The main problems affecting research are definitional issues, the lack of appropriate control or comparison groups, the failure to address child abuse from a multitheoretical and multivariate perspective, and the difficulties imposed by retrospective research designs. These problems mean that although violence is known to be transmitted in many ways and that violence breeds some form of violence among some people, the relationship is not clear. Thus, it is impossible to say with confidence what amount or frequency of abuse, administered by whom and under what conditions, will promote the transmission or inhibit it. Not all people who have experienced violence will grow up to be violent, and a minority of maltreated children grow up to maltreat their own children. Therefore, prospective, longitudinal studies are needed, because the widespread dissemination and acceptance of the cyclical hypothesis has distorted both professional and lay perspectives on the consequences of child abuse. 57 references