Factors affecting clergy-psychologist referral patterns.

Recent research on clergy-psychologist collaboration has resulted in helpful principles for clergy and psychologists working together, but very little is known about what specific characteristics in clergy are appealing to psychologists and vice versa. Two experimental survey studies are reported, both exploring characteristics that enhance or hinder collaboration. In Study 1, Southern Baptist pastors rated the likelihood of referring to a counselor who was identified as either a Biblical Counselor or a Christian Psychologist, and was identified as either being excellent in interpersonal skills or as using scripture and prayer in counseling. Pastors demonstrated a preference for counselors using scripture and prayer in counseling over those described as having excellent interpersonal skills. In Study 2, psychologist respondents rated the likelihood of referring to a clergyperson who was identified either as a Baptist or a Unitarian universalist, and was identified either as seminary trained or as having a doctoral degree from a prestigious divinity school. Psychologists demonstrated a preference for those with a doctoral degree. The weak effect sizes and comments offered on the surveys suggest that the individual relationship between clergy and psychologist is much more salient than particular demographic characteristics when considering the possibility of collaboration.

**********

Psychologists have reported very little systematic research on collaborating with clergy (Weaver et al., 1997), but this may be starting to change. Over the past decade various survey and exemplar studies have been reported (Benes, Walsh, McMinn, Dominguez, & Aikins, 2000; Budd, 1999; Chaddock & McMinn, 1999; Edwards, Lim, McMinn, & Dominguez, 1999; McMinn, Aikins, & Lish, 2003; McMinn, Chaddock, Edwards, Lim, & Campbell, 1998; McMinn, Meek, Canning, & Pozzi, 2001; McRay, McMinn, Wrightsman, Burnett, & Ho, 2001; Plante, 1999), and a recent special issue of Journal of Psychology and Christianity was devoted to the topic (McMinn & Dominguez, 2003). From this recent research, we now know various principles for effective collaboration (McMinn, Aikins, & Lish, 2003) and have learned from those who collaborate in their day-to-day professional activities (McMinn & Dominguez, 2005). However, we still do not know what specific characteristics of clergy and psychologists enhance or diminish the possibility of collaboration.

McMinn et al. (2005) used an experimental method to see what psychologist characteristics might influence clergy willingness to collaborate. They did not find any difference based on the sex of the psychologist or the type of institution at which the psychologist was trained, but there were three substantial methodological limitations that may have weakened the findings. First, a heterogeneous group of clergy from various Christian denominations was selected, which added to error variance and detracted from the study's power. Second, the clergy respondents were asked how likely they would be to collaborate in innovative ways, such as consulting with a psychologist for a new staff hire or to deal with conflict in the church. These innovative forms of collaboration appeared to be quite unfamiliar to the clergy respondents, as evidenced by their written comments on the survey. The vast majority of written comments pertained to traditional types of collaboration, such as referring a troubled parishioner for psychological services. Third, the selection and visual display of the independent variables may have lacked salience, thereby contributing to relatively modest power in the experimental design.

The first study described here was designed to eliminate these methodological problems by (a) selecting clergy from a single denomination, (b) having respondents rate the likelihood of referring a parishioner to the mental health professional, and by (c) choosing more salient independent variables. The second study was an attempt to look at a similar question from a reverse angle: How likely are psychologists to refer clients to clergy, and what factors enhance or inhibit these referral practices?

STUDY 1

We know from previous research that clergy are often sought in times of emotional difficulty, and so they function as front-line mental health workers (Weaver, Flannelly, Garbarino, Figley, & Flannelly, 2003). In some situations clergy refer troubled individuals on to a psychologist or other mental health professional, though this appears to be the exception rather than the rule for most clergy (Meylink & Gorsuch, 1988). Conservative clergy are relatively less likely to refer troubled parishioners to a mental health professional than liberal clergy (Mannon, & Crawford, 1996). Study 1 is an effort to identify the specific counselor characteristics that may promote referrals from conservative clergy.

Method

Participants. Participants recruited were clergy in the Southern Baptist Convention (SBC), which is a conservative tradition within Christianity (Smith, 1990). Through the website http://www.sbc.net, two hundred pastors were selected. In order to produce a diverse national sample, we used a quota sampling method selecting 40 names each from five predetermined regions of the United States. Selections were made by choosing large cities within each region and accepting participants within a 100-mile radius.

