Abstract
An important original study by Dorland and Fischer noted how the use of inclusive language can affect the therapeutic relationship positively for gay, lesbian, and bisexual clients. In this extension of that study with heterosexual participants (N = 179), there seemed to be low, but positive, salience of the language used by the therapist. These participants showed negligible, but positive, sensitization to issues related to inclusiveness, so the ethical practice of inclusive language use seems to have little downside (and potentially a large positive impact).
The broad areas in which we counsel, train, and do research are affected by GLB issues, whether or not we acknowledge them.
GLB clients access mental health services at higher rates than heterosexual clients (Burgess, Tran, Lee, & van Ryn, 2007). These higher rates of use are likely the result of the oppressive forces associated with the experience of having a stigmatized identity (Ritter & Terndrup, 2002). As a result of higher usage rates, most therapists interact with GLB clients. For example, Green and Bobele (1994) studied members of the American Association of Marriage and Family Therapy and found that 80% of respondents reported working with gay and lesbian clients. Because of the high rates of mental health care that GLB clients access, it is integral that counselors work toward building cultural competence with regard to GLB issues. This competence includes language use because “some of our most powerful therapeutic instruments are words” (Edelstein & Waehler, 2011, p. 17). Language that is inclusive, inviting, and nonjudgmental can increase client openness to therapy and their own concerns, whereas words that are discriminating and condescending will not promote therapeutic goals.
It is widely known that gay, lesbian, and bisexual (GLB) status is highly stigmatized (American Psychological Association, 2012). Researchers understand that on a macro level, heterosexist language use can lead to internalized stigma in GLB individuals (Herek, Gillis, & Coogan, 2009). Language use in therapy has the potential to strengthen therapeutic alliance; however, heterosexist language in this micro environment may also lead to increased levels of stigma experienced by GLB clients. In an important study of language in the counseling environment, Dorland and Fischer (2001) studied self-identified GLB individuals’ (N = 126) perceptions of counselors based on the language choice of the counselor. The authors randomly provided participants with two variations of a first counseling session. In one vignette, the counselor used heterosexist language (i.e., husband or wife, marriage). In the other vignette, the counselor used inclusive language (i.e., partner, separation). The authors then used the Counselor Effectiveness Rating Scale (CERS) and the Self-Disclosure Scale (SDS) to determine if there were different responses by participants based on the language variations.
Results indicated that the participants in the heterosexist condition rated the counselor as less credible, were less willing to disclose to the counselor, were less willing to see the counselor for a second session, and were less comfortable disclosing their sexual orientation to the counselor, even after controlling for participant outness levels. The effect sizes were largest with regard to whether or not participants would return for counseling (.28) and whether or not the participants would disclose their sexual orientations (.23). This study demonstrated that language use is linked to significant implications for GLB clients’ perceptions of counselors and their willingness to come to counseling and be open about their identities. If clients feel discomfort discussing their sexual orientation with their counselor, then there could be harmful consequences for the clients, such as leading them to avoid needed health services and perpetuating or possibly magnifying the clients’ experience of internalized heterosexism.
We consider this line of research to be consistent with other work investigating therapist’s multicultural orientation (MCO), which examines clients’ observations of their therapists’ “way of being” with the client (Owen, Tao, Leach, & Rodolfa, 2011). Psychotherapists’ MCO is thought to influence the therapeutic relationship, which in turn can affect clients’ psychological well-being. In their study, Owen et al. (2011) found that “clients’ perceptions of their psychotherapists’ MCO were positively associated with their psychological well-being” (p. 280). Therefore, actions that promote clients’ perceptions of MCO, such as the use of nonheterosexist language, may assist in promoting client psychological health and well-being.
