Age, Gender and Health Bias in Counselors: An Empirical Analysis.

A substantial body of research shows that psychologists, social workers and psychiatrists exhibit age and gender bias that affects the way they diagnose and treat older clients. Other findings show bias against people who are in poor health. Relatively little research has examined how vulnerable professional counselors are to these biases. The current study compared the responses of professional clinical counselors with psychologists and clinical social workers. Using the Age Bias Questionnaire the research found that mental health professionals judged older clients significantly less competent and less likely to improve than younger clients. They also judged female clients as less competent than male clients regardless of age. This pattern was consistent across all three professions. Length of time in practice had an inverse relationship to judgment of competency. No evidence of health bias was found. Implications of these findings for access to counseling services for older Americans are discussed.

Persons over the age of 65 underutilize mental health services. Between 10% and 40% of older adults suffer from some type of mental disorder (Molinari, 1996), yet this age group accounts for only 4% to 5% of the clients in community mental health agencies (Lagana, 1995; Lasoski, 1986). They also represent only 9% of all clients seen in private mental health offices (Gottlieb, 1994). Molinari calls this population the most underserved age group in the United States. A number of barriers to treatment exist for older adults (Nordhus, Nielson, & Kvale, 1998), but two merit special attention. The first is the belief systems of elders (Raue & Meyers, 1997). Their beliefs often include negative feelings about aging, acceptance of the society's stereotypes about them, and negative expectations about the capacity of counseling to be useful to them (Lagana, 1995; Lasoski, 1986). Attitudes of mental health professionals towards older adults form the second major barrier to treatment (Knight, 1996; Raue & Meyers, 1997). Researchers have found that mental health professionals are reluctant to work with older persons and perceive them of less interest than other adults (Gatz & Pearson, 1988; Kastenbaum, 1963; Myers, 1998). They also tend to view older individuals as having poorer prognoses than younger individuals (Busse, 1994; Ford & Sbordone, 1980: Ray, Raciti, & Ford, 1985), though the research is not unanimous on this point (Wadsworth, 1996). Myers (1998) concluded that counselors tend to view older individuals as being more set in their ways, less able to change, and less likely to benefit from mental health services in spite of the absence of any empirical evidence to support that conclusion.

In fact, recent evidence suggests that elders are more open to changing their beliefs than those at mid-life (Visser & Krosnick, 1998). Treating older adults for psychosocial stress and mental illnesses has been found beneficial not only to those concerns (Nordhus et al., 1998), but also in alleviating physical discomfort related to medical problems (Klausner & Alexopoulos, 1999). In spite of this substantial body of evidence that counseling is effective for older adults, Myers (1998) reported that counselors-in-training have little interest in specializing in issues faced by older clients. In addition, Woolfe and Biggs (1997) concluded that counselors have little insight into the concerns of this age group. Similarly, Gatz, Karel, and Wolkenstein (1991) found that only a small percentage of psychologists currently working with older clients have specialized training in this area.

Negative biases towards older adults have also been shown to influence the diagnostic process. Mental health professionals tend to attribute problems of older adults to situational factors over personality problems, even when evidence of personality dysfunction is present (James & Haley, 1995; McConatha & Ebener, 1992; Morrow & Deidan, 1992; Ray et al., 1985). For example, McConatha and Ebener (1992) found that, when presented with vignettes of client statements in opening counseling sessions that varied only by the age of the client, counselor trainees rated younger clients as having more perceived problems than older clients. In addition, trainees rated depression in younger clients more severe than depression in older clients, even though the symptoms described in each vignette were identical. The underlying theme seems to be that the experience of symptoms of mental illness is normal for older adults and abnormal for younger adults. McConatha and Ebener also found that counselor trainees felt that professionals over the age of 45 would be most appropriate to work with older clients.

These biases have been reported among a number of professionals serving community mental health needs in both outpatient and inpatient settings. Ray et al. (1985) reported that psychiatrists of varying theoretical orientations believed that older individuals were less ideal for psychological help and had poorer prognoses than younger individuals. James and Haley (1995) found that psychologists believed that psychotherapy was less appropriate for older individuals than for younger individuals. Rohan, Berkman, Walker, and Holmes (1994) found that social workers treating oncology patients had a higher number of contacts with younger patients than with older patients and spent less time addressing significant problems with the older patients.

