出版社：Faculty of Law, University of Zagreb, Department of Social Work
摘要：The article presents a short overview of basic notions and research techniques used in studies on delinquent behaviour, with an emphasis on methodological issues of the application of self-report measures and longitudinal research design. The majority of studies related to the development of delinquent behaviour are transversal. Although their purpose is to gather data on correlates of delinquent behaviour, they offer little evidence on the causes of such behaviour, i.e. how and why delinquent behaviour begins, persists and ends. Defining its causes is of great importance for the decision-making process and for the development of adequate prevention and therapeutic programmes, which can be made possible only through longitudinal research. Due to the fact that both longitudinal and transversal researches have their advantages and disadvantages, the two types are usually combined. The majority of current studies on the development of delinquent behavior use self-report measures as a primary source of data collection. However, its use is connected with a number of methodological problems, such as sampling, validity and reliability. There is a tendency to overcome these problems by the use of other data sources, such as official data or observation. The combination of different data sources and longitudinal/transversal research enables simultaneous study and comparison of several cohorts, research on development of delinquent and criminal behavior from childhood into adulthood, as well as data collection from several sources on different forms of social behavior, out of which some are age-specific.
