Methodological and ethical issues in research on lesbians and gay men.
Research has an important role in educating social workers about lesbians and gay men and their human services needs. However, these issues received little attention in social work research until recently (Mallon, 1997). Studies of lesbians and gay men now constitute an emergent area of research, a trend that is likely to continue in fight of new requirements for inclusion of curriculum on lesbian and gay issues in Council on Social Work Education-accredited social work programs (CSWE, 1994). Because it is an emergent area, some researchers in social work might feel unprepared to enter it. This article aims to help social work researchers prepare for studies involving lesbians and gay men by raising their awareness of methodological and ethical issues in such studies. These issues involve theory and problem formulation, population definition, sampling and generalizability of findings, and preventing harm to study participants.
THEORY AND PROBLEM FORMULATION
Researchers must always scrutinize for possible bias the theories underpinning their research. Heterosexist bias is especially likely in studies of lesbians and gay men. Theories that assume or promote the superiority or normalcy of heterosexual sexual orientation and the inferiority or marginalization of other sexual orientations contain heterosexist bias (Hunter, Shannon, Knox, & Martin, 1998). For example, some theories assume that normal human development leads to a heterosexual sexual orientation. With the use of these theories, it is logical to conclude that a gay or lesbian sexual orientation in an adult is the outcome of an abnormal developmental process (Bieber et al., 1962; Erikson, 1980; Kronemeyer, 1980; Nicolosi, 1991; Socarides, 1978; Strean, 1996). When underpinning theories assume that gay and lesbian sexual orientations are pathological, research problems are likely to focus on the causes of the orientation and methods for changing it. Morin (1977) illustrated this point by noting that more than two-thirds of all studies of gay men and lesbians before the 1970s focused on sickness, diagnosis, and causation. Heterosexist bias can affect all stages of the research process. For instance, studies may ignore sexual orientation altogether, use identification of sexual orientation to devalue or stigmatize gay men and lesbians, or treat groups of gay men and lesbians as if they were homogeneous (Herek, Kimmel, Amaro, & Melton, 1991).
Whereas heterosexist bias in underpinning theories might result in the reinforcement of prejudicial ideas, biases in a study's problem formulation can reinforce popular denial of systemic oppression against lesbians and gay men. In particular, researchers might reach conclusions that inadvertently blame lesbians and gay men for their own oppression as a result of focusing on person-centered variables in the problem formulation. According to Kagel and Cowger (1984), the scientific conventions of "specificity of measurement, methodological controls, and unidirectional assumptions of causation" (p. 349) may lead to victim-blaming conclusions. The need to be specific in measurement causes researchers to select person-centered variables for study because they tend to be more circumscribed and less complex than environmental or situational variables. The desire to rule out alternative explanations for results causes researchers to apply as much control as possible over extraneous contextual factors. However, it is much more difficult to control the context of environmental or situational variables than person-centered variables. Because researchers are able to implement unidirectional causal models in studies more easily than reciprocal or multiple causal models involving situational variables, they are more likely to use them. For instance, a hypothetical study examining the relationship of men's sexual orientation (person-centered variable) to the number of sexual partnerships they have in young adulthood might conclude that gay men are sexually promiscuous, especially if the study disregards situational and environmental variables such as age cohort, gender roles, and the legal status of same-gender partnerships. By contrast, studies focusing on complex interactions of person-centered and sociocultural variables are more likely to increase understanding of gay and lesbian sexual orientation (Garnets & Kimmel, 1993).
According to Kitzinger (1987), the "liberal humanistic" theory underpinning most contemporary research on lesbians and gay men also biases researchers toward selecting person-centered variables for study. Like pre-1970s psychological research that was based on a pathological view of homosexuality, much of contemporary research assumes that sexual orientation is an individual, psychological characteristic. Liberal humanistic theory asserts that lesbians and gay men are basically the same as heterosexuals; that sexual orientation is not a "central organizing principle of the personality" (p. 45); that homosexuality is natural and normal; and that lesbians and gay men pose no threat to mainstream society. Kitzinger criticized research conducted from this perspective for depoliticizing lesbians and instead advocated research from the perspective of "radical feminist" (p. 62) theory. According to this theory, a lesbian identity is a freely chosen political challenge to patriarchal oppression. For example, Kitzinger used Q methodology to examine self-constructions of lesbian identity.
Researchers' personal and political biases can also affect a study's problem formulation. As Ripple (1960) noted, the identification of an appropriate problem for research should pass the "test of attitudes toward the findings" (p. 32). Researchers examining lesbians, gay men, and their issues should not be indifferent toward the potential results, but they certainly should not be strongly invested in one particular solution to the problem. A researcher should be able to face, honestly and comfortably, the possibility that the hypothesized results might not be found. Therefore, researchers with strong political or personal views supporting or opposing lesbians and gay men should refrain from engaging in research on this population unless they are willing to accept that the results might not support their views. If they cannot accept this possibility, the conditions are inappropriate for research. Under these conditions, researchers are too vulnerable to allowing conscious and unconscious sampling biases to affect the selection of participants or to constructing biased measurement instruments with leading questions or selective omissions. They also are vulnerable to violating the ethic of scientific integrity by fabricating, falsifying, or misrepresenting results that do not support their view, or else they might choose not to publish them at all.
