THE USEFULNESS OF CRIMINAL STATISTICS IS LIMITED BY FACTORS SUCH AS THE INCREASING COMPLEXITY OF INTERPRETATION AS THE INSTRUMENT OF MEASUREMENT, PRACTICAL DIFFICULTIES IN REGISTERING DATA (TIME SHORTAGES, THE HIGH COSTS OF COLLECTING COMPLETE DATA), RESTRICTIONS ON INFORMATION GATHERING BECAUSE OF CONFIDENTIALITY CONFLICTS, AND THE QUESTIONABLE VALUE OF EVER-INCREASING QUANTITIES OF INFORMATION IRRELEVANT TO THE ACTUAL QUALITY OF INTERPRETATIONS. THE EXISTING ORGANIZATION OF CRIMINAL STATISTICS ACCORDING TO THE CATEGORIES SET UP IN LAW IS ACCEPTABLE, BUT COULD BE IMPROVED BY REGISTERING MODERN CRIMES (E.G., WHITE COLLAR CRIMES) AS CAREFULLY AS CLASSICAL ONES AND BY INCLUDING TRAFFIC CRIMES. STATISTICAL MEASUREMENT OF INTERNAL SECURITY IS HINDERED BECAUSE PURELY NUMERICAL REGISTRATION OF UNDIFFERENTIATED CRIMES BY STATES AND OVERLAPPINGS OF CRIME FORMS WHICH ARE ONLY PARTIALLY RECORDED MISREPRESENT THE NATURE OF CERTAIN CRIMES. SITUATIONAL INFORMATION SUCH AS CRIME SERIOUSNESS, OFFENDER-VICTIM RELATIONSHIPS, CRIME-RELATED FEAR, AND THE DEGREE OF PUBLIC ENDANGERMENT COULD BE INCORPORATED BY THE USE OF COMPUTER PROGRAMS. LITTLE EFFORT HAS BEEN MADE TO PROVIDE ADEQUATE OFFENDER PROFILES, AND ALTHOUGH SOCIAL PROFILES OF OFFENDERS ARE DESIRABLE, MOST INFORMATION REQUIRED IS NOT GATHERED DURING THE INVESTIGATIVE PROCESS BECAUSE OF MANPOWER SHORTAGES. ON THE OTHER HAND, DATA ON PREVIOUS CRIMINAL RECORDS COULD EASILY BE STORED, BUT OBJECTIONS HAVE BEEN RAISED TO THE POSSIBLE DAMAGE WHICH COULD RESULT FROM STORAGE OF SUCH INFORMATION. STATISTICS ON POLICE ACTIVITIES WOULD FACILITATE COST-BENEFIT ANALYSES, EFFECTIVITY ASSESSMENT, SECURITY ANALYSIS, EFFECTIVENESS OF RESOURCE DISTRIBUTION FOR CRIME PREVENTION, AND EVEN MULTILEVEL CRIME GEOGRAPHY STUDIES. ON THE WHOLE, MASS STATISTICS MUST REMAIN WITHIN MODEST PARAMETERS, AND USE OF SOCIAL RESEARCH BASED ON STATISTICAL SAMPLES SHOULD PROVE A BETTER SOLUTION TO THE INADEQUACIES OF STATISTICS ALONE. --IN GERMAN. (KMD)