Instrument. The front side of the questionnaire included information about a hypothetical psychologist, Dr. Pat Johnson, using the following narrative:</p> <pre> Meet Dr. Pat Johnson. Please note that he is a fictional person, made up for the purposes of this research. We are not marketing any services. We are interested in knowing how you would feel about referring a troubled person to someone like Dr. Johnson. </pre> <p>Following this brief introduction, participants viewed a descriptive profile of Dr. Johnson indicating a professional identity and a counseling method. The possible identities were as follows:

Dr. Johnson is a Biblical Counselor: He emphasizes the importance of faith in his counseling work, and he is a member of the American Association of Christian Counselors.

Dr. Johnson is a Christian psychologist: He recognizes the importance of faith in his psychotherapy work, and he is a member of the Christian Association of Psychological Studies.

The two possible methods of counseling were as follows:

Dr. Johnson uses Scripture and prayer in his counseling sessions: He understands the causal role of sin in human problems, and recognizes that healing comes through the power of God.

Dr. Johnson has excellent interpersonal and counseling skills: He uses proven counseling methods in order to facilitate healing in human problems.

There were four variations of this questionnaire. Each questionnaire asked the participant to rate his or her likelihood to refer to Dr. Johnson based on the information provided. This study utilized a 2 x 2 factorial design with the independent variables being Dr. Johnson's professional identity and his method of counseling.

Along with the independent variables on the front page, five referral situations were presented, including depression, addiction, schizophrenia, sexual abuse, and relationship problems. A sixth referral situation involved whether or not the respondent needed help for a personal problem in life. For each of these items, respondents were asked to rate their likelihood of contacting Dr. Johnson on a 5-point Likert scale ranging from 1 (Definitely No) to 5 (Definitely Yes). At the bottom of the front page, space was provided for written comments from the participants, asking them to describe the factors that influenced their answers.

On the back page of the questionnaire, participants were asked to rate how important certain factors were in considering whether to make a referral to a professional counselor. All respondents were given the same seven factors, including whether the counselor (a) is a Christian, (b) has formal theological training, (c) shares participant's theological beliefs, (d) has formal psychological training, (e) uses scripture and prayer in counseling, (f) has a reasonable fee, and (g) keeps the pastor informed about the parishioner's progress in counseling. Each factor was rated on a 5-point Likert scale ranging from 1 (Not at all important) to 5 (Extremely important). Along with the rating scales, participants were asked for demographic information and any additional comments.

Procedure. Questionnaires were mailed to participants in January 2005 along with $3 as incentive for responding. Participants were randomly assigned to receive one of four possible variations of the questionnaire. Of the 200 potential participants, 13 questionnaires were returned as undeliverable and 113 responded, resulting in a 60.4% response rate. There was a relatively even distribution across all four variations of the questionnaire, [chi square] (3, N = 113) for goodness of fit = 0.70, ns.

Results

Of the 113 respondents, 110 (97.3%) were male, 1 (0.9%) was female, and 2 (1.8%) did not report sex. Respondents were largely European American (77.9%), followed by Asian American (7.1%), African American (5.3%), and Native American (0.9%), with 1.8% reporting being of "Other" descent. Eight respondents (7.1%) did not report their ethnicity. The ages of the respondents ranged from 26 to 69 years, with an average age of 47 years. The years they had been in ministry ranged from 2 to 42 years, with an average of 22 years. Reported weekly church attendance averaged 376 parishioners with a range of 13 to 4,000. Respondents' level of education was as follows: Masters of Divinity (23.9%), Doctorate of Ministry (21.2%), Other Masters (18.6%), Other Doctorate (15.9%), Undergraduate (9.7%), Associates or some college (2.7%), High School (2.7%), and Other (0.9%). Five respondents (4.4%) did not report their level of education. The majority of respondents reported being senior or solo pastors (94.7%), followed by other pastors (2.7%), and assistant or associate pastors (0.9%).

As described previously, each survey instrument had one of two descriptors describing the identity of the mental health provider as either a Christian Psychologist or a Biblical Counselor. We will refer to this independent variable as Identity. The second independent variable will be called Counseling Method: half the respondents received surveys describing the counselor as having excellent interpersonal skills, and half received surveys stating that the counselor uses scripture and prayer in counseling. The dependent variables included a list of 6 scenarios that clergy might face: someone you know is depressed, someone you know needs help for an addiction problem, someone you know needs help for schizophrenia, someone you know is recovering from sexual abuse, a couple you know needs help for their relationship, you yourself need help for a problem in life. Each of these dependent variables was rated on a 5-point Likert scale ranging from 1 (Definitely would not refer to Dr. Johnson) to 5 (Definitely would refer to Dr. Johnson).