Taken together this body of research suggests that mental health professionals may be stereotyping older individuals in ways that limit their access to needed services and that compromise their rights as competent adults. Such attitudes are clearly contradictory to the ethical standards of the counseling profession and related mental health disciplines. Several sections of the American Counseling Association (ACA) code of ethics emphasize counselors' responsibility to avoid discrimination against any group and direct the professional to act in ways that respect the dignity of the client (ACA, 1995). For example, Section A.1.a reads: "The primary responsibility of counselors is to respect the dignity and to promote the welfare of clients," and Section C.5.a reads: "Counselors do not discriminate against clients, students, or supervisees in a manner that has a negative impact based on their age, color, culture, disability, ethnic group, gender, race, religion, sexual orientation, or socioeconomic status, or for any other reason."

The ethical standards of the profession also support the autonomy of the adult client, both in decision making about matters that concern them and in control of counseling information revealed to others (Welfel, 1998). Therefore, the tendency of counselors and other mental health professionals to stereotype older adults and to view the autonomy of older persons as different than other adults is troubling. However, most of the published research has not included practicing community mental health counselors as participants. Consequently, the degree to which counselors endorse such elder bias is not fully understood. The current study addresses these issues and the relationship between age, gender bias, and health bias.

Needless to say, stereotyping based on age is not the only type of bias found in research. Evidence of professional bias based on gender, gender role, and health problems has also been published (James & Haley, 1995; Matyi & Drevenstedt, 1989; Seem & Johnson, 1998). The influence of gender on mental health treatment and diagnosis has been studied since the early 1970s. At that time Broverman, Broverman, Clarkson, Rosenkrantz, and Vogel (1970) found that mental health professionals used different descriptors for healthy men and healthy women and that the descriptors for a healthy adult closely matched those for a healthy man. Healthy women were characterized as being more submissive, less aggressive, more excitable, and less competitive than healthy men or healthy adults. These findings prompted a flurry of research on gender bias. For example, the Broverman findings were replicated in 1985 by O'Malley and Richardson, who indicated that psychologists, social workers, and counselors continued to describe both males and females in stereotypical terms. Matyi and Drevenstedt (1989) found that psychology students perceived older female clients as having better memories and being more alert than older males. Thus, a second purpose of this study was to examine whether practicing mental health counselors showed evidence of gender bias and whether age and gender bias interacted to show different attitudes towards older women than older men.

A review of the published research for the last two decades revealed almost no empirical research that explored the interaction of age and gender as an influence on counselor judgment. One exception was the work of Busse (1994), which showed that male psychologists judged older female clients as less psychologically minded and less ideal therapy candidates than either males in general or younger women. Theories and anecdotal evidence of a stronger negative bias against older women than younger women or males abound however. Sherman (1997), for example, uses historical and literary analysis to argue that older females are especially devalued in Western society. Nadien (1996) connects the increased vulnerability of women to elder maltreatment to negative perceptions of them. This gap in the empirical literature led to the second goal of the study, exploring the relationship of age and gender bias among practicing clinicians.

The third purpose of the study was to examine whether the health of a client affects a counselor's judgment about client competence or prognosis. Healthism is the term coined by James and Haley (1995) who reported that psychologists rated clients who were in poor health as less able to establish an adequate counseling relationship, and less appropriate for counseling intervention than their healthier counterparts. They were also seen as having a poorer prognosis (in mental health areas) and more likely to commit suicide than those clients who were in good health. Ford and Elliott (1999) also found an influence of health considerations on the objectivity of mental health professionals in diagnosing depression. Specifically, their research revealed that clinicians tended to diagnose the depressive symptoms of clients with health problems more leniently than those without health problems. In both studies, the health of the client influenced the judgment of the professional in ways that are inconsistent with accurate diagnosis and treatment of psychological problems.

Finally, the study sought to explore the influence of the type of license of the clinician and the number of years as a practicing professional on their judgments about client diagnosis and treatment. Prior research has not explicitly compared practitioners from different disciplines in the same study nor has it tested the contribution of clinician experience on age, gender or health bias and their interaction.

In sum, the current study addressed four major research questions:

1. Do the age, health status, and gender of clients affect mental health professionals' perception of client competence?

2. Do the age, health status, and gender of clients affect mental health professionals' perception of client prognosis?

3. Does professional discipline (professional clinical counselor, psychologist, or clinical social worker) affect perception of client competency and client prognosis?

4. Does the amount of time practicing as mental health professionals affect clinicians' perception of client competency and client prognosis?

