Many of the technical problems of instrument construction/administration and sample selection are amplified when accurate and sufficient data on minorities are desired. An uneven response rate among racial groups may cast considerable doubt on the results of some self-report studies. Conceptual formulations and operational linkages used in current self-report research are also problematic. In casual analyses in particular, conceptualization and operationalization of independent variables such as 'broken home' and 'social class' are frequently not meaningful in the context of black experience. Some of these problems can be countered by selecting samples to ensure inclusions of sufficient numbers of minorities and officially labeled delinquents as well as adolescents not designated as delinquent. In the area of administration, reading the questions to small groups of respondents should increase the response rate among poor readers. When constructing scales of delinquent involvement, the use of standardized item values can correct bias due to scales often containing items of varying seriousness. Using a series of multiple regression equations can unravel the unique effects of intercorrelated independent variables. Variables to measure family structure should be reformulated to take into account that a significant number of youths, particularly minority youths, do not live with two parents in the home. If possible, the concept of social class should be broadened to include multiple measures usually considered significant in stratification research, such as education and income, as well as occupational prestige. A total of 12 notes and 24 references are provided. (Author summary modified)