Even in the absence of such strong views, researchers' assumptions "constitute an iron ring around the research" (Ripple, 1960, p. 39). Assumptions must be stated explicitly, especially when they are based on values rather than previous empirical research, or else their merits cannot be fairly judged. For instance, the use of sexual behavior as the sole dependent variable in a study of sexual reorientation therapy effectiveness reflected an unstated assumption by Bieber (1976) that sexual behavior and sexual orientation are synonymous.
DEFINING THE POPULATION OF INTEREST
Obtaining representative samples is one of the greatest challenges faced by researchers studying gay men and lesbians. One reason is that the population can be defined in a variety of ways. In one of the earliest attempts to estimate the number of gay men in the United States, Kinsey, Pomeroy, and Martin (1948) found that four percent of their sample reported engaging exclusively in same-gender sexual behavior as a lifelong pattern. Ten percent engaged primarily in same-gender sexual behavior for at least three years between ages 16 and 55, and 25 percent of the sample engaged in more than one same-gender sexual experience between ages 16 and 55. Which of these behavioral patterns reflected the best definition of "gay man?" Are some ways of defining gay men and lesbians more accurate than others? Should the definition be strictly behavioral, or should it include affectional preferences and fantasies? Can people of any age be gay or lesbian, or is there a minimum age limit?
Many studies use a self-labeling definition, in which gay men or lesbians are simply people who voluntarily label themselves as such (Hare, 1994; Martin & Knox, 1997c; Nardi & Sherrod, 1994; Peters & Cantrell, 1991; Pilkington & D'Augelli, 1995). Because these labels have a social stigma, studies assume that people who are not actually gay or lesbian would be unlikely to endorse them. This definition is appropriate for many studies. However, such studies must indicate that the population of interest is self-identified gay men or lesbians. That is because participants who self-identify themselves as such might actually have a range of sexual orientations. According to Haslam (1997) there is evidence that male sexual orientation is a continuous variable, although its behavioral manifestions might be discretely organized. Rothblum (1994) noted that self-definitions of sexual orientation are subject to regional and cohort differences. Lesbian participants in San Francisco might conceptualize their sexual orientation differently than those in Topeka, and participants who came out in the 1950s might conceptualize it differently than those coming out today. In addition, some people who think of themselves as gay or lesbian might not identify themselves as such because of the stigma associated with those labels or because of the fear of discrimination, violence, or other possible harms.
Reliance on self-labeling could be particularly limiting in studies of women, who might identify with a variety of labels other than "lesbian," even if they are involved in long-term, same-gender relationships (Denenberg, 1994). The writings of several authors (Cook, 1977; Rich, 1980; Rothblum & Brehony, 1993) exemplify a longstanding feminist discourse on considering same-gender sexual attraction only "one point on the lesbian continuum" (Golden, 1996, p. 230) of women who have intimate connections with other women. On the basis of interviews with over 100 women, Golden found that same-gender sexual attractions and involvements could not predict women's adoption of lesbian, bisexual, or heterosexual identities. For example, some adult women considered themselves heterosexual for many years before becoming aware of any attraction to other women. Others considered themselves lesbians only until deciding to have a relationship with a man. Still others reported choosing not to consider themselves lesbians because they thought it would make their lives more difficult. Golden charged that the essentialist conceptualization of sexual orientation is based primarily on studies of men conducted from a male perspective. Evidence that biological determinants of sexual orientation are more salient in men than women (Hu et al., 1995) may support this charge. Winkle (1997) cautioned that studies of women should include a variety of labels for participants to endorse to capture the most accurate self-definition for each participant. Labels should allow for all combinations of same-gender sexual desire, same-gender sexual or affectional behavior, and lesbian identity.
Not all men who engage in sex with other men identify themselves as gay either, especially married men (Hays & Samuels, 1989) and ethnic minority men (Greene, 1994; Peterson & Marin, 1988; Zamora-Hernandez & Patterson, 1996). For this reason, studies involving ethnic minority men often use a behavioral definition of the population of interest. For example, Ramirez, Suarez, de la Rosa, Castro, and Zimmerman (1994) simply asked participants whether they had ever had sex with another man. Ethnic differences such as these may exist because both gay and nongay male identifies are social constructions bound by culture and class, as explained by Foucault (1978) and Chauncey (1994). Today American men in majority culture are likely to identify themselves as gay if their primary objects of sexual and affectional interest are same-gender. However, working-class men in the early 20th century often engaged in same-gender sexual behavior without considering themselves "different," as long as their social and sexual behavior conformed to a masculine role. Before the concept of "homosexuality" became popularized, men considered themselves "different" only if they dressed or acted like women (Chauncey).