METHOD

Participants

Forty two Licensed Independent Social Workers (LISW), 42 psychologists and 43 Licensed Professional Clinical Counselors (LPCC) were randomly selected from licensure lists obtained from their respective state licensing boards in Ohio, producing a pool of 127 potential participants. A minimum sample size of 30 participants from each discipline was established in order to provide enough power to detect any differences between the groups (Gay & Airasian, 2000). Potential participants were contacted by telephone by the primary researcher and an assistant and asked to volunteer. The assistant, a doctoral student in counseling, was trained by the researcher to present the procedures to participants and was then observed to ensure consistency in procedure.

Of the 127 professionals contacted, 28 individuals declined to participate. Of the 99 professionals surveyed, 94 (95%) returned the survey to the researcher. One survey was returned blank so it was eliminated from the analysis. Ultimately, 93 licensed mental health practitioners produced useable results. The practitioners were divided among 32 professional clinical counselors (LPCCs), 31 clinical social workers (LISWs), and 30 licensed psychologists.

Participants ranged in age from 28 years old to 75 years old (M = 46.88, SD = 9.37) with psychologists being older on average (M =49.57, SD = 8.94) than clinical social workers (M = 46.58, SD = 13.40) or professional clinical counselors (M = 46.28, SD = 9.12). The clinical experience of the clinicians ranged from one year in practice to 45 years in practice (M = 15.06, SD = 8.34) with psychologists (M = 17.80, SD = 9.27) and clinical social workers (M = 17.59, SD = 16.95) having more years in practice than professional clinical counselors (M = 12.71, SD = 6.87). The sample was 71% female. Gender by discipline was divided as follows: 81% of the professional clinical counselors were female, 77% of the clinical social workers were female, and 53% of the psychologists were female. 99% of the participants were White, 1% was Asian; 61% of the sample worked in public community mental health centers, and 39% were employed in private practice.

Instrument

The instrument used in this study, the Age Bias Questionnaire, was developed by the researchers. It included four vignettes of hypothetical client situations. Following is a sample vignette of a woman who is experiencing major depression.

Ms. Green is a (52- or 72-)year-old married woman in (good health or with diabetes). She went to her physician several weeks before seeking mental health counseling complaining of being tired all the time yet not being able to sleep. She also reported that she felt agitated a good part of the time and had trouble sitting still. The physician prescribed sedatives to help the agitation and insomnia and recommended that she contact a mental health professional. Ms. Green is a well-dressed attractive woman. She reports that she has felt "down" for almost a year. Ms. Green reports that she has felt "down" before, but it has never seemed as severe or lasted as long. She complains of lethargy and insomnia and is averaging only three to four hours of sleep per night. Over the last several months she has experienced a decrease in appetite and feels like she has to remind herself to eat. She has lost five pounds. She also complains that she seems to be very forgetful and finds that her thinking gets confused on occasion. She has trouble concentrating and finds making decisions difficult. She has feelings of hopelessness and sadness and states that sometimes she wishes she could just go to sleep and never wake up. She states repeatedly that she doesn't wish any of her friends or family to know she is experiencing difficulties.

The other vignettes dealt with a male client with generalized anxiety disorder, a widowed female client experiencing major depression, and a retired male client experiencing an adjustment disorder.

Symptoms for each hypothetical client were taken from the Diagnostic and Statistical Manual of Mental Disorders, (4th ed.) (American Psychiatric Association, 1994). The vignettes were designed to represent three diagnostic categories from the DSM-IV: major depression (used twice because of its frequent occurrence among older adults), generalized anxiety disorder, and adjustment disorder. The vignettes were identical except for variations in age and health status of the clients. In the variations that represented clients with health problems, one client was depicted as having heart disease, one as having kidney disease, one as having diabetes, and one as having severe arthritis. These medical conditions were selected because they represented appropriate health problems for both age groups and implied that the client was experiencing chronic rather than acute health problems (Leventhal, Idler, & Leventhal, 1999). The vignettes were designed to represent two female clients and two male clients. Immediately after reading each vignette, participants were asked to respond to seven questions dealing with diagnosis, evaluation of client prognosis, and evaluation of client competence. The participants were given nine possible diagnoses (major depression, bipolar disorder, panic disorder w/agoraphobia, generalized anxiety, adjustment disorder, v-code designation, cyclothymia, primary insomnia, or other condition) and asked to judge the best diagnosis for the hypothetical client. Next, participants were asked to evaluate the competency of the client by rating their estimate of the client's competency to understand counseling and to give informed consent. They were also asked to rate the client's prognosis for improvement. Perceived level of competency and prognosis were measured on 7-point Likert-type scales ranging from (1) poor to (7) excellent.

