Is Allison More Likely Than Lakisha to Receive a Callback From Counseling Professionals? A Racism Au

Is Allison More Likely Than Lakisha to Receive a Callback From Counseling Professionals? A Racism Audit Study

Abstract

Is Allison More Likely Than Lakisha to Receive a Callback From Counseling Professionals? A Racism Audit Study Richard Q. Shin1, Lance C. Smith2, Jamie C. Welch1, and Ijeoma Ezeofor1 Abstract Using an audit study, we examined racially biased callback responses in the mental health field by leaving voicemails soliciting services with practicing counselors and psychologists (N = 371). To manipulate perceived race, an actor identified herself with either a stereotypically Black- or non-Latino White-sounding name. Although the difference in callback rate between the two names was not significant, the difference in voice messages from therapists that either promoted potential services or impeded services was significant. The caller with the stereotypically White-sounding name received voice messages that promoted the potential for services at a 12% higher rate than the caller with the stereotypically Black-sounding name. Limitations, future directions for research, and counseling implications are discussed.

"Extensive, pervasive, and persistent" (Flores, 2010, p. e979) summarizes the inequitable patterns of mental health service delivery between non-Latino White and Black individuals in the United States (Brondolo, Gallo, & Myers, 2009; Cook, McGuire, Meara, & Zuvekas, 2009; Cook, McGuire, & Zaslavsky, 2012; Owen, Imel, Adelson, & Rodolfa, 2012). Black populations are less likely than White populations to enter mental health treatment for both generalist and specialty providers (Lê Cook, McGuire, Lock, & Zaslavsky, 2010; Miranda, McGuire, Williams, & Wang, 2008). Compared to White consumers, when seeking services from primary care providers, Black consumers are less likely to receive referrals for counseling, antidepressant prescription, or any form of care for depression or anxiety (Agency for Healthcare Review and Quality, 2008; Alegría et al., 2008; Lagomasino, Stockdale, & Miranda, 2015). A systematic review of the literature by Flores and the Committee on Pediatric Research (2010) found that Black children have lower adjusted odds of receiving a psychological evaluation, a diagnosis, medication, or treatment while simultaneously having higher proportions of parents with negative expectations about treatment helpfulness. Racial disparities in access to mental health treatment are a microcosm of the persistent and pervasive inequities in the broader U.S. health care system. A considerable body of research conducted over the past few decades in the United States has demonstrated that Black patients receive lower quality health care than White patients after controlling for differences in income, insurance coverage, geographic location, access, disease status, and other clinically germane factors (Beard, Gwanmesia, & Miranda-Diaz, 2015; Kennedy, 2015; Penner & Dovidio, 2016; Valire, 2005; van Ryn et al., 2011). When compared with White patients, Black patients are less likely to receive a prescription for pain (Sabin & Greenwald, 2012), less likely to receive breast and colorectal cancer screening (Rao, Breen, & Graubard, 2016), more likely to experience amputation as a result of diabetes (Trawalter, Hoffman, & Waytz, 2012), and more likely to receive general anesthesia rather than an epidural during Cesarean delivery; they also have poorer heath outcomes with regard to respiratory health, cancer, and cardiovascular disease (Butwick, Blumenfeld, Brookfield, Nelson, & Weiniger, 2016; Celedón, Ewart, & Finn, 2016; Hinchcliff et al., 2016; Taylor et al., 2005). Researchers have attempted to offer explanatory frameworks for the systemic disparities between White and Black individuals. One dominant model centers on how help-seeker behavior attributed to within-group cultural factors of Black people may contribute to the problem (Lee & Richardson, 1991). Constructs such as attitudes toward care (Diala et al., 2000), caregiver strain (McCabe, Yeh, Lau, Garland, & Hough, 2003), receptiveness to treatment (Corrigan & Miller, 2004), intracultural stigma toward services (Corrigan, 2004), culturally distinct beliefs about mental illness and mental health (Rüsch, Angermeyer, & Corrigan, 2005), and styles of expressing mental health-related suffering (Snowden & Yamada, 2005) have been investigated to understand this disparity. This body of literature is useful for understanding the cultural factors that may contribute to racial disparities in mental health care. However, an explanatory model, which focuses exclusively on help-seeker behavior, falls short of addressing the complexity of the problem. This approach fails to consider helpprovider behavior, for example, potential influences of service provider bias, and macrolevel systemic injustices. A lack of attention to potential barriers that are created and sustained on the part of counselors and psychologists sends the implicit message that the disparities in access to mental health treatment are due to deficits among Black populations. The purpose of this study was