Because we were not looking for any particular relationship among the six dependent variables, we opted not to use a multivariate analysis of variance (MANOVA). We first tested to see if response patterns on the six items were inter-related. Both a test of internal consistency (Cronbach's [alpha] = .92) and a principle component factor analysis where all items loaded on a single factor, accounting for 73% of the variance, suggested that these six items belonged to a common construct. Thus, we added clergy ratings on the six items to derive an overall Likelihood of Referral score.

The results of a 2 x 2 analysis of variance (ANOVA) are presented in Table 1. A main effect for Counseling Method was observed, F (1, 107) = 4.8, p < .05, with preference shown to counselors who use scripture and prayer in their sessions. No main effect for Identity or interaction effects were observed. Given the main effect observed for the overall Likelihood of Referral score, we proceeded with a series of 2 x 2 ANOVAs looking at each of the six dependent variables. Given the exploratory nature of the study, we retained an [alpha] of .05 though recognizing that the multiple hypothesis tests inflate the possibility of Type I error. No main effects or interaction effects were observed for referring a depressed client or a schizophrenic client. Two main effects were observed when the clergyperson perceived a need for personal help; they preferred a Biblical Counselor, F (1, 107) = 4.6, p < .05, who uses scripture and prayer in session, F (1, 107) = 5.0, p < .05. Similarly, both main effects were observed for referring those with relationship problems; clergy preferred a Biblical Counselor, F (1, 108) = 7.9, p < .01, who uses scripture and prayer in session, F (1, 108) = 4.3, p < .05. Clergy preferred a counselor who used scripture and prayer in session when referring a client recovering from addiction, F (1, 108) = 4.9, p < .05, or sexual abuse, F (1, 108) = 5.7, p < .05. No interaction effects were found in any of the analyses.

Next, we compared the ratings on the six dependent variables to see which situation is most likely to result in a referral to a mental health professional. We looked for overall differences using a repeated-measures MANOVA. The differences, which can be seen in Table 2, were significant, Wilks' [lambda] (6, 105) = .054, p < .001. The overall differences justified a profile analysis, by which adjacent means are compared with paired-sample t tests. Item differences are reported in Table 2.

We also asked clergy respondents to identify the factors that they consider when looking for a counselor to whom they can refer parishioners. All respondents were given the same list of factors, listed in Table 3. We used a repeated-measures MANOVA to test for interitem differences, Wilks' [lambda] (7,101) = .010, p < .001, followed by a profile analysis. Item differences are identified in Table 3.

Next, we computed several additional analyses to see how these findings compared with previous reports. Previous researchers have reported that factors such as age, level of education, denominational affiliation and congregational size may affect collaboration (Shabazz, 2003). More specifically, clergy age 40 years and under were more likely to refer people for counseling problems than clergy over the age of 40 years. These findings were not replicated in the present study: We found no significant differences in overall likelihood of referring between clergy over age 40 and those 40 and younger. Also, previous research has suggested that clergy with more education are more likely to refer than minimally trained or untrained clergy (Thomas, Quinn, Billingsley, & Caldwell, 1994; Fultz, 2002). Again, we did not find support for any relationship between education and likelihood of referring in the current study, r = 0.05, ns. Finally, previous research has suggested clergy from larger congregations are more likely to refer than clergy from smaller congregations (Mannon, & Crawford, 1996). This idea was not supported by the current study, r = .09, not significant. Our failure to replicate findings from previous studies may be related to the relatively small sample size in this study or to the homogeneity of our respondents.

Finally, we computed a multiple regression in an attempt to predict likelihood of referring with the predictor variables hypothesized to be meaningful in this study and those found to be meaningful in previous studies: method (scripture/prayer or interpersonal skills), identity (biblical counselor or Christian psychologist), size of congregation, age, and education. We found only 1 significant predictor, method (scripture/prayer vs. interpersonal skills), and it only accounted for 5% of the variance.