Validity and Reliability

The Age Bias Questionnaire was initially reviewed by three expert mental health professionals and two doctoral-level counseling students in order to gather data regarding the content validity of the instrument. The reviewers were asked to assess its clarity and appropriateness and to verify the match between the symptoms presented in the vignettes and the DSM-IV criteria. In addition, the reviewers were informed of the purpose of the study and asked to assess the proposed age and health manipulations. Subsequent to this review, the wording of the vignette representing the woman who had recently been widowed was changed in order to clarify her situation. As an additional check on face validity, the instrument was piloted with six counseling trainees. That pilot revealed no further problems with item clarity.

Several statistical analyses of the instrument based on data obtained in the pilot study were conducted to ascertain reliability. First, frequencies were obtained to test diagnostic accuracy. Accuracy of the pilot study participants' diagnoses for three of the vignettes representing hypothetical clients was 66%. The fourth vignette (representing a diagnosis of major depression) yielded an 83% accuracy rate. Given the wide variability in diagnostic accuracy reported in the literature (Kirk & Kutchins, 1992), these rates seemed acceptable for a new instrument. Second, as a check of internal consistency a Cronbach's alpha was computed ([Alpha] = .63). After one item that proved to be confusing was removed from the instrument, the Cronbach's alpha level of the survey was 0.78. There is no agreed upon acceptable level of reliability, but Conway, Jako, and Goodman (1995) state that .69 is acceptable by conventional standards.

Procedure

Once a mental health professional agreed to participate, the primary researcher or the assistant arranged a mutually convenient time with the participant to deliver the measures. Actual delivery and retrieval of the instrument was done by the researcher and the researcher saw each participant at the time of delivery. The researcher and participant completed the informed consent procedures and then agreed upon a date and time for the researcher to retrieve the completed instrument. The participants were asked to read each vignette and to answer the questions that followed. They were told to put the instrument in the accompanying envelope when they had completed it. They were also asked not to put any identifying information on the envelope to ensure anonymity of response. The personal delivery method was chosen instead of a mailing to increase response rate, and it appeared to have the desired effect. All instruments were coded to exclude identifying data and kept in a locked file.

Design

Two independent variables were manipulated in the study--age and health of the hypothetical clients. The vignettes were designed so that each participant reviewed one hypothetical client in each combination of age and health status (younger-healthy client, younger-unhealthy client, older-healthy client, and older-unhealthy client). Three other independent variables were examined--the gender of the client (two hypothetical clients were male and two were female), the professional discipline of the clinician (professional counselor, social worker, or psychologist), and number of years in practice. In addition, an effort was made to survey approximately equal numbers of each professional discipline. The two dependent variables were counselor judgment of client competence and counselor judgment of prognosis.

RESULTS

Because data on the psychometric properties of the instrument were limited, the first step in analysis of the results was an analysis of the accuracy of diagnosis. Overall accuracy of diagnosis was 77%. The accuracy per vignette ranged from a high of 81% for generalized anxiety to a low of 72% for adjustment disorder. Clinical social workers and psychologists had a rate of diagnostic accuracy (81% and 80% respectively) somewhat higher than professional clinical counselors (71%), though chi square analysis, [[Chi].sup.2](2) =.926, p [is greater than] .05, indicated that these differences were not statistically significant.

The relationship between the two dependent variables (judgment of competence and judgment of prognosis) was assessed using the Pearson Product Moment correlation coefficient. A significant correlation between the two variables emerged (r = .49, p [is less than] .05), justifying the use of a multivariate model (Hair, Anderson, Tatham, & Black, 1995).

The multivariate analysis of variance showed that the following independent variables each had a significant effect on at least one of the dependent variables, as measured by Hotteling's Trace: client age (F(1,367) = 3.48, p =.002), client gender (F(1,367) = 2.58, p =.019), number of years in practice (F(1,367) = 2.22, p =.041), and the interaction of client age, client gender, and clinician license (F(1,367) = 4.32, p =.014). Client health status had no statistical effect on either of the outcome variables.

Tests of between subject effects were performed in order to determine which independent variables had significant effects on which of the two dependent variables. The Table contains the means and standard deviations for these variables.

