THE DISCUSSION FOCUSES ON THREE AREAS: WHAT LEGITIMATELY CAN BE EXPECTED FROM PUBLIC OPINION RESEARCH IN CRIMINOLOGY; THE CONCEPTUAL AND METHODOLOGICAL DIFFICULTIES MOST OFTEN ENCOUNTERED IN THIS TYPE OF RESEARCH; AND BASIC PRINCIPLES FOR CARRYING OUT REPRESENTATIONS RESEARCH CORRECTLY. A REVIEW OF REASONS FOR CONDUCTING PUBLIC OPINION RESEARCH POINTS OUT THAT THE REALITY OF THE CRIMINAL JUSTICE SYSTEM IS NOT RESTRICTED TO WHAT THE SYSTEM ITSELF IS BUT ALSO ENCOMPASSES THE CONTRADICTIONS, CONFLICTS, AND EXPECTATIONS THAT EXIST IN SOCIETY RELATIVE TO THE SYSTEM. IT IS THIS ASPECT OF THE SYSTEM AND ITS FUNCTIONING THAT REPRESENTATIONS RESEARCH ADDRESSES. AN ANALYSIS OF THE PITFALLS OF BOTH QUANTITATIVE (QUESTIONNAIRE) AND QUALITATIVE (INTERVIEW) PROCEDURES IN PUBLIC OPINION RESEARCH TOUCHES ON SAMPLING, DATA-GATHERING INSTRUMENTS DESIGN, TYPES OF INTERVIEWS, AND DATA INTERPRETATION AND EXPLOITATION. REFLECTIONS ON THE FORMULATION OF CONCEPTS AND HYPOTHESES IN REPRESENTATIONS RESEARCH ARE PRESENTED. TWO METHODOLOGICAL PRINCIPLES TO GUIDE THE IMPLEMENTATION OF SUCH RESEARCH ARE SET FORTH: (1) PUBLIC OPINION RESEARCH SHOULD USE SUCCESSIVE APPROACHES--BATTERIES OF INVESTIGATIONS DESIGNED TO COMPLEMENT EACH OTHER AND TO AVOID PROBLEMS STEMMING FROM RESEARCHERS' PRECONCEPTIONS--AND (2) QUALITATIVE AND QUANTATIVE PROCEDURES SHOULD BE ALTERNATED IN THESE BATTERIES. A SEQUENTIAL RESEARCH BATTERY THAT EMBODIES THESE PRINCIPLES IS ILLUSTRATED SCHEMATICALLY. THE BATTERY CONSTITUTES A STRATEGY FOR PUBLIC OPINION RESEARCH ON THE CRIMINAL JUSTICE SYSTEM. (LKM)