STUDY 2

Books published recently by the American Psychological Association highlight the importance of focusing on the spiritual needs of clients (Miller, 1999; Miller & Delaney, 2005; Richards & Bergin, 1997, 2000, 2004; Shafranske, 1996; Sperry & Shafranske, 2005). Involved in the renewed concern with client spirituality is the possibility of referring

clients to non-psychologists or clergy for spiritual or religious guidance. Furthermore, the current economics of mental health care may sometimes prompt referrals to clergy and others who provide pro bono services.

Method

Participants. Participants were psychologists recruited from http://www.superpages.com and included a sample of individuals from across the United States. Questionnaires were sent to 200 psychologists throughout the United States. In order to obtain a diverse sample, 40 psychologists were selected from five predetermined geographic regions. Selections were made by choosing large cities within each region and selecting participants within a 100-mile radius of the city.

Instrument. Participants were asked to fill out a brief questionnaire about a hypothetical minister named "Pat Johnson". All questionnaires began with the following introduction:</p> <pre> Meet Pat Johnson, a minister. Please note that he is a fictional person, made up for the purposes of this research. We are not marketing any services. We are interested in knowing how you would feel about referring someone with spiritual questions to Pat Johnson. </pre> <p>Subsequently, participants viewed a descriptive profile of Pat Johnson in which information was manipulated. There were four variations of this questionnaire. The possible educational levels were as follows:

Pat Johnson is seminary trained: He received His Masters in Divinity from a prominent seminary where he was trained in theology and pastoral counseling.

Pat Johnson is highly educated: He received a doctoral degree in theological studies from a world-renowned divinity school, and is an ordained minister with training in pastoral counseling.

The liberal/conservative affiliations were as follows:

Pat Johnson is the minister of a Unitarian Church: Unitarian Universalism is a liberal religion that keeps an open mind to the spiritual questions people have struggled with in all times and places.

Pat Johnson is the minister of a Baptist Church: Baptists hold firm to a confession of faith in Jesus Christ, pledging faithfulness to the doctrines revealed in the Bible.

This study utilized a 2 x 2 factorial design. The independent variables were Pat Johnson's level of education and his liberal/conservative affiliation. Participants completed 1 of 4 possible variations of the questionnaire. On the front page, participants were also presented with six hypothetical referral situations and were asked to rate their likelihood of referring to Pat Johnson in each situation. The first four hypothetical situations involved client referral and included: whether the client has spiritual questions, has been wounded by a religious organization in the past, has no social support network and cannot afford to see a psychologist. Questions 5 and 6 asked psychologists about self-referral and specifically asked them to rate their likelihood of seeking answers for spiritual questions and problems in life from the aforementioned minister. Responses were rated on a 5-point Likert scale ranging from 1 (definitely no) to 5 (definitely yes). At the bottom of the front page, participants were given the opportunity to describe factors that influenced their answers.

The back of the questionnaire included questions about the factors involved in referring to ministers. Respondents were asked to rate the importance of six factors on a Likert scale ranging from 1 (not at all important) to 5 (extremely important). The six factors included: whether the minister is highly educated, has formal counseling training, shares the participant's beliefs, demonstrates respect for psychology, and has an open and inclusive view of religious faith. Participants were also provided with an "Other" category in which to write in another factor and rate its importance. Furthermore, participants were asked brief demographic questions, and asked to rate the importance of their religion on a 5-point Likert scale. The final item on the questionnaire gave participants the opportunity to offer additional comments.

Procedure. Questionnaires and a $2 response incentive were mailed to participants in January of 2005 (the incentive was slightly lower than in Study 1, both for budgetary reasons and because our previous research suggests that psychologists are more likely than clergy to respond to surveys). Participants were randomly assigned to receive one of four possible variations of the questionnaire. Of the 200 potential participants, 18 questionnaires were returned as undeliverable whereas 120 responded, resulting in a 65.9% response rate. There was a relatively even distribution across all four variations of the questionnaire, [chi square] (3, N = 120) for goodness of fit = 0.50, ns.

Results

Of the 120 respondents, 45.8% were male, 51.7% were female, and 2.5% did not report sex. Respondents were largely European American (84.2%), followed by Latino/Hispanic (3.3%), Asian American (2.5%), other (2.5%) and African American (.8%). Notably, 6.7% of participants did not report their ethnicity. The participants ranged in age from 26 to 82 years of age with an average age of 51. The majority of respondents (73.3%) held a Ph.D., followed by 16.7% with a Psy.D., and 1.7% with an Ed.D. A small number of participants (2.5%) did not report degree title. Participants' number of years in psychology ranged from 1 to 44, with a mean of 20.

