Alcoholism has long been considered a disease that primarily affects men. Researchers in the alcoholism field have contributed to this perception by using all-male or predominantly male samples in their investigations; a comparable amount of research on the unique issues involved in women's alcoholism has not been done. The assumption is that female alcoholism is or should be the same as male alcoholism. Because most alcoholism research has been done on men, the result has been a male-as-norm bias in the definition and analysis of alcoholism.
The male-as-norm bias describes male alcoholic behavior as the norm and the standard by which female alcoholic behavior is judged. This bias suggests that alcoholic women are sicker and harder to treat than are alcoholic men (Hoar, 1983) and defines women as a special population in the alcoholism field. Reed (1987) argued that the special population framework is used to incorporate the unexpected experiences of women, when what is necessary is to revise the fundamental assumptions of the male model of alcoholism. There is an inherent contradiction in that women's alcoholism is viewed as abnormal or different when compared to men's but women are researched as if their alcoholism were the same. Research that starts with a male-as-norm assumption has led to problems for both clients and social work practitioners. These problems are reflected in standardized assessment measures that lead to an imprecise or inadequate diagnosis of alcoholism for women, few options for specialized women's treatment programs, and a lack of reliable treatment outcome research (Hughes, 1990). Social workers in all areas of practice confront issues of alcoholism and alcohol abuse in their clients regardless of whether it is the presenting problem. This article identifies some gaps in the current research on alcoholism in women so that social workers who provide treatment services can recognize how a male-as-norm bias has shaped perceptions of women's alcoholic behavior and their responses to treatment.
WOMEN'S HEALTH PERSPECTIVE
An alternative framework in which to do research is a women's health perspective, which begins with the belief that women's health is distinct from men's and that "imbalances in social roles, and subsequently in power, equality, and control, are likely to affect women's health adversely" (Rodin & Ickovics, 1990, p. 1018). Although more research on women is needed, Hughes (1990) argued that "research conducted with unexamined assumptions derived from traditional scientific models ultimately may be more harmful to women than no research at all".
When doing research on women, it is important to acknowledge that society is male dominant and that lack of power can affect the development of a woman's self-esteem and sense of power over her own life (Sapiro, 1990). Research has shown consistently that women who have traditional sex-role values are more depressed and anxious (Rendely, Holmstrom, & Karp, 1984), have lower self-esteem and social adjustment (deMan & Benoit, 1982; Leavy & Adams, 1986; Orlofsky & O'Heron, 1987), and have a poorer sense of purpose and mastery (Frank, Towell, & Huyck, 1985) than women with nontraditional sex-role values. Much of the research on women's physical health has been limited to reproduction or the consequences of a woman's behavior on her fetus (Hughes, 1990). In addition to the negative effects of drinking on fetal development, research on women's alcoholism has often focused on the consequences of drinking related to a woman's failure as a wife and mother. For example, some early research (Parker, 1972; Wilsnack, 1974) found that sex-role conflict was at the root of women's alcoholism. Wilsnack (1974) suggested that because drinking produced feminine feelings in women, drinking reduced the conflict between conscious and unconscious sex-role identification. However, Lundy (1987) criticized the methods of these studies because they assumed that pencil-and-paper tests can assess the level of one's masculinity and femininity. She argued that these measures are the result of socially constructed ideologies of masculinity and femininity and are culturally biased. Indeed, measures that reflect ideologies that change are products of their times. Further, more recent disconfirming evidence has revealed that alcoholic women do not tend to drink abusively because of sex-role conflict (Hoar, 1983; MacAndrew, 1982-1983). Researchers looking at sex-role conflict in the early 1970s may have assessed a cohort effect, given the ambivalence about women's rights and equality (see Ryder, 1965, for a review of the cohort). Alcoholic women in the early 1970s may have struggled with their conscious and unconscious sex-role identification not because the conflict is endemic to alcoholic women, but because the society as a whole was ambivalent about women's changing roles and their effects on perceptions of women. Therefore, a male-as-norm bias in addiction models not only affects the theoretical questions in research, but also ultimately affects assessment techniques, treatment interventions, and follow-up care.
There are many gender differences in the …