Although counselor language use with regard to GLB issues plays a role with these clients, little is known about the effect of GLB-related issues on heterosexual clients. For example, Bieschke (2008) points out that internalized heterosexism, which powerfully affects GLB individuals, may affect heterosexual clients as well. Furthermore, Bieschke notes that heterosexism contributes to a systemic limitation of self-expression, which is costly for all individuals, not just members of the GLB community. In addition, because sexual orientation is an invisible identity that continues to remain highly stigmatized, it is imperative for counselors to offer a welcoming environment for all clients, so as to not unwittingly block client openness and further extend social trends that discriminate. The implications are that heterosexism and its manifestation in language use may be important therapeutic factors for both GLB and heterosexual clients.
Dorland and Fischer’s (2001) study, illustrating just how powerful language can be, is consistent with Wachtel’s (1993) observation that in the counseling relationship we focus on every utterance and nuance of the client’s language, but little attention is paid to the language of the counselor. As Wachtel noted, “The particular way one says something has a powerful impact” (p. 5). Edelstein and Waehler (2011) suggest that style and language combine powerfully when therapists speak, stating, “Clients listen closely to your language and remember what you say to them” (p. 17). Robertson and Fitzgerald (1992) demonstrated empirically the impact of language as they found that changing the way a counseling service is described (i.e., as a workshop as opposed to counseling) would lead some people to engage in psychological services more readily. Atkinson and Carskaddon (1975) studied the effects of a prestigious introduction and the counselor’s use of abstract psychological jargon on clients’ perceptions of counselor credibility and found that clients perceive counselors who use psychological jargon as more credible and clients were more willing to see a counselor if the counselor was given a prestigious introduction. In a study of bilingual Mexican American college students, Ramos-Sanchez (2007) explored the effect of counselor bilingualism and counselor ethnicity on client emotional expression. The author found that the European American (as opposed to Mexican American) counselors who used bilingualism elicited more emotional responses from the participants than did counselors who solely used English, regardless of counselor ethnicity. This result implies that clients are aware of differences in counselor language and that language use plays a powerful role in therapy. The results of these studies demonstrate that the language a counselor uses has an impact on counselor credibility and client reactions to counselors and, therefore, has an impact on treatment outcomes. Language can be used to strengthen therapeutic alliance, which has been shown in the literature to be an important factor in the outcome of therapy (Bender, 2005).
Not all results give credence to the concept that language in the counseling environment is powerful. Arokiasamy and Strohmer (1994) found that the use of politically correct versus non–politically correct language did not have a significant effect on clients’ ratings of counselor credibility when they looked at this variable with disabled and nondisabled participants. Counselor credibility ratings did not vary based on politically correct, incorrect, or ultracorrect language, and in fact counselor skill level (good or poor attending skills) had a significant effect on the dependant variables. The authors attribute the lack of significant effects based on language to previously cited data showing that clients with disabilities tend to discount insensitivity to their disability among therapists. In a study that tested bilingual ability, counselor bilingualism clues, and counselor ethnicity on Mexican American college students’ perceptions of counselor credibility, Ramos-Sanchez, Atkinson, and Fraga (1999) found that there were no significant differences on counselor credibility ratings whether bilingualism was used or not. Results from this study did not support the hypothesis that counselor language affected client perceptions of counselor credibility. In their discussion, the study’s authors reported that levels of acculturation may have influenced participants’ ratings of counselor credibility and acculturation may have played a larger role than language use in this particular study.
Although results have yielded mixed results regarding counselor language use and perceptions of counselor credibility, there are currently few contemporary studies that investigate the effects of language on credibility. Dorland and Fischer (2001) found clear evidence of an impact of language use on counselor credibility for GLB participants, making the use of inclusive language important with these individuals. Extending the Dorland and Fischer study to heterosexual participants would help explore how counselors’ communication strategies with clients of various backgrounds may increase counselor credibility, client willingness to disclose, and utilization intent. For example, if participants prefer a counselor using inclusive language or if there are no significant differences between perceptions of either counselor, then counselors’ use of inclusive language with all clients, regardless of the clients’ reported sexual orientation, is supported further. A practical and ethical goal for all mental health professionals ought to be creating an open and welcoming environment for GLB and heterosexual clients alike. We hypothesized that heterosexual participants in this study who read the vignette with the counselor who uses inclusive language would rate that counselor as more credible, would indicate that they would be more willing to see that counselor, and would be more willing to disclose personal information to that counselor than the participants in the group with the heterosexist counselor. Conversely, indications of no difference between the two conditions would also be a result that would support counselor use of inclusive language.