to initiate a shift away from the emphasis on why Black individuals "fail" to seek mental health services to a focus on what counseling professionals may be doing to block the provision of services to potential Black clients. We believe that examining help-provider behavior is a necessary step in constructing a more comprehensive understanding of the persistent, inequitable patterns of mental health service delivery between White and Black clients. Multicultural Competencies and Social Justice The question of racial bias being perpetrated by mental health professionals has not been ignored within the mental health fields. In fact, counseling scholars have investigated potential bias as a result of race and ethnicity for many years (Jones & Seagull, 1977; Rosenthal & Berven, 1999; Wampold, Casas, & Atkinson, 1981). To reduce the probability of biased clinical behaviors and to reduce disparities in treatment, the field of psychology has emphasized the need to increase multiculturally competent practitioners (Pederson, Carter, & Ponterotto, 1996; Whaley & Davis, 2007). For over 30 years, Sue's (1982) tripartite model of multicultural competence has served as the dominant framework for how counseling psychologists and related mental health professionals can provide effective counseling services to individuals from diverse racial, ethnic, and cultural backgrounds. In 1992, this model was transformed into a set of competency guidelines called the Multicultural Counseling Competencies (MCC; Sue, Arredondo, & McDavis, 1992). The MCC operationalize multicultural counseling as a multidimensional construct with 31 original competencies undergirding the awareness, knowledge, and skills necessary for therapists to be able to provide culturally competent services. 1190 The Counseling Psychologist 44(8) The development of the MCC has had a significant impact on the advancement of the multicultural competency movement in the fields of counseling and psychology. For example, in 2002, the American Counseling Association endorsed the MCC put forth by the Association of Multicultural Counseling and Development (AMCD). Also in 2002, the American Psychological Association (APA) built on the MCC to establish guidelines for addressing cultural issues in clinical practice (APA, 2003; Sue et al., 1992). In 2008, the APA Task Force on the Implementation of the Multicultural Guidelines offered specific recommendations for integrating multicultural competencies into the primary activities of psychologists, including practice, research, education, and policy (APA, 2008). Finally, the most recent iteration of the competencies are now named the Multicultural and Social Justice Counseling Competencies (Ratts, Singh, Nassar-McMillan, Butler, & McCullough, 2015) and were endorsed by AMCD on June 29, 2015, and by the American Counseling Association on July 20, 2015. Key areas of revision include a stronger link between multicultural and social justice competence, greater emphasis on intersectionality, and a new action competency category. A question that has plagued MCC scholars from the inception of the tripartite model is whether it can be supported by empirical evidence. In other words, is there research to support the claim that higher levels of multicultural competence result in more effective counseling practices and better client outcomes? Recent studies analyzing MCC research have provided evidence to support an affirmative answer to this question (Tao, Owen, Pace, & Imel, 2015; Worthington, Soth-McNett, & Moreno, 2007). However, it is unclear whether the proliferation of the MCC has eliminated racially biased practices at the entry point of counseling services because there is a dearth of research investigating this specific question. Beyond the MCC, counseling psychologists and related mental health professionals have been charged to embrace a social justice paradigm to effectively interrupt racial disparities (Vera & Speight, 2003). Scholars in the field define social justice work as a commitment to transforming systems and institutions that prevent an equitable distribution of resources and opportunities (Fouad, Gerstein, & Toporek, 2006). Social justice-oriented counseling psychologists have asserted that the counseling field continues to myopically focus on the intrapsychic struggles of the client rather than examining macrosystemic structures that perpetuate client suffering (Vera & Speight, 2003). The pervasive professional bias toward privileging intrapsychic and intracultural conceptualizations over the effects of oppressive environmental conditions serves to maintain the inequitable societal status quo experienced by members of marginalized social groups (Fouad et al., 2006; Prilleltensky, 1997). The predominant and reductionist help-seeker explanation regarding the persistent disparity in mental health service delivery between White and Black populations provides an excellent opportunity for counseling psychologists to further integrate social justice principles into the field. One way this can be accomplished is by investigating the potential presence of systemic, help-provider implicit racial bias operating within the counseling