A MODIFIED CRITICAL INCIDENT PERFORMANCE EVALUATION FORM FOR HOURLY MACHINING PERSONNEL WAS DEVELOPED, AND CERTAIN RATER AND RATING INSTRUMENT CHARACTERISTICS WERE INVESTIGATED FROM THE STANDPOINT OF THEIR EFFECTS ON THE VARIANCE OF PERFORMANCE. INTERVIEWS WERE CONDUCTED WITH FRONTLINE FOREMEN TO COLLECT PERFORMANCE SPECIMENS WHICH WERE EVALUATED AND INCORPORATED INTO A RATING FORM AND A RATING SCALE FOR OVERALL PERFORMANCE. THE FORM, CONSISTING OF 26 EFFECTIVE AND 26 INEFFECTIVE SPECIMENS, WAS TESTED BY 22 FOREMEN RATING A SAMPLE GROUP OF 198 MACHINISTS. THE RATIO OF EFFECTIVE TO INEFFECTIVE SPECMENS PROVED TO BE RELIABLE AND A RELEVANT MEASURE OF OVERALL PERFORMANCE. THE REGRESSION EQUATION OF THE EFFECTIVE-INEFFECTIVE VERSUS RATING SCORE WAS CONSISTENT WITH PREVIOUS RESULTS. A CRITICALITY FACTOR WAS OBTAINED FOR EACH SPECIMEN FROM THE AVERAGE FOREMEN WEIGHTS. THE INCORPORATION OF THE CRITICALITY FACTORS (THE MEASURES OF THE IMPORTANCE OF A PERFORMANCE SPECIMEN IN RELATION TO THE TOTAL SPECIMEN POPULATION) IN CALCULATING THE SPECIMEN RATIO DID NOT REDUCE THE VARIANCE OF ESTIMATE IN PREDICTING EVALUATION FROM OBSERVATION. BY REDUCING THE SPECIMEN LIST TO ITEMS WITH HIGH CRITICALITY FACTORS, THE EFFECTIVE-INEFFECTIVE RATIO WAS SIGNIFICANTLY REDUCED. THE INCREASE IN THE VARIANCE OF ESTIMATE DUE TO RATER DIFFERENCES DRAMATIZES THE NEED FOR A GENERAL EQUATION THAT WILL ESTIMATE AVERAGE PERFORMANCE EVALUATION RESPONSE FOR A GIVEN SET OF OBSERVED PERFORMANCES. THE FORM DEVELOPED IS RELIABLE AND RELEVANT AND MAY BE USED IN THE INDUSTRIAL COMPLEX FOR WHICH IT WAS DESIGNED. THE FORM IS INCLUDED. THIS DOCUMENT IS AVAILABLE FOR $1.50 FROM CLEARINGHOUSE FOR FEDERAL SCIENTIFIC AND TECHNICAL INFORMATION, U.S. DEPARTMENT OF COMMERCE, 5285 PORT ROYAL ROAD, SPRINGFIELD, VIRGINIA 22151. (HC)