Method
Participants
Participants were recruited from psychology classes at a midsized Midwestern university. After 2 participants with missing data and 13 individuals who identified as GLB were removed, there were a total of 179 (72 men and 107 women) self-identified heterosexual individuals. This sample size yielded a power above .80 (t test) and .95 (correlations) with an assumed moderate effect size and an alpha level set at .05. In terms of level of education, 41.8% of participants were freshmen, 16.8% sophomores, 20.7% juniors, 19.6% seniors, and 1.1% graduate students. Ages of participants ranged from 18 to 61 (M = 21.46, SD = 5.47) with the greatest concentration of ages being 18 (14.0%), 19 (31.8%), 20 (12.3%), and 21 (15.6%). Most participants were European American (83.8%); the second largest percentage reported race was Black or African American, with 11.2%.
Procedures
A convenience sample of college students was used in the current study to increase sample size and because this was an initial study in this area. Participants were offered extra credit for their participation in the study. Participants were told that the study explored language perceptions in counseling. They were then provided with questionnaire packets that requested that they read the informed consent and mark the form to indicate consent to participate, complete a demographic form, read the counseling vignette, complete subsequent dependent measures, complete a manipulation check, and read the debriefing form. Participants were provided with a vignette in which the client’s gender matched the participant’s gender; the vignette type (heterosexist vs. inclusive) was randomly assigned to each participant.
Counseling Vignettes
We used (with permission) the original vignettes from Dorland and Fischer’s (2001) study. When developing the vignettes, Dorland and Fischer created two variations of an initial counseling session that were identical except for key word changes with regard to heterosexist language. For example, in the heterosexist vignette, the counselor asks if the client is “going through a divorce,” and in the inclusive vignette the counselor asks if the client is “going through a separation.” The vignette contains 10 client responses and 11 counselor responses. When creating the vignettes, the authors consulted with six individuals who had expertise in GLB counseling issues. These individuals included licensed psychologists, paraprofessionals, and counseling psychology graduate students who had academic knowledge, publications, and/or clinical experience with regard to GLB counseling issues. Half of the consultants read the heterosexist vignette, and the other half read the inclusive vignette. After making minor adjustments, the authors had a second team of three different reviewers read the vignettes, and no revisions were requested. In the final version of the vignette, there were five instances in the counselor’s dialogue in which the language was altered between the two conditions.
Manipulation Check
As a final procedural activity, participants were asked two questions regarding the content of the sessions to ascertain whether they were paying attention to the material. Each participant was asked two questions regarding the content of the vignettes and answered at least one question correctly (96.1% and 92.2%, respectively) so all data were retained for the final analysis. In addition, the participants were asked to respond to the statement “I felt that the counselor exhibited heterosexist bias in the vignette” on a 5-point scale from 1 (strongly disagree) to 5 (strongly agree). Responses from participants in the inclusive language condition averaged 2.35 (SD = 0.93), which differed significantly, t(177) = 3.74, p < .001, Cohen’s d = 0.56, from the participants who read the heterosexist condition (M = 2.91, SD = 1.06), indicating that participants were aware of the language usage in the different conditions.