professions. Implicit Racial Bias Current research that measures racial bias among mental health care providers and those in training has begun to shift attention away from the helpseeker explanatory model and toward a help-provider model. Recent empirical work regarding racial discrimination examines a form of bias that may be unknowingly held by beneficent, egalitarian mental health professionals and students. In contrast to conscious and overt racist attitudes, this form of bias is implicit and is driven by automatic, subconscious prejudicial associations (Boysen, 2010; Cooper et al., 2012; Katz & Hoyt, 2014). Boysen and Vogel (2008) found significant implicit bias against African Americans and sexual minorities in a sample of 105 counseling students. Castillo, Brossart, Reyes, Conoley, and Phoummarath's (2007) research found implicit anti-Black bias among a sample of 84 counseling graduate students. More recently, Katz and Hoyt (2014) sought to examine explicit and implicit antiBlack bias among 173 mental health practitioners. Through the use of the Implicit Associations Test (IAT; Greenwald, McGhee, & Schwartz, 1998) along with other instruments, Katz and Hoyt (2014) concluded that the automatic anti-Black prejudice found in their sample of mental health professionals and trainees was "worrisome" (p. 302). It is interesting that Katz and Hoyt discovered that the ethnic or racial identity of the professionals in their sample had no statistically significant effect on rates of racial bias. Implicit bias has been conceptually connected to inequitable diagnosis and assessment (Gushue, 2004). For example, Black clients experience biased clinical judgment from their mental health providers (Monnot, Quirk, Hoerger, & Brewer, 2009), such as being disproportionately diagnosed with psychotic spectrum disorders relative to White clients (Feisthamel & Schwartz, 2006; Perry, Neltner, & Allen, 2013). Pro-White/anti-Black implicit bias has been observed among the larger population of health care providers as well (Ewen et al., 2015; Green et al., 2007; Johnson et al., 2016). In a recent review of literature that examines the extent and measurement of health care provider interpersonal racism, 26 out of 37 studies produced statistically significant evidence of interpersonal racism on behalf of providers toward people of color (Paradies, Truong, & Priest, 2014). Hall et al. (2015) also conducted a review of implicit racial bias by examining 15 cross-sectional studies that used the IAT (Greenwald et al., 1998) to assess bias amongst health care providers. Their review found low to moderate levels of implicit pro-White/anti-Black bias among health care professionals in all but one study, demonstrating "that implicit bias was significantly related to patient-provider interactions, treatment decisions, treatment adherence, and patient health outcomes" (Hall et al., 2015, p. e60). The extant literature regarding implicit anti-Black bias perpetrated by mental health professionals and those in training provides support for the notion that there may be systemic and/or institutionalized provider-helper barriers operating for Black mental health consumers. If indeed mental health professionals are contributing to the inequitable patterns of service delivery experienced by Black individuals, then additional social justice-oriented research should be conducted to uncover our field's potential complicity in this systemic disparity. In this study, we sought to explore potential racial bias at the entry point of mental health service delivery. Audit Studies Although the research measuring implicit bias during diagnosis and treatment is expanding, missing from the literature is empirical work that examines implicit or explicit bias at the entry point for mental health care. In other words, could racial bias impede potential Black clients from stepping through the door of mental health practitioners' offices? Could such racial bias at the entry point of services be so widespread that it is institutional in its effect, comprising a significant component of the systemic racial disparity in access to services? The literature on racial bias within the mental health fields warrants such a line of inquiry. Thus, we chose to conduct an audit study targeting practicing counselors and psychologists to test for an anti-Black bias effect in the mental health field. Racial audit methodology is a type of field experiment originally designed to measure discrimination in the labor and housing markets (Galster, 1987; Jowell & Prescott-Clark, 1970). Given the significant limitations of selfreport methodology for measuring racial bias, audit designs are constructed around an elegantly simple, experimental premise: Comparable White actors and actors of color are asked to pursue identical goals, such as obtaining employment or housing, and variance in outcomes is then measured. Early studies used actors who were matched according to physical characteristics. The actors received significant training and were then sent into actual recruitment meetings or job interviews (Newman, 1978). Fisher and Massey (2004) expanded the design by conducting a phone-based audit study 

