The definitions used for these concepts will determine what behaviors and behavioral consequences are included in research that purports to focus on violence against women; thus, how one defines violence against women is one of the most important research decisions that a methodologist will make. One of the key issues in definitions of the components of violence against women pertains to broad versus narrow definitions of such violence. Typically, research focuses on narrow definitions of violence that encompass physical abuse, sexual abuse, or both. Few researchers will include a broader definition of violence that includes psychological, verbal, or economic abuse. Those researchers who favor broad definitions of violence argue that psychological or emotional abuse can be even more painful and damaging to one's sense of well-being than physical and sexual violence. Although there is a trend toward using broad definitions of violence against women, there is still significant variance in the incidence and prevalence rates of violence against women across studies, even when they use similar measures. This is due to sampling differences, different data-collection techniques, and other methodological factors. A significant issue in methodology is whether the questions asked of women reveal the information that is sought. Perhaps the most important need in newer surveys is for measures of context, meaning, and motive in incidents and patterns of violence. Another problem in surveys is how respondents view marital violence in relation to the content of criminal law. Since many women do not view how they are treated by their husbands as being criminal, they will not acknowledge criminal victimization in a family context when queried about victimization in crime surveys. In an effort to address the many problems of definition and perception in surveys, multiple measures of violence should be used. Various wordings of questions that probe respondents' experiences and perceptions of various types of violence should be used. Further, surveys that adequately address the complexities of violence against women in a variety of contexts and social settings must begin with a qualitative component. This might include indepth interviews with women and men, researchers, friends, family members, criminal justice officials, and shelter workers. Such information will sensitize research teams to the importance of using several supplementary open-ended and closed-ended questions that reflect women's subjective experiences. 56 references