Dependent Measures
Counselor credibility
The CERS (Atkinson & Wampold, 1982), which was also used in Dorland and Fischer (2001), consists of 10 items. There are 3 items each for counselor expertness (expertness, competence, and skill), trustworthiness (sincerity, reliability, and trustworthiness), and attractiveness (friendliness, approachability, and likability). The measure has one item related to counselor utilization intent. A 7-point Likert-type scale with a range from 1 (good) to 7 (bad) was used, and half of the items were reverse scored. The CERS has been found to correlate highly (Atkinson & Wampold, 1982) with the Counselor Rating Form (LaCrosse & Barak, 1976), a measure of counselor credibility that has been established to have high predictive ability (LaCrosse, 1980). The previous study’s internal consistency of the nine credibility items was α = .92 (current study α = .87).
Utilization intent
The utilization intent item from the CERS (Atkinson & Wampold, 1982) was used to measure participant perceived willingness to see the counselor for counseling. The question “The counselor is someone I am willing to see for counseling in the future” was used with an additional item that was developed by Akutsu et al. (1990). The second question assessed if the participant would return for a second hypothetical session (“If I were a client, I would come back to see the counselor for another session”). Both items used a 7-point Likert-type scale (higher scores indicated greater willingness to see the counselor). Akutsu et al. (1990) and Dorland and Fischer (2001) found that these two items were correlated highly (r = .73 and r = .88, respectively; r = .83 in the current study), and therefore Dorland and Fischer averaged them into one composite measure of utilization intent. Higher scores indicated greater utilization intent (current sample, α = .90). No additional validity data were available for this measure.
Willingness to disclose
The SDS (Plasky & Lorion, 1984) was adapted from the Self-Disclosure Inventory (Jourard & Lasakow, 1958) and the Self-Disclosure Questionnaire (Vondracek & Marshall, 1971). The SDS consists of six categories: disclosure about sex, emotions, family, relationships, work, and morality. There are seven items for each category. Items are scored 1 (yes) or 0 (no) as participants indicate their comfort with regard to disclosing to the counselor. Dorland and Fischer (2001) reported an alpha of .95 (current study α = .91).
Results
Participants were randomly assigned to condition, and there were no statistical differences between the demographic variables (age, education level, gender, or race/ethnicity) of the two groups. There were also no mean differences in the outcome measures based on these demographic variables, so all data were collapsed into the comparison groups based solely on language use condition. Correlations between the condition (inclusive vs. heterosexist language) and counselor credibility (r = –.08, p = .32) and utilization intent (r = –.05, p = .49) were nonsignificant; willingness to disclose was significantly different (r = –.17, p = .03), wherein participants in the inclusive language condition were more willing to disclose than participants in the heterosexist condition. As would be expected, counselor credibility and utilization intent were correlated significantly (r = .76, p < .001), as were counselor credibility and willingness to self-disclose (r = .32, p < .001) and utilization intent and willingness to self-disclose (r = .26, p = .001). These latter three results lend credibility to the scores reported.
The t tests comparing the means of the dependent measures by condition followed a pattern similar to the correlations (see Table 1). Counselor credibility ratings were not significantly different between the inclusive and heterosexist conditions, t(177) = 1.01, p = .32, Cohen’s d = 0.15. Utilization intent was also not significantly different between the two conditions, t(177) = 0.69, p = .34, Cohen’s d = 0.10. Willingness to disclose showed a significant difference between the two conditions, t(177) = 2.24, p = .03, Cohen’s d = 0.34, with a greater willingness to disclose reported by the participants in the inclusive language condition.
Heterosexist Versus Inclusive Language: Comparison of Outcome Variables.
Discussion
We extended Dorland and Fischer’s (2001) original quasi-experimental study that explored GLB individuals’ perceptions of counselor credibility, their utilization intent, and their willingness to disclose based on the heterosexist and inclusive conditions being practiced. Their study demonstrated that language use mattered greatly to the GLB people in their study. We expanded the original study to test whether their script of a counselor using either inclusive language or heterosexist language would alter heterosexual individuals’ perceptions of counselor credibility, their utilization intent, and their willingness to self-disclose in therapy.