Table 1. Mean and Standard Deviations for Judgment of Competency by Client Variables

Variable                       Younger Client         Older Client
                               M       SD      N*     M       SD

Judgment of Competence
         Client Gender
                Male           5.25    1.05    89     4.98    1.08
                Female         4.94    1.20    97     4.55    1.22
         Client Health
                Healthy        5.13    1.28    92     4.73    1.17
                Not Healthy    5.04    0.98    94     4.82    1.17
         Total                 5.09    1.14    186    4.77    1.16

Judgment of Prognosis
         Client Gender
                Male           4.92    0.92    89     4.51    1.02
                Female         4.80    1.04    97     4.38    1.08
         Client Health
                Healthy        4.92    1.01    92     4.46    1.03

                Not Healthy    4.80    0.96    94     4.46    1.03
         Total                 4.86    0.98    186    4.45    1.05

Variable                      Older Client         Total
                               N            M       SD      N*

Judgment of Competence
         Client Gender
                Male           97           5.11    1.07    186
                Female         89           4.75    1.22    186
         Client Health
                Healthy        94           4.93    1.23    186
                Not Healthy    92           4.93    1.08    186
         Total                186           4.93    1.16    372

Judgment of Prognosis
         Client Gender
                Male           97           4.70    0.99    186
                Female         89           4.60    1.08    186
         Client Health
                Healthy        92           4.63    1.01    186
                Not Healthy    92           4.63    1.01    186
         Total                186           4.65    1.04    372

* N = Number of vignettes

Question 1: Judgments of Client Competence

The independent variables of client age, F(1,338) = 4.59,p = [is less than].05 and client gender, F(1,338) = 10.76, p [is less than].05, had statistically significant effects on the dependent variable of judgment of competence, though the multivariate effect sizes for client age ([[Eta].sup.2] = .021) and client gender ([[Eta].sup.2] = .029) were both relatively small. Participants tended to judge older clients as somewhat less competent (M = 4.77, SD = 1.16) than younger clients (M = 5.09, SD = 1.14) as well as judging female clients (M = 4.75, SD = 1.22) somewhat less competent than male clients (M = 5.11, SD = 1.07). There were no statistically significant interactions and health status had no statistical effect on this variable.

Question 2: Judgments of Client Prognosis

The independent variable of client age had a statistically significant effect on the dependent variable of judgment of prognosis, F(1,314) = 14.43, p [is less than].05 though the multivariate effect size was relatively small ([[Eta].sup.2] = .027). This result suggests that participants tended to view prognosis somewhat more negatively for an older client (M = 4.45, SD = 1.05) than for a younger client (M = 4.86, SD = 0.98). The independent variable of gender had no statistically significant effect on judgment of prognosis nor did health status.

Question 3: Role of Professional License

The professional license of the participant had no direct effect on either judgment of competence or judgment of prognosis. However client age, client gender, and participant discipline had a statistically significant interaction effect on participants' judgment of competence, F(3,359) = 6.62, p [is less than].05, with a relatively small effect size of [[Eta].sup.2] = .034. Post hoc analysis of means for the three-way interaction of client age, client gender, and professional discipline indicated that professional clinical counselors were less likely to be affected by the interaction of the client's age and gender when judging the client's competence than psychologists or clinical social workers. In other words, clinical counselors were less likely than other professionals to view older female clients as less competent.

Question 4: Years in practice

Participants' years in practice had a statistically significant effect on the dependent variable of judgment of competence (F(1,338) = 5.6, p [is less than].05) though the multivariate effect size was small ([[Eta].sup.2] =.014). This indicated that the longer clinicians were in practice, the less competent they judged clients to be, regardless of the age, gender, or the health status of the client. Years in practice had no significant effect on judgments of prognosis by age, gender, or health.

DISCUSSION

In the vignettes used in this study the descriptions of the clients' symptoms were held constant, regardless of age or health status. There was a small but statistically significant difference in how participants rated client competency based on age and gender. This result is consistent with previous research that found that age bias influences the way mental health professionals diagnose and treat older clients (McConatha & Ebener, 1992; Morrow & Deidan, 1992; Ray et al., 1985), although the limited effect sizes suggest that these differences may be subtle rather than obvious. Nevertheless, the statistical findings of this study suggest that participants in this study were vulnerable at least to some degree to the mistaken assumption that elderly clients were less able to make decisions autonomously, a finding inconsistent with ethical standards of the profession, with the judgment of scholars in the field such as Myers (1998), developmental theory of cognitive change (Knight & McCallum, 1998), and the competencies for gerontological counseling presented by the profession (Moye & Brown, 1995). The finding that participants tended to judge female clients of any age as somewhat less competent to make autonomous decisions is troubling both in light of current ethical standards and the extensive effort that has taken place in counselor education programs over the last 30 years to educate counselors, who are able to guard against such gender bias in client interactions. This finding also gives additional credibility to the conclusions of Daniluk, Stein, and Bockus (1995) and of Stevens-Smith (1995) who have argued that the implementation of training in gender bias is inconsistent. When considered in light of the substantial body of research on the nature and effects of gender bias on females (Worell & Remer, 1992), this research further supports the need for continued attention to this problem in both training and practice settings.

