Aim. The literature on domestic violence makes it possible to perceive significant discrepancies in its presented scale, on the basis of the studies described (even several dozen percentage points). Such a situation leads to a number of problems of both a theoretical and a practical nature, which mainly come down to issues of diagnosing the phenomenon. The article presents three types of these differences: methodological (research procedures), cultural (ways of understanding phenomena) and ethical (norms of implementing scientific research), with deeper considerations focused on the importance of changes in the ethical rules of conducting research for the depiction of phenomena. Materials and methods. The authors use the example of their own research on domestic violence against minors (questionnaire auditory interviews in research samples for selected municipalities of Western Pomerania) to address the issue of the impact of ethical requirements for sociological research on the results obtained. Results and conclusion. A comparative view of the results of surveys carried out using the same methodology, but with the changing requirements of the ethical procedures, reveal a difference of up to 17.3% in terms of the overall scale of domestic violence, and up to 23% in terms of experiencing psychological violence. Faced with the cognitive needs of depicting social problems as accurately as possible, and the need to depict them in a manner consistent with the canon of ethical requirements, it turns out to be good practice to set up competent boards to optimise research projects. The authors share their experience of setting up and operating research ethics committees at their home university.