Defining the population solely by its sexual behavior introduces other problems. There are no commonly agreed-on standards for labeling such behavior. As the Kinsey et al. (1948) study illustrated, there could be various standards for the amount of same-gender sexual behavior needed to consider people gay or lesbian. Defining the population by its sexual behavior is also likely to eliminate people who identify themselves as gay or lesbian, although they have not had any same-gender sexual experience, especially youths and people that discover their lesbian or gay sexual orientation later in life.
In the absence of universally accepted definitions of "gay" and "lesbian," researchers should define the population they wish to sample in a way that fits the purpose of the study (Rothblum, 1994). Studies of people of color should consider using a behavioral definition or a definition that does not require participants to endorse a gay or lesbian label. Studies of youths should consider using a definition that does not require participants to have same-gender sexual experiences. Studies aiming for maximum inclusiveness should use a complex definition that takes into consideration sexual behavior, affectional preferences, and identity self-labeling, such as the one used by Laumann, Gagnon, Michael, and Michaels (1994). However, the existence of varying definitions of this population makes it difficult to compare findings across studies.
Measures of Sexual Orientation
Use of standardized measures to identify participants' sexual orientation might avoid some of the limitations associated with self-identification, and it also may increase the comparability of findings across studies. However, the reliability and validity of some of the measures described below are not well established. To ensure the most accurate definition of the population, studies can use a combination of self-definition questions and standardized measures.
The Multidimensional Scale of Sexuality (Berkey, Perelman-Hall, & Kurdek, 1990) is a 45-item, self-report questionnaire that measures cognitive, affective, behavioral, and dynamic (non-static) components of sexual orientation. It categorizes sexual orientation into nine groups: heterosexuality, homosexuality, asexuality, and six types of bisexuality (for example, concurrent or sequential). These categories are not mutually exclusive. Although the scale uses a true-or-false response format for most items, a single item asks participants to endorse one of nine possible descriptions (corresponding to the nine categories) of their sexual orientation. For each category, the sum of four items yields a cognitive-affective score, and a single item yields a behavioral score. Berkey et al. reported internal consistency (Cronbach's alpha) ranging from .63 to .87 for the nine sets of cognitive-affective items.
The Klein Sexual Orientation Grid (Klein, Sepekoff, & Wolf, 1985) is a 21-item self-report questionnaire that conceptualizes sexual orientation as a "dynamic multi-variable process." Participants rate themselves from 1 = other sex only to 7 = same sex only on seven dimensions of sexual orientation (sexual attraction, behavior, fantasies, emotional preference, social preference, self-identification, and lifestyle) for the current year, for previous years, and according to their ideals. Ideal ratings might indicate potential future changes in any of the dimensions of sexual orientation. Scores for each of these 21 categories are arranged on a seven-by-three grid. Klein, Sepekoff, and Wolf reported excellent internal consistency (Kuder-Richardson formula) for the entire grid based on data from 351 male and female participants and higher estimates of internal consistency for the time dimensions (past, present, and future) than for each of the seven dimensions of sexual orientation.
The Gay Identity Questionnaire (Brady & Busse, 1994) is a 45-item self-report measure that assesses the extent to which men have a gay identity, as conceptualized by the Cass six-stage model. It uses a categorical (true or false) response format. Three items verify that respondents have same-gender sexual or affectional thoughts, feelings, or behavior. The rest of the questionnaire consists of seven items in each of six dimensions, corresponding to the six stages proposed by the Cass model. The dimension with the highest point total indicates the current stage of identity development. Brady and Busse reported acceptable internal consistency (Kuder-Richardson formula) on three of the dimensions, ranging from .71 to .78, based on administration to 225 male participants. One dimension had low internal consistency (.44). Brady and Busse did not report internal consistency for the other three dimensions. Use of the Gay Identity Questionnaire with 196 male participants provided some support for the validity of the Cass model (Brady & Busse).
SAMPLING AND GENERALIZABILITY
Sampling methodology can present significant challenges for research on gay men and lesbians. Studies cannot obtain a true random sample of lesbians or gay men because there is no conceivable sampling frame for them, especially because an unknown percentage will not reveal their sexual orientation (Garnets & Kimmel, 1993). They often use purposive or convenience samples, drawn from members of gay and lesbian organizations (Moran, 1992; Nardi & Sherrod, 1994; Pilkington & D'Augelli, 1995; Proctor & Groze, 1994), readers of gay and lesbian publications (Deenen, Gijs, & van Naerssen, 1994; Hare, 1994; Kurdek, 1991), patrons of gay and lesbian business establishments (Folkman, Chesney, Pollack, & Phillips, 1992; Kelaher, Ross, Rohrsheim, Drury, & Clarkson, 1994), recipients of social or medical services that target gay men and lesbians (Fisher, Goldschmidt, Hays, & Catania, 1993; Paul, Stall, Crosby, Barrett, & Midanik, 1994; Siebt et al., 1995), or a combination of strategies (Esterberg, 1994; Martin & Knox, 1997c; McLean et al., 1994; Ridge, Plummer, & Minichiello, 1994; Siever, 1994).