Mental health professionals also tended to judge the prognosis of older clients significantly more negatively, although no gender effect was noted on this variable. These results are consistent with the literature that found that many mental health professionals do not want to work with elderly clients because they feel that these clients have poorer outcomes than younger clients (Dye, 1978; Ford & Sbordone, 1980; Gatz & Pearson, 1988; Kastenbaum, 1963; Ray et al., 1985). Clients are dependent on the professional's view of their problems and whether or not professionals believe that they can help a client function better. When professionals start with the assumption that they cannot be of much help to the client, it is likely that the counseling process will be structured around coping with day-to-day problems rather than substantive changes in the client's behavior, affect, or cognition. When this happens, the client may not be gaining the full benefit from counseling. In addition, this negative view of the client's prognosis may negatively impact the rapport between the client and counselor. Such a negtive belief stands in contradiction to the research evidence, which suggests older adults are less resistant to changing their beliefs than middle-aged adults (Visser & Krosnick, 1998) and which also suggests that counseling with older adults can be substantially similar to counseling with younger adults with similar outcomes (Knight & McCallum, 1998). Overall, these findings also show that participants' judgments were not in accordance with the ethical guidelines of the American Counseling Association (1995), American Psychological Association (1992), nor the National Association of Social Workers (1996), all of which include language which terms stereotyping based on age or gender as unethical.

Number of years as a mental health practitioner had an inverse relationship with clinicians' judgment of client competency. In other words, the longer participants had been in practice, the more likely they were to judge clients as somewhat less competent. This finding is interesting in light of the work of Kastenbaum (1963). Kastenbaum theorized that one reason mental health professionals were reluctant to treat older clients was that it was difficult for clinicians to face their own aging. Research suggests that older people view aging negatively largely as a consequence of negative stereotypes of aging (Levy & Langer, 1995; Rodin & Langer, 1980). Insofar as length of time in practice is indicative of older clinicians, their negative beliefs about their own aging might cause them to view older clients as less competent. Because training in gerontological counseling is a rather new aspect of mental health training, it is also possible that those in practice longer had not been exposed to this education. Further research into the questions is recommended to sort out these issues.

Contrary to the findings presented by James and Haley (1995) and Ford and Elliott (1999), the health status of the client did not have a significant effect on how professionals viewed client competence or prognosis. In addition, there was no interaction between age and health status. Thus, this research does not support the position that mental health professionals exhibit a negative bias towards clients in poor health. Because of the frequency with which clients with health problems seek counseling and the small number of studies in this area (Roth-Roemer & Kurpius, 1998), additional research is needed before firm conclusions can be drawn about health bias in mental health practice.

Finally, these results show that the type of license of the professional interacted with the age and gender of the client to affect to a small degree the judgment of competency of the client. The "good news" of this study is that licensed clinical counselors were least likely to judge older female clients as less competent. Further research is needed to determine if differential license requirements or training experience of professional clinical counselors, clinical social workers, and psychologists affect the fairness of their responses to clients based on age and gender.

In sum, these results point to the importance of continuing both research and professional training in age and gender bias. Professionals in this study seemed to be making important judgments about diagnosis and treatment based at least partly on factors that had little relevance to the problems for which the clients were seeking help.

There are several limitations in this research that must be considered in interpreting its findings. First, the study sampled only one region of the country and was relatively small in size. In addition, because participation in the study was voluntary, the sample must be considered as self-selected in spite of randomization procedures. Second, the study used an analogue format, and so its findings may not be comparable to the judgments of professionals presented with real clients. Third, the instrument was new, with limited, though promising data on its reliability and validity. Moreover, the instrument sampled only three major diagnostic categories from the DSM-IV. The degree to which the bias demonstrated in this study might be present with other diagnostic categories is unknown. Future research that includes a larger, national sample of mental health professionals and more clinical vignettes from a wider variety of diagnostic categories would significantly advance this line of research.

