[p.220] Many studies of child abuse, like other social science research, rely on promises of confidentiality to respondents. [...] If confidentiality is promised and cases of abuse are discovered, legal and ethical questions arise. In many states, but not all states, researchers fall into the category of those who are mandated to report cases of suspected and actual child abuse. [...]
However, although [these legal problems may be solved] the ethical problem still remain. [...]
Researchers should consider at least two alternatives: One is not to make promises of confidentiality that would then preclude the reporting of cases of child abuse. The second is that when such promises of confidentiality are given to make contingency plans for special intervention initiated and supervised by the researcher. Promises of confidentiality do not preclude the possibility of the investigator performing some sort of intervention. At the minimum, that intervention should include a great deal of encouragement to, and even pressure on, the respondent to report the case of abuse or to allow a report to be made.