Nonprobability samples, used in studies such as those identified above, call for significant limitations on the generalizability of findings to the population of all gay men and lesbians. Harry (1990) noted that studies of gay men that use nonprobability samples often underrepresent men who are not well-connected to the "social structures of the gay male world, e.g., bars, gay organizations, and friendship networks" (p. 90). Young, old, less-educated, rural, nonwhite, and heterosexually married gay men are among those who tend to be excluded. According to Stevens (1992), lesbians have always been relatively invisible, making it difficult to draw nonprobability samples that represent their diversity. Researchers must devote special effort to maximizing the diversity of nonprobability samples according to age, education, race, and ethnicity. Nonprobability samples also tend to incur the problem of self-selection, since participants who volunteer might have different motivations or characteristics than other members of this population. Nevertheless, Garnets and Kimmel (1993) asserted that studies that endeavor to challenge existing stereotypes about lesbians or gay men do "not require representative samples to be compelling" (p. 4). In other words, findings that challenge universal stereotypes about this population can be important even if they lack generalizability.
Laumann et al. (1994) and Harry (1990) advocated the use of probability sampling in research on this population. However, this sampling strategy has its own pitfalls, especially in assuming that gay men and lesbians are randomly distributed in the general population. With the use of a stratified, multistage cluster sample of the entire United States, Laumann et al. found the percentage of gay men and lesbians to be considerably higher in urban settings and lower in rural settings. Studies using probability samples that do not account for this uneven distribution risk underrepresenting or overrepresenting the general gay and lesbian population.
PREVENTING HARM TO PARTICIPANTS
The NASW Code of Ethics (1996) directs social workers engaged in research activities to be mindful of the effects of these activities on participants; to ensure participants' privacy, dignity, and self-determination; and to make sure that participants are not harmed in any way. These standards are particularly relevant when the research participants are lesbians and gay men.
According to Walsh-Bowers and Parlour (1992), research on minority groups "is necessarily an ethical and political intervention with participants" (p. 109). Unless researchers take specific steps to prevent harm to participants and their environments, they should not assume that their study is benign. Walsh-Bowers and Parlour voiced particular concern about hierarchical and parasitic relationships in which researchers use and then discard members of gay and lesbian communities. For example, if a study of health risks among lesbians were to leave participants feeling used or demeaned, it might add to existing community mistrust of the scientific--medical establishment (Winkle, 1997). Such mistrust might make it less likely that community members will seek preventive or interventive health care services. The strategies of significant involvement and functional relevance (Bowman, 1983) can be used to ensure ethical standards in research on this population. Research reports should describe the nature of researcher--participant relationships and the way in which investigators entered and exited from the research setting.
Many gay men and lesbians may feel uncomfortable about participating in research, especially when investigators are external to their community. For many of them, the danger of discrimination, violence, and other serious harms resulting from breaches of privacy and confidentiality cannot be overstated. Because a person's sexuality or sexual behavior is extremely sensitive information, some studies might need to take extraordinary measures to ensure the protection of privacy and confidentiality or anonymity (Boruch, 1989).
Lesbian and gay participants in cross--sectional research should remain anonymous, because confidential participation may provide insufficient assurance. However, this strategy prevents researchers from knowing the number and characteristics of potential participants who self-select out of the study. Alternatively, a trusted member of the community could function as custodian of the participant list. As intermediary, this person would maintain records of those who select in and out of the study. In longitudinal research it is common to ensure privacy and confidentiality by assigning aliases to each participant. Researchers must also minimize the potential for deductive disclosures. Postmarks, responses to demographic questions, and codes used to identify distribution sites can allow for the identification of individual participants when the sample is small or lacking in age, gender, or ethnic diversity. Gay or lesbian participants may be particularly suspicious of personal interviews, which may provide less protection of their privacy and anonymity. Several strategies can be used to ensure the confidentiality of information gained from interviews, including randomized response tactics and other complex statistical methods (Boruch, 1989).
Practitioners who evaluate their interventions with lesbian or gay clients must take similar steps not to violate ethical research standards. Nelsen (1994) recommended that practitioners using single-case practice research obtain informed consent for all nonroutine aspects of their practice, such as data collection, third-party monitoring, or delay of treatment during a baseline period. Practitioner--researchers should make full disclosure to clients about the reason for using evaluation methodology, its impact on the treatment itself, and all ways in which the information will be used. Agencies routinely using single-case research in practice should have review panels to ensure adherence to ethical standards.