REFERENCES

American Counseling Association. (1995). Code of ethics and standards of practice. Alexandria, VA: Author.

American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author.

American Psychological Association. (1992). Ethical principles of psychologists and code of conduct. Washington, DC: Author.

Broverman, I. K., Broverman, D. M., Clarkson, F. E., Rosenkrantz, P. S., & Vogel, S. R. (1970). Sex-role stereotypes and clinical judgments of mental health. Journal of Counseling and Clinical Psychology, 34, 1-7.

Busse, W. M. (1994). Age and gender bias among mental health professionals: The effect of therapist age, client age and gender on clinical judgment. Dissertation Abstracts International, 55(4), (University Microfilms No. 9425004).

Conway, J. M, Jako, R. A., & Goodman, D. F. (1995). A meta-analysis of interrater and internal consistency reliability of selection interviews. Journal of Applied Psychology, 80, 565-579.

Daniluk, J. D., Stein, M., & Bockus, D. (1995). The ethics of inclusion: Gender as a critical component of counselor training. Counselor Education and Supervision, 34, 294-307.

Dye, C. J. (1978). Psychologists' role in the provision of mental health care for the elderly. Professional Psychology: Research and Practice, 9, 38-49.

Ford, C. V., & Sbordone, R. J. (1980). Attitudes of psychiatrists toward elderly patients. American Journal of Psychiatry, 137, 571-575.

Ford G. R. M., & Elliott, T. R. (1999). Clinicians' reactions to depressive behavior and health. Professional Psychology: Research and Practice, 30, 269-274.

Gatz, M., Karel, M. J., & Wolkenstein, B. (1991). Survey of providers of psychological services to older adults. Professional Psychology: Research and Practice, 22, 413-415.

Gatz, M., & Pearson, C. G. (1988). Ageism revised and the provision of psychological services. America Psychologist, 43, 184-188.

Gay, L. R., & Airasian, P. (2000). Educational research: Competencies for analysis and application (6th ed.). Upper Saddle River, NJ: Merrill.

Gottlieb, G. L. (1994). Barriers to care of older adults with depression. In L. S. Schneider, C. F. Reynolds, B. D. Lebowitz, & A. J. Friedhoff (Eds.), Diagnosis and treatment of depression in late life (pp. 119-131). Washington DC: American Psychiatric Association.

Hair, J. F., Anderson, R. E., Tatham, R. L., & Black, W. C. (1995). Multivariate data analysis (4th Ed.). Upper Saddle River, NJ: Prentice Hall.

James, J. W., & Haley, W. E. (1995). Age and health bias in practicing clinical psychologists. Psychology and Aging, 10, 610-616.

Kastenbaum, R. (1963). The reluctant therapist. Geriatrics, 19, 296-301.

Kirk, S. A., & Kutchins, H. (1992). The selling of the DSM: The rhetoric of science in psychiatry. New York: Aldine DeGruyter.

Klausner, E. J., & Alexopoulos, G. S. (1999). The future of psychosocial treatments for elderly patients. Psychiatric Services, 50, 1198-1204.

Knight, B. G. (1996). Psychotherapy with older adults. Thousand Oaks, CA: Sage.

Knight, B. G., & McCallum, T. J. (1998). Adapting psychotherapeutic practice of older clients: Implications of the contextual, cohort-based, maturity, specific challenge model. Professional Psychology: Research and Practice, 29, 15-22.

Lagana, L. (1995). Older adults' expectations about mental health counseling: A multivariate and discriminant analysis. International Journal of Aging and Human Development, 40, 297-316.

Lasoski, M. C. (1986). Reasons for low utilization of mental health services by the elderly. Clinical Gerontologist, 5, 1-18.

Leventhal, H., Idler, E. L., & Leventhal, E. A. (1999). The impact of chronic illness on the self system. In R. J. Contrada & R. D. Ashmore (Eds.), Self, social identity, and physical health: Interdisciplinary explorations (pp. 185-208). New York: Oxford University Press.

Levy, B., & Langer, E. (1994). Aging free from negative stereotypes: Successful memory in China and among the American deaf. Journal of Personality and Social Psychology, 66, 989-997.

Matyi, C. L., & Drevenstedt, J. (1989). Judgments of elderly and young clients as functions of gender and interview behaviors: Implications for counselors. Journal of Counseling Psychology, 36, 451-455.