Multicultural Competencies and Social Justice

of the rental housing market: Auditors called listings to inquire about availability of units while using either stereotypical White middle-class English (WME), stereotypical Black-accented English (BAE), or Black English Vernacular (BEV). The researchers found that all indicators of housing access declined as callers' accents moved from WME to BAE to BEV. Bertrand and Mullainathan (2004) significantly increased the potential of audit methodology by designing a study that eschewed the need for live auditors. Rather than using live actors, they developed a correspondence audit design wherein identical fictitious resumes were randomly assigned stereotypically Black- or White-sounding names, and then sent to potential employers who had posted help wanted ads. The researchers then measured the variance in callback rates and found that individuals with White-sounding names received 50% more invitations for interviews. More recently, Nunley, Pugh, Romero, and Seals (2015) utilized correspondence audit study methodology and found that, within a sample of 9,400 online job openings, Black-named recent college graduate job seekers were approximately 14% less likely to receive interview requests than applicants with White-sounding names. Racial audit methodology has also been recently applied to the field of higher education. Milkman, Akinola, and Chugh (2012) conducted a study that entailed sending emails to university faculty members from fictional prospective doctoral students seeking a meeting to discuss research opportunities either on the day the email was sent (now condition) or in one week (later condition). Like the work done by Bertrand and Mullainathan (2004) and Nunley et al. (2015), perception of race and gender were manipulated via the name of the fictitious prospective doctoral student. Milkman et al.'s (2012) findings revealed that White men were granted access to faculty members 26% more often than were White women and members of racially marginalized groups. The discrimination that has been documented in the housing and labor markets, and in higher education through the use of audit studies, varies in degree; however, deleterious outcomes are consistently observed for names or voices associated with persons of color. Because audit study methodology provides "some of the cleanest nonlaboratory evidence of differential treatment by race" (Bertrand & Mullainathan, 2004, p. 993), it is a valuable means by which counseling psychology scholars can examine racial bias and inequitable service delivery within the mental health care fields. Purpose of the Study The purpose of the present study was to assess racial bias at the entry point of counseling services. Specifically, we were interested in examining two dependent variables (a) callback response rates and (b) responses that promote the potential for services. For the purposes of this study, callback response rates were defined simply as the rates at which counselors called back when a message was left in their voicemail. Responses that promote the potential for services was defined as instances where a service provider responded by both returning the actor's phone message and extending an invitation for a conversation indicative of the counselor's interest in promoting future services. Invitations for a phone conversation may have been offered for several reasons. For instance, invitations may have stemmed from a therapist's desire to learn more about the client's presenting issue to determine goodness of fit. Therapists may also have been interested in scheduling an appointment or they wanted to refer the client to another provider. Conversely, responses where service providers either failed to return a phone call or indicated they could not take on a client due to their caseload being full were defined as responses that impeded services. Based on the previously cited counseling psychology literature and similarly designed audit studies (e.g., Bertrand & Mullainathan, 2004; Milkman et al., 2012), we predicted that counselors and psychologists would exhibit greater bias against a caller using a stereotypically Black-sounding name. As such, we hypothesized the following: 1. There would be a statistically significant association between the name of the caller (i.e., the Black-sounding name versus the Whitesounding name) and the likelihood of a returned phone call by the therapist (i.e., receiving a callback vs. not receiving a callback). The caller with the White-sounding name would receive significantly more callbacks from therapists than the caller with the Black-sounding name. 2. There would be a statistically significant association between the name of the caller and the likelihood a therapist would respond in a manner that promoted the potential for services (i.e., calling back and receiving an offer for a conversation vs. not calling back or declining service). The caller with the White-sounding name would receive significantly more responses that promoted the potential for services from therapists than the caller with the Black-sounding name.