As with the clergy survey described in Study 1, the psychologist survey had two independent variables and six dependent variables in the 2 x 2 design. The first independent variable, Education, had one of two descriptors on each survey: Pat Johnson is highly educated with a doctoral degree from a world-renowned divinity school, or Pat Johnson is seminary trained with a masters of divinity from a prominent seminary. The second independent variable, Liberal/Conservative Affiliation, indicated that Pat Johnson was either the pastor of a Baptist church or a Unitarian Universalist church. The dependent variables included a list of six scenarios that might call for referral to a clergyperson, each of which was rated on a 5-point Likert scale ranging from 1 (definitely would not refer) to 5 (definitely would refer).

As with Study 1, we were not looking for any particular relationship among the six dependent variables. Thus, we opted not to use a multivariate analysis of variance (MANOVA), but tested to see if response patterns on the six items were inter-related. A test of internal consistency (Cronbach's [alpha] = .87) and a principle component factor analysis--where all items loaded on a single factor accounting for 61% of the variance--suggested that these six items belonged to a common construct. Thus, we summed psychologist ratings on the six items to derive an overall Likelihood of Referral score.

Likelihood of Referral scores for each condition in the 2 x 2 ANOVA design are shown in Table 4. A main effect was observed for Education, F (1, 108) = 6.0, p < .05, but not for Liberal/Conservative Affiliation. No interaction effects were observed. We then performed a series of 2 x 2 ANOVAs, using each of the six dependent variables in separate analyses. We again used an [alpha] of .05 in light of the exploratory nature of the study. Psychologists were more likely to refer to Unitarians than Baptists when a client had a spiritual question, F (1, 114) = 8.5, p < .01. Main effects were observed for both Education and Liberal/Conservative Affiliation when a client had been wounded by religion in the past, with psychologists preferring Unitarian clergy, F (1, 114) = 7.6, p < .01, and those with doctoral degrees, F (1, 112) = 6.7, p < .05. When a client indicated having a poor support network, psychologists preferred to refer to a highly educated clergyperson, F (1, 115) = 10.5, p < .01, but no main effect was observed for Liberal/Conservative Affiliation. The same was true for a depressed client who could not afford a psychologist, F (1, 113) = 4.8, p < .05. No main effects were observed when psychologists had spiritual questions or needed help with a personal problem. No interaction effects were found in any of the analyses.

We then compared the ratings on the six referral scenarios to see when psychologists perceive themselves to be most likely to refer to clergy. Overall differences--shown in Table 5--were detected with a repeated-measures MANOVA, Wilks' [lambda] (6, 106) = .115, p < .001. A profile analysis was then conducted to compare adjacent means (see Table 5).

We asked psychologists to rate several factors they may consider when looking for a clergyperson to whom they can refer clients. All respondents were given the same list of factors, listed in Table 6. We used a repeated-measures MANOVA to test for interitem differences, Wilks' [lambda] (5, 106) = .025, p < .001, and then followed it with a profile analysis. Item differences are identified in Table 6.

Next, we investigated the possibility of demographic variables' relation to likelihood of referring to a clergyperson. No significant correlations were observed between referral likelihood and various demographic variables (age, years in practice, and importance of personal religious values). Similarly, men and women did not differ in their likelihood of referring to a clergyperson.

Finally, we computed a multiple regression in an attempt to predict likelihood of referring with the following predictor variables: education (doctoral degree or recipient of seminary training), church (Unitarian or Baptist), age, sex, and personal importance of religion. We found only one significant predictor, education, and it only accounted for 5% of the model's variance.

QUALITATIVE ANALYSES

In addition to the quantitative analyses described in Studies 1 and 2, we also collected and coded all written comments on the questionnaires, using QSR N6 software. Participants were asked to give their reasons for referring to Pat Johnson or not, and they were also given the opportunity for open-ended commentary at the end of the questionnaire. A single coder was used for purposes of content analysis (i.e., categorizing and tallying the various comments from clergy and psychologist respondents).

The most dominant themes for both groups of respondents were suspicion and values-related tensions. A total of 36 clergy respondents emphasized the importance of the psychologist's religious faith and practice when deciding about a referral, and many voiced suspicion about the worldview they associate with psychology. For example, one clergyperson wrote, "So much of psychological training seems to undermine God's role." Conversely, 18 psychologist respondents expressed concerns about the ideological narrowness of Baptist pastors and suspicion of clergy in general. One psychologist wrote, "Why can't ministers stick to their own training and business. If they want to do psychotherapy let them get a graduate degree at an accredited program like everyone else. These guys are walking time bombs." Most respondents were gentler in their appraisal, but some level of suspicion toward clergy was a common theme.