Researchers who are themselves lesbian or gay might have an advantage when recruiting lesbian or gay research participants, but they are not immune to potential ethical violations. For example, Woodman, Tully, and Barranti (1995) cautioned lesbian researchers about potential ethical dilemmas involving privacy, confidentiality, and dual--role relationships when participants are other lesbians living in the same community. In these situations researchers are likely to encounter participants in social situations. Researchers might also be participants' friends, professors, or therapists. Participants might pressure researchers to divulge confidential information. When reading research reports, participants might be able to identify other participants even after researchers made efforts protect their anonymity.
Because gay men and lesbians are stigmatized in U.S. society, researchers must take special precautions to ensure that studies do not exploit them and that study outcomes provide them with some benefit. To reduce the possibility of exploitation of racial or ethnic minority groups by research, Bowman (1983) called for the significant involvement of members of these groups in the planning of studies that focus on them. For instance, community members might be used as consultants or liaisons in the design of measurement tools to make sure that language is appropriate. They might also assist with the selection of the sample and ease acceptance of the project by the community. In addition, they might be employed as interviewers or disseminators of self-report surveys.
These recommendations are extremely relevant to studies of gay men and lesbians. Members of gay and lesbian communities are likely to be suspicious of outsiders seeking to study them and especially their sexuality (Catania, McDermott, & Pollack, 1986; Silvestre, 1994). Through the significant involvement of community members as consultants or liaisons, both researchers and the community may benefit. For example, researchers involved members of a local gay and lesbian social services agency in every step of the planning of a study of HIV risk behavior among gay men (Martin & Knox, 1997a, 1997b, 1997c). Members of the HIV education department contributed to the development of the survey instrument by contributing items, estimating the relevance of items to the community, and evaluating the appropriateness of item wording. These efforts may have improved the instrument's content validity and may have helped to reduce the potential for response bias. The members helped develop the sampling strategy by identifying ways to reach potential participants and by stimulating community interest in the study. They also helped to disseminate the questionnaire at a variety of sites in the community. Gay male employees of the agency pilot-tested the questionnaire, providing valuable information regarding its format and instructions and the face validity of its components. Researchers incorporated their criticisms to shape the final version of the questionnaire. They debriefed agency staff about the study's findings and made sure they received copies of publications produced from the study.
Bowman (1983) also encouraged researchers to ensure that studies of racial or ethnic minority populations are functionally relevant to minority communities. In other words, research that uses minority group members as participants must respond to the needs of minority communities. People in communities experiencing widespread poverty, hunger, and violence may not care about research that ignores these pressing problems. Similarly, members of gay and lesbian communities face discrimination, marginalization, and violence (Bryant & Demian, 1994; Ettelbrick, 1996; Herek, 1993; Herek, Gillis, Cogan, & Glunt, 1997). Researchers should be able to demonstrate how the study will lead to some benefit for participants or their community. Martin and Knox (1997a, 1997b, 1997c) used the questionnaire cover sheet to explain the relevance and potential benefit of their study to the local gay and lesbian community and its members.
Silvestre (1994) suggested a brokering model for relationships between researchers and gay male communities. According to this model researchers and community leaders exchange a variety of goods and services for their mutual long-term benefit. In the case of a longitudinal epidemiological study of HIV in gay men, researcher contributions included access to expert information about HIV, university facilities for meetings and events, and donations to community charities. In exchange, community leader contributions included expressions of public support for the project, distribution of information about it, and advice regarding research protocols. Researchers used the advice of established leaders in filling the position of community organizer. This person's primary responsibilities were to establish a community advisory board and to broker the relationship between the community and the researchers. Thus the model provided a concrete structure for community organization and development while it helped to ensure the success of the research project.
Renzetti (1995) suggested using a feminist research model for studies of lesbians. The ultimate purpose of research conducted according to this model is to improve participants' life conditions. Researchers using this model engage in highly collaborative relationships with participants, rejecting the objective, value-free stance of positivist research. Participants might help to determine the research questions and methods for measuring them, as well as the analysis, interpretation, and use of collected data. Researchers might disclose as much as participants, and data are often both qualitative and quantitative.