Hypothesis testing indicated that there were no significant differences between the use of heterosexist or inclusive language for two of the three outcome variables—counselor credibility and utilization intent. The third hypothesis, that “willingness to disclose” would be increased for the inclusive language group, was supported. These findings confirm what might be considered common sense: There seemed to be low salience of the language used for individuals who may not be particularly sensitized to issues related to inclusiveness, and where there was a difference, it was in the direction that favored the use of inclusive language. In fact, all three of the participants’ outcome measures were correlated with their ratings of the therapist’s use of inclusive language regardless of condition (counselor credibility r = .32, p < .001; utilization intent r = .29, p < .001; willingness to disclose r = .18, p = .025). Certainly causation direction is an issue with these correlations, but we are encouraged to see participants at least associate the use of inclusive language with positive endorsements of the counselor and therapy. Our conclusion is similar to the observation of Owen et al. (2011), who noted that their findings expanded on previous studies examining the positive relationship between clients’ ratings of their therapists’ MCO and satisfaction with psychotherapy.
The original Dorland and Fischer (2001) study was conceived based on the hypothesis that if a therapist is unaware that a client identifies as GLB, use of noninclusive language will negatively affect the likelihood of that individual feeling comfortable, feeling validated, and being willing to engage in a therapeutic relationship. In general, the importance of inclusive language in the therapeutic endeavor is critical because clients may not choose to disclose sexual orientation at the onset of therapy. We note that although the advantage observed for using inclusive language with heterosexual persons is not particularly strong, it is important that a negative finding was not observed; that is, counselors using inclusive language were not seen as operating in a way that would diminish them in the eyes of these potential service consumers. Therefore, the ethical practice of inclusive language, which has such upside potential with GLB individuals, seems to have little downside potential with other individuals.
In addition, because sexual orientation is an invisible and highly stigmatized identity, it is important for therapists to remember that clients may not always be willing to openly discuss their sexual orientations, particularly at the onset of therapy. In using inclusive language, therapists create environments in which clients are more likely to be open and forthcoming with regard to their sexual orientations. Recent literature has found that sexual orientation can be fluid for many individuals, particularly women (Diamond, 2008). The possible fluid nature of sexual orientation, in addition to the stigma associated with GLB status, lends credence to the use of inclusive language to create a welcoming environment for all clients.
Although there are benefits to using inclusive language with clients because GLB clients may benefit from this, there are also benefits to using inclusive language with heterosexual clients. Clients often keep aspects of themselves hidden, and it is possible that by creating an atmosphere of openness, clients will follow suit and be more willing to disclose with an inclusive therapist. In the present study, the participants in the inclusive language group were significantly more willing to disclose to the therapist than those in the heterosexist language group. It is possible that using language that portrays sensitivity to others may increase clients’ openness to share struggles they may otherwise have kept hidden.
Limitations to the current study include the use of college students who were also recruited from one university. Future research should explore the effect of counselor language with participants of various ages, from different regions of the country, and from various socioeconomic statuses to determine their preferences for counselor type. It may be possible that college students are more open-minded than the general population and the results found in the present study may not generalize beyond a well-educated sample. Rather than simply making this assumption, however, future studies might include a measure of heterosexist attitudes along with the examination of language use impact. In addition, the current study was limited with regard to ethnic diversity. Future research should examine the effect of counselor language with more ethnically diverse samples of individuals because cultural biases may play a role in sensitivity and appreciation of inclusivity with regard to sexual orientation.
Despite these limitations, we expect that the current findings will further encourage individual counselors and agencies that are showing nonheterosexist sensitivity. Because language is an influential tool in therapy, combining these results with the earlier Dorland and Fischer (2001) work suggests that language that can benefit GLB individuals will have little impact on heterosexual individuals. In addition, even though there were only slight differences noted on the three outcome measures used in this study, we expect that there may be some as-yet-unmeasured positive effects that occur with heterosexual people who are exposed to inclusive language.
Footnotes
Acknowledgements
We would like to acknowledge the assistance of Vincent Marconi who helped with data input.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.