Both clergy (n = 14) and psychologists (n = 18) emphasized the importance of relationship in deciding whether or not to refer to Pat Johnson. The scant information they were given was inadequate--many expressed a need to have a prior relationship before deciding whether or not to collaborate with Pat Johnson. For example, "I usually only refer people to those counselors I know personally or those who have been referred to me by someone I highly respect and value."

Psychologists (n = 18) noted the importance of matching their client's religious values with the values of the pastor. Many noted that they might refer to a Baptist minister if the client were Baptist. For example, "I prefer to work with clients within their own spiritual framework and I am happy to collaborate with clergy who already work with them."

DISCUSSION

An earlier study by McMinn et al. (2005) used a similar methodology, but: the sample of clergy was more heterogeneous, the independent variables were not as clear and distinct in the questionnaire, and unfamiliar forms of collaboration were presented to clergy as dependent variables. We attempted to correct for these methodological problems in the current study to see if particular characteristics of psychologists influence clergypersons' decisions to refer. We added a parallel study to see if particular clergy characteristics influence psychologists' willingness to refer.

Still, there are limitations to the studies reported here. Sample sizes for each study were relatively small, with some cells in the 2 x 2 designs having as few as 26 respondents. Study 1 is difficult to generalize because we selected clergy from only one denomination. With this choice, we sacrificed external validity for the sake of internal validity; the homogeneity of the sample helped reduce error variance, but it also makes generalizing the results problematic. Finally, the statistical power of the independent variables should be questioned. For example, to what extent do clergy recognize distinctions between Christian psychologists and biblical counselors? What seems salient to those involved in the integration of psychology and Christianity may not be as important to those not involved in integration.

In each study, significant main effects were discovered. Southern Baptist clergy prefer counselors who use spiritual techniques (such as prayer and scripture in counseling) over counselors who are identified as having excellent interpersonal skills. They also prefer the identity of Biblical Counselor over Christian Psychologist for some particular issues (i.e., referring a couple with relationship problems, and if the clergyperson needs personal help). Psychologists prefer to work with highly educated clergy--those with doctoral degrees--as compared with seminary-trained clergy. They also prefer Unitarian clergy over Baptist clergy when a client has spiritual questions or has been wounded by a religious organization in the past. Though these findings are statistically significant, they remain mostly uninteresting because the magnitude of the relationships among these variables is quite modest. These statistically significant differences have effect sizes (d) on the order of 0.4 to 0.5, which indicates a moderate effect size at best. Moreover, multiple regression analyses failed to find any more powerful predictors, even among demographic variables that have been shown to be connected with referral patterns in previous studies.

Thus, it seems that neither clergy nor psychologists are looking for collaborators that fit in neat demographic categories or go by particular labels. The comments offered suggest a more nuanced approach. An established, trusting relationship with the other professional seems particularly important. Perhaps one reason this is so important is related to the mistrust that is evident among both clergy and psychologists. A good deal of tension and unease is evident in written comments and by the low likelihood of referral ratings offered by psychologists. Some psychologists are willing to refer to some clergy, especially if the clergyperson's values align with the clients. Southern Baptist clergy are often willing to refer parishioners to a counselor or psychologist, but only if the psychologists' values align with the values of the church.

An important implication for those involved in training doctoral students in psychology or clergy is to consider ways of enhancing contact between psychology and religious professionals during training. If collaboration is based more on relationship than demographic categories, then modeling cross-disciplinary relationships in training is important. This might take the form of inviting a clergyperson to give a special lecture in a psychology training program, or vice versa. It might also involve helping students learn the language of the other discipline. For example, psychology students might be encouraged to read some theology, seminarians might be encouraged to read some psychology, and both groups could be encouraged to interact with religious leaders and psychologists about these readings. Team-taught classes can also be a useful way to both model collaboration and enhance dialog between religious and mental health professionals (see Edwards et al., 1999).