Social work programs require research-based knowledge about lesbians and gay men to educate students about their characteristics and needs, their institutions, and their communities. Without research-based knowledge, students are likely to gain little more than a superficial understanding of the needs of this population, resulting in limited abilities to provide its members with effective services. The number of studies on lesbians and gay men is increasing rapidly, and we hope that it will continue to do so. Researchers who choose to study lesbians and gay men provide an important service to social work education and practice by expanding knowledge about this population. They also provide a foundation on which students interested in research on this population can build. However, lesbians and gay men should also benefit from these efforts. One way to maximize the potential to benefit is through implementation of significant involvement of lesbians and gay men in study development to ensure functional relevance (Bowman, 1983). Researchers should also formulate research questions that address real problems experienced by members of this population, rely on theory that is free of heterosexist bias, use methodology that takes into account the complexity of gay and lesbian identities and communities, and exercise diligence in protecting against breaches of confidentiality and other potential harms. An expanding network of researchers committed to these efforts will help forge partnerships between the social work profession and gay and lesbian communities that have great potential for mutual benefit.
Original manuscript received September 11, 1998
Final revision received April 1, 1999
Accepted May 3, 1999
Berkey, B. R., Perelman-Hall, T., & Kurdek, L. A. (1990). Journal of Homosexuality, 19, 67-87.
Berkman, C. S., & Zinberg, G. (1997). Homophobia and heterosexism in social workers. Social Work, 42, 319-332.
Bieber, I. (1976). Homosexuality: The ethical challenge. Journal of Consulting and Clinical Psychology, 44, 163-166.
Bieber, I., Dain, J. J., Dince, P.R., Drellich, M. W., Rifkin, A. H., Wilbur, C. B., & Bieber, T. B. (1962). Homosexuality: A psychoanalytic study. New York: Basic Books.
Boruch, R. F. (1989). Resolving privacy problems in AIDS research: A primer. In L. Sechrest, H. Freeman, & A. Mulley (Eds.), Health services and research methodology: A focus on AIDS. (DHHS Publication No. PHS 89-3439, pp. 165-180). Rockville, MD: U.S. Government Printing Office.
Bowman, P. J. (1983). Significant involvment and functional relevance: Challenges to survey research. Social Work Research and Abstracts, 19(4), 21-26.
Brady, S. & Busse, W. J. (1994). The Gay Identity Questionnaire: A brief measure of homosexual identity formation. Journal of Homosexuality, 26, 1-22.
Bryant, A. S., & Demian. (1994). Relationship characteristics of American gay and lesbian couples: Findings from a national survey. Journal of Gay & Lesbian Social Services, 1(2), 101-117.
Cain, R. (1996). Heterosexism and self-disclosure in the social work classroom. Journal of Social Work Education, 21, 65-76.
Catania, J., McDermott, L., & Pollack, L. (1986). Questionnaire response bias and face-to-face interview sample bias in sexuality research. Journal of Sex Research, 22, 52-72.
Chauncey, G. (1994). Gay New York: Gender, urban culture, and the making of the gay male world, 1890-1940. New York: Basic Books.
Cook, B. W. (1977). Female support networks and political activism. Chrysalis, 3, 43-61.
Council on Social Work Education. (1994). Curriculum policy statement for master's degree programs in social work education [Online]. Available: http://www.cswe.org/mswcps.htm.
Deenen, A. A., Gijs, L., & van Naerssen, A. X. (1994). Intimacy and sexuality in gay male couples. Archives of Sexual Behavior, 23, 421-431.
Denenberg, R. (1994). Report on lesbian health. Washington, DC: National Gay and Lesbian Task Force Policy Institute.
Erikson, E. (1980). Identity and the life cycle. New York: W. W. Norton.
Esterberg, K. (1994). Being lesbian and being in love: Constructing identity through relationships. Journal of Gay and Lesbian Social Services, 1(2), 57-82.
Ettelbrick, P. L. (1996). Legal issues in health care for lesbians and gay men. Journal of Gay & Lesbian Social Services, 5(1), 93-109.
Fisher, L., Goldschmidt, R. H., Hays, R. B., & Catania, J. A. (1993). Families of homosexual men: Their knowledge and support regarding sexual orientation and HIV disease. Journal of the American Board of Family Practice, 6, 25-32.
Folkman, S., Chesney, Pollack, L, & Phillips, C. (1992). Stress, coping, and high-risk sexual behavior. Health Psychology, 11, 218-222.
Foucault, M. (1978). The history of sexuality. New York: Pantheon.
Garnets, L. D., & Kimmel, D. C. (1993). Introduction: Lesbian and gay male dimensions in the psychological study of human diversity. In L. D. Garnets & D. C. Kimmel (Eds.), Psychological perspectives on lesbian and gay male experiences (pp. 1-51). New York: Columbia University Press.
Golden, C. (1996). What's in a name? Sexual self-identification among women. In R. C. Savin-Williams & K. M. Cohen (Eds.), The lives of lesbians, gays, and bisexuals: Children to adults (pp. 229-249). Ft. Worth, TX: Harcourt Brace.
Greene, B. (1994). Ethnic-minority lesbians and gay men: Mental health and treatment issues. Journal of Consulting and Clinical Psychology, 62, 243-251.