Method

The present study employed a between-subjects field experiment design with a sample of licensed, professional counselors and psychologists. Drawing from similar racial audit studies that have used names to signify race (Bertrand & Mullainathan, 2004; Fisher & Massey, 2004; Milkman et al., 2012), we selected the name "Allison" to represent a non-Latina White client and the name "Lakisha" to represent a non-Latina Black client. Bertrand and Mullainathan (2004) reported that within the U.S. population, the likelihood that "Allison" represents a White individual is quantified at .925 and the likelihood that "Lakisha" represents a Black individual is quantified at .967. To "recruit" mental health professionals, the researchers identified phone numbers for licensed counselors and psychologists within an East Coast, Mid-Atlantic state using online therapist referral databases. The researchers compiled these phone numbers into a database and randomly assigned the numbers into two groups. Procedures were identical for each group, except that one group received voice recordings from Allison and the other received voice recordings from Lakisha (see Appendix). Collected phone numbers were assembled from national databases of licensed counselors and psychologists. Because these phone numbers were gleaned from online databases, the researchers were not able to collect information regarding the race, ethnicity, gender, or socioeconomic background of the participants, nor any other demographic variables. Each phone number was randomly assigned to one condition, and precautions were taken to ensure that no service providers received messages from both Allison and Lakisha. Because the phone numbers collected were randomly assigned to their conditions, we can assume that the demographics for the sample were similar between conditions and similar to the demographics of all licensed counselors and psychologists in the East Coast, Mid-Atlantic state from which the sample was drawn. Random distribution of conditions across the sample strengthens the likelihood that statistically significant findings are likely due to the racial-sounding names. As our aim was to uncover structural, systemic racial bias, we determined that the demographic information of the individual counselors called was not germane to our findings. Regardless of the social identities held by the individual therapists called, the focus of this study was to assess for potential bias operating within the counseling and psychology fields broadly, rather than spotlight individual groups within it. The actor did not speak with participants directly, but rather prerecorded audio clips were left in participants' voicemail boxes by members of the research team. The same actor recorded both of the audio clips (see Appendix) to ensure they would be virtually identical in vocal cadence, tone, and manner of speaking. The only differences between the recordings was the name of the caller (e.g., "Hi, my name is Allison . . ." or "Hi, my name is Lakisha . . .") and the number the mental health professional was provided to callback. Prior to the start of the primary investigation, a pilot study was conducted using 86 undergraduate students. The students were asked to listen to a nameless clip of the recording and describe the speaker's gender, race, and social class. Without hearing a racially identified name, participants perceived our confederate to be a woman (100%), middle to upper class (85%), and Black (55%). During the field experiment, separate prepaid phones were used to establish voicemail inboxes that would store the callbacks for Allison and Lakisha, respectively. In this way, callback responses for Lakisha and Allison could be tracked separately. To increase the chances that the calls would go unanswered and instead be transferred to the participants' voicemail boxes, calls were made on a Sunday evening. When a member of the research team heard the participant's voicemail prompt, the appropriate recording was played over the phone. Messages were only left directly on providers' voicemail systems. If a call was picked up by a message service, researchers ended the call without leaving a voicemail. In the event that a participant picked up the call, the researchers would explain that they had dialed a wrong number and end the call. For calls that remained unanswered and were not transferred to voicemail, the calls were abandoned after one minute of ringing. In total, 198 voicemails were successfully placed requesting services for Allison and 173 voicemails were successfully placed requesting services for Lakisha. Return voice messages from participants for both the Allison condition and the Lakisha condition began to arrive in the researcher's designated prepaid cell phone voicemail boxes within 24 hours. One week after the initial round of phone calls were made, researchers called each participant for whom a service soliciting voicemail had been left (i.e., regardless of whether the participant had followed up with a callback for Allison or Lakisha) to let the participant know that Allison or Lakisha was no longer interested in pursuing counseling. The responses were then categorized into two mutually exclusive groups: (a) responses that promoted the potential for services and (b) responses that impeded services. The current study complied with both U.S. Department of Health and Human Services Office for Human Research Protections (OHRP) and the APA Ethical Principles of Psychologists and Code of Conduct (Ethics Code; APA, 2010). Data collection and analysis were supervised by the first two authors with the approval of the institutional review boards of the University of Maryland, College Park, and the University of Vermont. The use of an audit research design requires the consideration of several ethical issues given that deception is an essential component of the procedures. The authors and their institutional review boards determined that deception was permissible given that (a) the study involved no more than minimal risk to participants, (b) the participants did not represent a vulnerable population, (c) the waiver of consent would not adversely affect the welfare of the participants, (d) the study could not practically be carried out without the waiver, and (e) the knowledge being sought was important enough to justify deception. It was determined that our participants, as highly educated professionals, met the criteria for a deception study. Furthermore, the effects of institutional racism are so damaging to Black people, that the minimal risk to our participants warranted the use of deception as we sought to document potential racism within the field of mental health. Finally, as opposed to most studies that use deception, participants in audit research designs are typically not debriefed. In our study, we decided not to debrief participants because doing so would not correct stressful misperceptions nor reduce anxiety but, to the contrary, could result in generating more harm to participants than good (see Standard 8.08b of the APA Ethics Code; OHRP, 2001). Given that practicing counseling professionals regularly receive inquiries from potential clients who then change their mind or decline services, it was decided that participants were unlikely to experience anything unusual or stressful by participating in our study