With regard to psychological practice, conservative clergy may sometimes respond more favorably to psychologists who use overt spiritual practices in their work and those who call themselves biblical counselors: Psychologists prefer to collaborate with highly educated and theologically liberal clergy in some situations. But the effect sizes in these studies are so modest that the more important implication is how little psychologists and clergy are influenced by titles and spiritual practices. Relationships between psychologists and clergy are essential for effective collaboration, and these relationships are ultimately more important than specific variables that can be described on a questionnaire. Meaningful collaboration occurs in the context of relationships characterized by trust, common values and goals, mutual respect, and communication (McMinn et al., 2003). Relationships between clergy and psychologists can be enhanced by psychologists volunteering time to present seminars at a local church, offering no-cost or low-cost consultations to pastors, inviting pastors to psychologists' staff meetings to consider topics of psychological and theological importance, and so on.

Regarding research implications, most of the previous research on clergy-psychologist collaboration has involved either descriptive survey research or the in-depth qualitative study of exemplars, and very little has been done with an experimental methodology. The two studies reported here, and a previous study (McMinn et al., 2005), have combined a survey and experimental methodology in an effort to determine which psychologists and clergy characteristics are most appealing for purposes of collaboration. None of these experimental studies have yielded robust conclusions. Either professionals develop collaborative relationships for reasons not easily captured in a questionnaire, or the studies have lacked statistical power, or both. Unless experimental methods can be adapted to much more practical and real-to-life scenarios, the future of experimental research in clergy-psychologist collaboration seems rather bleak. It may be better to pursue the other methods of exemplar research and descriptive studies.

CONCLUSION

A sign posted in a riverside construction area reads: "Bridge Out. Local Traffic Only." Though this gives a puzzling message as a construction sign, it aptly reflects the state of clergy-psychologist collaboration, at least when conservative Christian clergy are trying to collaborate with psychologists. The bridge is "out": many values conflict and tensions create suspicion and distrust. But the road may still be open to local traffic--those who have built a trusting collaborative alliance through mutual respect and communication. One cannot become "local traffic" by fitting any of the demographic categories we have studied here, but only by the time-honored work of building an effective relationship.

REFERENCES

Benes, K. M., Walsh, J. M., McMinn, M. R., Dominguez, A. W., & Aikins, D. C. (2000). Psychology and the church: An exemplar of psychology-clergy collaboration. Professional Psychology: Research and Practice, 31, 515-520.

Budd, F. C. (1999). An Air Force model of psychologist-chaplain collaboration. Professional Psychology: Research and Practice, 30, 552-556.

Chaddock, T. P., & McMinn, M. R. (1999). Values affecting collaboration among psychologists and evangelical clergy. Journal of Psychology and Theology, 27, 319-328.

Edwards, L. C., Lim, R. K. B., McMinn, M. R., & Dominguez, A. W. (1999). Examples of collaboration between psychologists and clergy. Professional Psychology: Research and Practice, 30, 547-551.

Fultz, S. A. (2002). Clergy referral for counseling problems. Dissertation Abstracts International: Section B: The Sciences & Engineering, 62(8-B), 3799.

Mannon, J. D., & Crawford, R. L. (1996). Clergy confidence to counsel and their willingness to refer to mental health professionals. Family Therapy, 23, 213-231.

McMinn, M. R., Aikins, D. C., & Lish, R. A. (2003). Basic and advanced competence in collaborating with clergy. Professional Psychology: Research and Practice, 34, 197-202.

McMinn, M. R., Ammons, J., McLaughlin, B. R., Williamson, C., Griffin, J. W., Fitzsimmons, C. R., et al. (2005). Clergy responses to psychologist characteristics. In M. R. McMinn & A. W. Dominguez (Eds.), Psychology and the church (pp. 9-17). Hauppauge, NY: Nova Science Publishers.

McMinn, M. R., Chaddock, T. P., Edwards, L. C., Lim, R. K. B., & Campbell, C. D. (1998). Psychologists collaborating with clergy: Survey findings and implications. Professional Psychology: Research and Practice, 29, 564-570.

McMinn, M. R., & Dominguez, A. W. (Eds.) (2003). Psychology and the church [Special issue]. Journal of Psychology and Christianity, 22(4), xxx-xxx.

McMinn, M. R., & Dominguez, A. W. (Eds.) (2005). Psychology and the church. Hauppauge, NY: Nova Science Publishers.

McMinn, M. R., Meek, K. R., Canning, S. S., & Pozzi, C. F. (2001). Training psychologists to work with religious organizations: The Center for Church-Psychology Collaboration. Professional Psychology: Research and Practice, 32, 324-328.