Hare, J. (1994). Concerns and issues faced by families headed by a lesbian couple. Families in Society, 75, 27-35.
Harry, J. (1990). A probability sample of gay males. Journal of Homosexuality, 19(1), 89-104.
Haslam, N. (1997). Evidence that male sexual orientation is a matter of degree. Journal of Personality and Social Psychology, 73, 862-870.
Hays, D., & Samuels, A. (1989). Heterosexual women's perceptions of their marriages to bisexual or homosexual men. Journal of Homosexuality, 18, 81-100.
Herek, G. M. (1993). Documenting prejudice against lesbians and gay men on campus: The Yale Sexual Orientation Survey. Journal of Homosexuality, 25, 15-30.
Herek, G. M., Gillis, J. R., Cogan, J. C., & Glunt, E. K. (1997). Hate crime victimization among lesbian, gay, and bisexual adults: Prevalence, psychological correlates, and methodological issues. Journal of Interpersonal Violence, 12, 195-215.
Herek, G. M., Kimmel, D. C., Amaro, H., & Melton, G. B. (1991). Avoiding heterosexist bias in psychological research. American Psychologist, 46, 957-963.
Hu, S., Pattatucci, A. M., Patterson, C., Li, L., Fulker, D. W., Cherny, S. S., Kuglyak, L., & Hamer, D. (1995). Linkage between sexual orientation and chromosome Xq28 in males but not females. Nature Genetics, 11, 248-256.
Hunter, S., Shannon, C., Knox, J., & Martin, J. I. (1998). Lesbian, gay, and bisexual youths and adults: Knowledge for human services practice. Thousand Oaks, CA: Sage Publications.
Kagel, J. D., & Cowger, C. D. (1984). Blaming the client: Implicit agenda in practice research? Social Work, 29, 347-351.
Kelaher, M., Ross, M. W., Rohrsheim, R., Drury, M., & Clarkson, A. (1994). Dominant situational determinants of sexual risk behaviour in gay men. AIDS, 8, 101-105.
Kinsey, A. C., Pomeroy, W. B., & Martin, C. E. (1948). Sexual behavior in the human male. Philadelphia: W. B. Saunders.
Kitzinger, C. (1987). The social construction of lesbianism. Beverly Hills, CA: Sage Publications.
Klein, F., Sepekoff, B., & Wolf, T. J. (1985). Sexual orientation: A multi-variable dynamic process. Journal of Homosexuality, 11(1-2), 35-49.
Kronemeyer, R. (1980). Overcoming homosexuality. New York: Macmillan.
Kurdek, L. A. (1991). Correlates of relationship satisfaction in cohabitating gay and lesbian couples: Integration of contextual investment, and problem-solving models. Journal of Personality and Social Psychology, 61, 910-922.
Laumann, E. O., Gagnon, J. H., Michael, R. T., & Michaels, S. (1994). The social organization of sexuality. Chicago: University of Chicago Press.
Mackelprang, R. W., Ray, J. A., & Hernandez-Peck, M. (1996). Social work education and sexual orientation: Faculty, student, and curriculum issues. Journal of Gay & Lesbian Social Services, 5, 17-31.
Mallon, G. P. (1997). Knowledge for practice with gay and lesbian persons. In G. P. Mallon (Ed.), Foundations of social work practice with lesbian and gay persons (pp. 1-30). Binghamton, NY: Harrington Park Press.
Martin, J. I. (1997). Political aspects of mental health treatment. In T. R. Watkins & J. W. Callicutt (Eds.), Mental health policy and practice today (pp. 32-48). Thousand Oaks, CA: Sage Publications.
Martin, J. I., & Knox, J. (1997a). Loneliness and sexual risk behavior in gay men. Psychological Reports, 81, 815-825.
Martin, J. I., & Knox, J. (1997b). The surprising role of intimacy in HIV prevention. National Social Science Perspectives Journal: Proceedings of the April 1997 National Social Science Association Conference, 11, 126-130.
Martin, J. I., & Knox, J. (1997c). Self-esteem instability and its implications for HIV prevention among gay men. Health & Social Work, 22, 264-273.
McLean, J., Boulton, M., Brookes, M., Lakhani, D., Fitzpatrick, R., Dawson, J., McKechnie, R., & Hart, G. (1994). Regular partners and risky behaviour: Why do gay men have unprotected intercourse? AIDS Care, 6, 331-341.
Messinger, L., & Topal, M. (1997). Are you married? Two sexual-minority students' perspectives on field placement. Affilia, 12, 106-113.
Moran, M. R. (1992). Effects of sexual orientation similarity and counselor experience level of gay men's and lesbians' perceptions of counselors. Journal of Counseling Psychology, 39, 247-251.
Morin, S. F. (1977). Heterosexual bias in psychological research on lesbianism and male homosexuality. American Psychologist, 32, 629-637.