Results

We conducted two chi-square tests of independence for 2 × 2 contingency tables and calculated a Pearson χ2 statistic for each test. The chi-square test of independence allows for hypothesis testing with categorical variables, in which the null hypothesis is that the probability of one event occurring is independent of the probability of another event occurring, that is, that the events are independent of each other (B. H. Cohen, 2001; Reynolds, 1977). A chi-square test is appropriate when variables are nominal or ordinal, the data are randomly sampled, the levels of every variable are mutually exclusive, the events are independent, and the expected value for all of the cells exceeds five (B. H. Cohen, 2001). Among the total phone calls made by the research team, 24 and 29 of the calls placed for Lakisha and Allison, respectively, were answered by therapists. The number of phone calls answered by therapists between each condition did not differ significantly, χ2(1) = 0.034, p = .854. Answered phone calls were excluded from the chi-square analyses used to test our hypotheses. To test the first hypothesis, we examined the relationship between name of caller and frequency of calls returned. A total of 371 calls were made. Of these calls, 198 were placed for Allison and 173 were placed for Lakisha. Allison and Lakisha received 131 and 99 callbacks, respectively. In terms of percentages, of the calls that were completed, 66% were returned for Allison and 57% were returned for Lakisha. We found that the observed not differ significantly from the expected values, χ2(1) = 3.129, p = .077. The phi for this test was .092. Phi (φ) is a measure of the strength of association between the two variables with ranges from 0 to 1, and it provides an effect size for the χ2 test statistic. In other words, phi represents the mean percentage difference (9.2%) between the numbers of calls returned between Allison and Lakisha. These findings suggest that the caller's name was not related to the likelihood of therapists returning her call. To test our second hypothesis, we conducted a chi-square test to compare responses that promoted the potential for services between Allison and Lakisha. This analysis was conducted using the same 371 calls analyzed in the first hypothesis. Again, 198 calls were placed for Allison and 173 calls were placed for Lakisha. Of the outgoing calls completed, Allison was invited to participate in a phone conversation with a therapist 63% of the time (126 voice messages that promoted the potential for services), whereas Lakisha was invited to participate in a phone conversation with a therapist 51% of the time (89 voice messages that promoted the potential for services). The second hypothesis presumed that of the total calls made by each caller, there would be a significant difference in the responses in terms of promoting the potential for services. This was of interest because receiving an invitation for a conversation that indicates a counselor's interest in promoting future services, more so than a simple callback, is the primary concern for callers who are seeking mental health treatment. The results of this test revealed that the observed values deviated significantly from the expected values, providing evidence of a statistically significant association between the name of the caller and the likelihood of her receiving a response that had the potential to promote services, χ2(1) = 5.631, p = .018, φ = .123. Once again, phi represents the mean percentage difference (12.3%) between the numbers of responses that promoted potential mental health services for Allison and Lakisha.

