TRAINING PROGRAMS ARE DESIGNED TO BRING ABOUT DESIRED BEHAVIOR CHANGES IN ATTITUDES, SKILLS, AND KNOWLEDGE. TRAINING EVALUATION SHOULD DETERMINE IF THE CHANGES OCCURRED AS A RESULT OF THE TRAINING; THUS, RESEARCHERS SHOULD EXAMINE RELIABLE EVIDENCE WHICH CAN DEMONSTRATE EFFECTIVENESS. LEARNING IS A RELATIVELY PERMANENT BEHAVIOR CHANGE RESULTING FROM EXPERIENCE OR PRACTICE. LEARNING CONDITIONS ARE COMPOSED OF THREE CUMULATIVE PROCEDURES: (1) SPECIFICATION OF GOALS AT ONSET, (2) ARRANGEMENT OF AN EXPERIENCE, AND (3) EVALUATION OF THE OUTCOME. THE MEASUREMENT CRITERION TO BE USED IS A SUBJECTIVE MANAGEMENT PRESPECIFICATION OF GOALS THAT IS RELEVANT, RELIABLE, AND BIAS FREE. THE MODEL CONSISTS OF A TRAINING GROUP AND AN UNTRAINED CONTROL GROUP. BOTH GROUPS ARE MEASURED ACCORDING TO THE CRITERION VARIABLE PRIOR TO AND FOLLOWING THE TRAINING EXPERIENCE. A MINIMALLY ADEQUATE DESIGN INCORPORATES THE USE OF THE POSTTEST AND PRETEST ALONG WITH THE CONTROL GROUP PROCEDURE. A DESIGN LACKING EITHER OF THESE TWO CHARACTERISTICS IS NOT QUALITY EVIDENCE. THE MODEL CAN BE GENERALIZED TO ANY TRAINING EXPERIENCE AND WILL THUS BE USEFUL IN GOVERNMENT, EDUCATION, BUSINESS, AND INDUSTRY TRAINING EVALUATION. CHARTS AND REFERENCES ARE INCLUDED.