McConatha, J. T., & Ebener, D. (1992). Relationship between client age and counselor trainees' perceptions of presenting problems, therapeutic methods, and prognoses. Educational Gerontology, 18, 795-802.

Molinari, V. (1996). Current approaches to therapy with elderly clients. In The Hatherleigh guide to psychotherapy (pp. 193-213). New York: Hatherleigh Press.

Morrow, K. A., & Deidan, C. T. (1992). Bias in the counseling process: How to recognize and avoid it. Journal of Counseling and Development, 70, 571-577.

Moye, J., & Brown, E. (1995). Postdoctoral training in geropsychology: Guidelines for formal programs and continuing education. Professional Psychology: Research and Practice, 26, 591-597.

Myers, J. E. (1998). Combating ageism: The rights of older persons. In C. C. Lee & G. R. Walz (Eds.), Social action: A mandate for counselors (pp. 137-160). Alexandria, VA: American Counseling Association.

O'Malley, L. M, & Richardson, S. S. (1985). Sex bias in counseling: Have things changed? Journal of Counseling and Development, 63, 294-299.

Nadien, M. (1996). Aging women: Issues of mental health and maltreatment. In J. A. Sechzer & S. M. Pfafflin (Eds.), Women and mental health (pp. 129-145). New York: New York Academy of Sciences.

National Association of Social Workers. (1996). Code of ethics. Silver Spring, MD: Author.

Nordhus, I. H., Nielson, G. H., & Kvale, G. (1998). Psychotherapy with older adults. In I. H.

Nordhus, G. R. Vandenbos, S. Berg, & P. Fromholt (Eds.), Clinical geropsychology (pp. 289-312). Washington, DC: American Psychological Association.

Raue, P. J., & Meyers, B. S. (1997). An overview of mental health services for the elderly. In L. S. Schneider (Ed.), Developments in geriatric psychiatry (pp. 3-12). San Francisco, CA: Jossey-Bass.

Ray, D. C., Raciti, M. A., & Ford, C. V. (1985). Ageism in psychiatrists: Associations with gender, certification, and theoretical orientation. The Gerontologist, 25, 497-500.

Rodin, J., & Langer, E. (1980). Aging labels: The decline of control and the fall of self-esteem. Journal of Social Issues, 16, 12-29.

Rohan, E. A., Berkman, B., Walker, S., & Holmes, W. (1994). The geriatric oncology patient: Ageism in social work practice. Journal of Gerontological Social Work, 23, 201-221.

Roth-Roemer, S., & Kurpius, S. R. (Eds.). (1998). The emerging role of counseling psychology in health care. New York: Norton.

Seem, S. R., & Johnson, E. (1998). Gender bias among counselor trainees: A study of case conceptualization. Counselor Education and Supervision, 37, 257-268.

Sherman, S. B. (1997). Images of middle-aged and older women. Historical, cultural, and personal. In J. M. Coyle (Ed.), Handbook on women and aging (pp. 14-28). Westport, CT: Greenwood.

Stevens-Smith, P. (1995). Gender issues in counselor education: Current status and challenges. Counselor Education and Supervision, 34, 283-293.

Visser, P. S., & Krosnick, J. (1998). Development of attitude strength over the life cycle: Surge and decline. Journal of Personality and Social Psychology, 75, 1389-1410.

Wadsworth, J. S. (1996). Clinical judgment bias in case management decision-making for older persons with mental retardation. Dissertation Abstracts International, 57(5), (University Microfilms No. 969502).

Welfel, E. R. (1998). Ethics in counseling and psychotherapy: Standards, research and emerging issues. Pacific Grove, CA: Brooks/Cole.

Woolfe, R., & Biggs, S. (1997). Counseling older adults: Issues and awareness. Counseling Psychology Quarterly, 10, 189-194.

Worell, J., & Remer, P. (1992). Feminist perspectives in therapy: An empowerment model for women. New York: Wiley & Sons.

Paula R. Danzinger, Ph.D., is an assistant professor, Department of Special Education and Counseling, William Paterson University, Wayne, NJ. Elizabeth Reynolds Welfel, Ph.D., is a professor, Department of Counseling, Administration, Supervision, and Adult Learning, Cleveland State University, OH. This research is based on Paula R. Danzinger's doctoral dissertation under the direction of Dr. Welfel and Frank L. O'Dell of Cleveland State University.

Correspondence concerning this article should be addressed to Dr. Danzinger, William Paterson University, 300 Pompton Rd., Raubinger Hall 416, Wayne, NJ 07470; Email [e-mail]