Discussion

This study represents one of the first racial audit field studies set in the context of the mental health profession. The results provide evidence to suggest the existence of systemic racial discrimination at the entry point of services among practicing counselors and psychologists. This finding is consistent with the literature that has found racial inequity levied against Black individuals in every major sector of U.S. society, including medicine (Blair et al., 2013), employment (Bertrand & Mullainathan, 2004), housing (Fisher & Massey, 2004), and education (Farkas, 2003). Because of the relatively small body of literature in the mental health fields regarding systemic racial bias, Shin et al. 1199 the findings from the current investigation may be best situated within the larger problem of disparities in access to quality health care between Black and White populations (Brummer, Reyes, Martin, Walker, & Heron, 2016; Valire, 2005). In our audit study, which targeted practicing counselors and psychologists, a statistically significant difference was not observed in number of callbacks between Lakisha and Allison. However, among the participants who were sampled in our study, Allison was significantly more likely than Lakisha to receive a response that promoted the potential for services. We found that for the 198 calls Allison "made," 63% resulted in invitations for a conversation, which indicated the therapist's interest in promoting services for her. In contrast, of the 173 calls Lakisha "made," only 51% resulted in invitations for a phone conversation. In other words, the fictitious client with a stereotypically White-sounding name had a 12% greater chance of having a therapist open the door to potential mental health services by returning her phone call and offering the opportunity to have a conversation, rather than closing the door by failing to return her phone call or leaving a message that declined services. The observed effect size, which can serve as a measure of clinical or practical significance, was .123. This indicates that there was a nontrivial association between the caller's name and her ability to access services. However, caution must be taken in drawing firm conclusions because an effect size of .123 is considered small (J. Cohen, 1992). The findings from the current study must be replicated before we can confidently conclude that there is systemic bias toward stereotypically Black-sounding names occurring within the fields of counseling and psychology. That being said, we propose that help-provider behavior, specifically implicit racial bias among counselors and psychologists, should continue to be investigated as a possible factor contributing to the persistent inequitable patterns of mental health service delivery for Black consumers. Furthermore, we suggest that our findings may be suppressed for several reasons. First, as opposed to other telephone audit studies investigating disparities between White and Black participants (Fisher & Massey, 2004; Massey & Lundy, 2001), we chose not to employ the use of Black linguistic styles, which may have provided a stronger prime for racialized or classed stereotypes based on the name Lakisha. In other words, we chose the lowest dosage (a Black-sounding name using Standard American English and not Black accented speech) for our experimental condition and still found a statistically significant difference in voice messages that promoted potential services. Another potential reason for our somewhat modest findings is that mental health professionals must operate according to the ethical principle of 1200 The Counseling Psychologist 44(8) "do no harm," which obligates them to return a potential client's phone call. This professional responsibility distinguishes the participants in our investigation from those in audit studies that sought to measure bias in employment, housing, and academia. Therefore, it is important to consider the possibility that some of the participants who called Lakisha back and declined services due to full caseloads did indeed have openings. These therapists may have been simply seeking to meet a minimal standard of compliance with ethical practice, thus potentially suppressing our findings. It is actually quite disturbing that despite the ethical obligations of counselors and psychologists, a little under half of the messages that Lakisha left and one-third of the messages that Allison left, were unreturned. The existence of disparities between White and Black populations in access to mental health treatment has been well documented. However, the majority of explanations explored in the counseling and psychology literatures have focused on help-seeker behaviors and within-group cultural factors like trust, receptiveness to treatment, stigma, and incongruent cultural beliefs about mental illness and mental health (Snowden & Yamada, 2005). Aside from a limited group of dated studies within the rehabilitation counseling field documenting that Black potential clients are less likely to be accepted for mental health services compared to White clients (Atkins & Wright, 1980; Dziekan & Okocha, 1993; Feist-Price, 1995) and some current research focused on biased clinical judgments and diagnoses (Feisthamel & Schwartz, 2006; Monnot et al., 2009; Perry et al., 2013), there is a relatively small body of research that has placed the spotlight inward and toward the existence of potentially biased practices within the counseling fields. The lack of attention to potential barriers that are created and sustained on the part of counselors and psychologists subtly reinforces the message that the disparities in access to treatment are due to deficits among Black populations. The findings from the current study may provide support for the burgeoning body of research investigating the processes and consequences of implicit racial bias among practicing mental health professionals and those in training. A growing body of empirical evidence has documented that therapists are not immune from internalizing the racial biases that continue to be prevalent throughout contemporary U.S. society (Burkard & Knox, 2004; Sue, 2005). In addition, it is possible that counselors and psychologists are particularly susceptible to the type of automatic prejudice triggered merely by a racialized name like Lakisha. Research has shown that individuals can genuinely hold egalitarian values and view themselves as low in prejudice and simultaneously demonstrate harmful prejudicial associations at an automatic level (Devine, 1989). Therefore, although a counselor or psychologist consciously possesses an overt and strong attachment to values such as fairness Shin et al. 1201 and egalitarianism, they may be simultaneously in denial of subconscious biases that can be activated by a racialized name.

Answers the following questions

Discuss the methods that were used for the "Allison Vs. Lakisha" Study . This discussion should include conversation around the sampling method, measures (IV and DV), and design. You should also discuss potential internal validity threats and how they managed some of these. What design did the researchers use? Why did the researchers use the design that they did? How would this have been different if they had used a correlational or survey design instead?