVALUATION AND SELECTION PRESSURE CONTINUES AND PROGRAM ELEMENTS REMAIN GENERALLY THE SAME. IN RESEARCH AND DEVELOPMENT WORK, HOWEVER, GOALS AND RESOURCES MAY CHANGE SO RAPIDLY THAT CLIMAX IS NEVER REACHED. FOR KEEP, ADVANTAGES OF CONTINUING EVALUATIONS APPEAR TO OUTWEIGH DISADVANTAGES. A CLIMAX CONDITION CANNOT BE DESIGNED IN ADVANCE, YET TYPICAL INTERVENTIONS TRY TO DO THIS. THE BASIC RESEARCH AND DEVELOPMENT MODEL IS TO MAINTAIN SELECTION PRESSURE FROM ALL WAYS OF KNOWING, SO THAT THE CLIMAX WILL REVEAL ITSELF. A TABLE AND A REFERENCE LIST ARE INCLUDED. (CFW)