Morrow, D. F. (1996). Heterosexism: Hidden dimension in social work education. Journal of Gay & Lesbian Social Services, 5, 1-16.
Nardi, P. M., & Sherrod, D. (1994). Friendship in the lives of gay men and lesbians. Journal of Social and Personal Relationships, 11, 185-199.
NASW Code of Ethics. (1996, December). NASW News. S1-S4. Nelsen, J. C. (1994). Ethics, gender, and ethnicity in single-case research and evaluation. Journal of Social Service Research, 18, 139-152.
Nicolosi, J. (1991). Reparative therapy of male homosexuality. Northvale, NJ: Jason Aronson.
Paul, J. P., Stall, R. D., Crosby, G. M., Barrett, D. C., & Midanik, L. T. (1994). Correlates of sexual risk-taking among gay male substance abusers. Addiction, 89, 971-983.
Peters, D. K., & Cantrell, P. J. (1991). Factors distinguishing samples of lesbian and heterosexual women. Journal of Homosexuality, 21, 1-15.
Peterson, J., & Marin, G. (1988). Issues in the prevention of AIDS among black and Hispanic men. American Psychologist, 43, 871-877.
Pilkington, N. W., & D'Augelli, A. R. (1995). Victimization of lesbian, gay, and bisexual youth in community settings. Journal of Community Psychology, 23, 34-56.
Proctor, C. D., & Groze, V. K. (1994). Risk factors for suicide among gay, lesbian, and bisexual youths. Social Work, 39, 504-513.
Ramirez, J., Suarez, E., de la Rosa, G', Castro, M. A., & Zimmerman, M. A. (1994). AIDS knowledge and sexual behavior among Mexican gay and bisexual men. AIDS Education and Prevention, 6, 163-174.
Renzetti, C. M. (1995). Studying partner abuse in lesbian relationships: A case for the feminist participatory research model. Journal of Gay & Lesbian Social Services, 3, 29-42.
Rich, A. (1980). Compulsory heterosexuality and lesbian existence. Signs, 5, 631-660.
Ridge, D. T., Plummer, D. C., & Minichiello, V. (1994). Young gay men and HIV: Running the risk? AIDS Care, 6, 371-378.
Ripple, L. (1960). Problem identification and formulation. In N. Polansky (Ed.), Social work research (pp. 24-47). Chicago: University of Chicago Press.
Rothblum, E. D. (1994). "I only read about myself on bathroom walls": The need for research on mental health of lesbians and gay men. Journal of Consulting and Clinical Psychology, 62, 213-220.
Rothblum E. D., & Brehony, K. (1993). Boston marriages: Romantic but asexual relationships among contemporary lesbians. Amherst, MA: University of Massachusetts Press.
Siebt, A. C., Ross, M. W., Freeman, A., Krepcho, M., Hedrich, A., McAlister, A., & Fernandez-Esquer, M. E. (1995). Relationship between safe sex and acculturation into the gay subculture. AIDS Care, 7, S85-S88.
Siever, M. D. (1994). Sexual orientation and gender as factors in socioculturally acquired vulnerability to body dissatisfaction and eating disorders. Journal of Consulting and Clinical Psychology, 62, 252-260.
Silvestre, A. J. (1994). Brokering: A process for establishing long-term and stable links with gay male communities for research and public health education. AIDS Education and Prevention, 6, 65-73.
Socarides, C. W. (1978). Homosexuality. New York: Jason Aronson.
Stevens, P. E. (1992). Lesbian health care research: A review of the literature from 1970 to 1990. Health Care for Women International, 13(2), 91-120.
Strean, H. S. (1996). Psychoanalytic theory and social work treatment. In E J. Turner (Ed.), Social work treatment (4th ed.) (pp. 523-554). New York: Free Press.
Walsh-Bowers, R. T., & Parlour, S. J. (1992). Researcher-participant relationships in journal reports on gay men and lesbian women. Journal of Homosexuality, 23(4), 93-112.
Winkle, C. L. (1997). A health and mental health needs assessment of lesbians in the Dallas metropolitan area. Unpublished master's thesis, University of Texas at Arlington.
Wisniewsky, J. J., & Toomey, B. G. (1987). Are social workers homophobic? Social Work, 32, 454-455.
Woodman, N. J., Tully, C. T., & Barranti, C. C. (1995). Research in lesbian communities: Ethical dilemmas. Journal of Gay & Lesbian Social Services, 3, 57-66.
Zamora-Hernandez, C. E., & Patterson, D. G. (1996). Homosexually active Latino men: Issues for social work practice. Journal of Gay & Lesbian Social Services, 5(2-3), 69-91.
James L Martin, PhD, ACSW
Associate Professor, Ehrenzkranz School of Social Work
New York University
Jo Knox, PhD, ACSW
School of Social Work
University of Alabama, Tuscaloosa