McRay, B. W., McMinn, M. R., Wrightsman, K., Burnett, T. D., & Ho, S. D. (2001). What evangelical pastors want to know about psychology. Journal of Psychology and Theology, 29, 99-105.

Mcylink, W. D., & Gorsuch, R. L. (1988). Relationship between clergy and psychologists: The empirical data. Journal of Psychology and Christianity, 7(1), 56-72.

Miller, W. R. (Ed.) (1999). Integrating spirituality into treatment: Resources for practitioners. Washington, DC: American Psychological Association.

Miller, W. R., & Delaney, H. D. (Eds.) (2005). Judeo-Christian perspectives on psychology: Human nature, motivation, and change. Washington, DC: American Psychological Association.

Plante, T. G. (1999). A collaborative relationship between professional psychology and the Roman Catholic Church: A case example and suggested principles for success. Professional Psychology: Research and Practice, 30, 541-546.

Richards, P. S., & Bergin, A. E. (1997). A spiritual strategy for counseling and psychotherapy. Washington, DC: American Psychological Association.

Richards, P. S., & Bergin, A. E. (Eds.) (2000). Handbook of psychotherapy and religious diversity. Washington, DC: American Psychological Association.

Richards, P. S., & Bergin, A. E. (Eds.) (2004). Casebook for a spiritual strategy for counseling and psychotherapy. Washington, DC: American Psychological Association.

Shabazz, B. F. (2003). African-American clergy receptivity to relationship-building with Christian psychologists. Dissertation Abstracts International: Section B: The Sciences & Engineering, 63(8-B), 3938.

Shafranske, E. P. (1996). Religion and the clinical practice of psychology. Washington, DC: American Psychological Association.

Smith, T. W. (1990). Classifying Protestant denominations. Review of Religious Research, 31, 225-245.

Sperry, L., & Shafranske, E. P. (2005). Spiritually oriented psychotherapy. Washington, DC: American Psychological Association.

Thomas, F. B., Quinn, S. C., Billingsley, A., & Caldwell, C. H. (1994). The characteristics of northern black churches with community health outreach programs. American Journal of Public Health, 84, 575-579.

Weaver, A. J., Flannelly, L. T., Garbarino, J., Figley, C. R. & Flannelly, K. J. (2003). A systematic review of research on religion and spirituality in the Journal of Traumatic Stress: 1990-1999. Mental Health, Religion & Culture, 6, 215-228.

Weaver, A. J., Samford, J. A., Kline, A. E., Lucas, L. A., Larson, D. B., & Koenig, H. G. (1997). What do psychologists know about working with the clergy? An analysis of eight APA Journals: 1991-1994. Professional Psychology: Research and Practice, 28, 471-474.

AUTHORS

McMINN, MARK R.: Address: Department of Psychology, Wheaton College, 501 College Avenue, Wheaton, IL 60187. Title: Rech Professor of Psychology. Degrees: Ph.D. (clinical psychology), Vanderbilt University. Specializations: Church-psychology collaboration, cognitive therapy, integration of psychology and Christianity.

RUNNER, STEVEN J.: Address: Department of Psychology, Wheaton College, 501 College Avenue, Wheaton, IL 60187. Title: Doctoral student in clinical psychology. Degrees: M.A. (clinical psychology), Wheaton College.

FAIRCHILD, JENNIFER A.: Address: Department of Psychology, Wheaton College, 501 College Avenue, Wheaton, IL 60187. Title: Doctoral student in clinical psychology. Degrees: M.S. (clinical psychology), Emporia State University.

LEFLER, JOSHUA D.: Address: Department of Psychology, Wheaton College, 501 College Avenue, Wheaton, IL 60187. Title: Doctoral student in clinical psychology. Degrees: M.A. (clinical psychology), Wheaton College.

SUNTAY, RACHEL P.: Address: Department of Psychology, Wheaton College, 501 College Avenue, Wheaton, IL 60187. Title: Doctoral student in clinical psychology. Degrees: M.S. (general psychology), DePaul University.

MARK R. MCMINN, STEVEN J. RUNNER, JENNIFER A. FAIRCHILD, JOSHUA D. LEFLER, and RACHEL P. SUNTAY

Center for Church-Psychology Collaboration

Wheaton College

This research was funded in large part through the generous support of Joel S. Milner, PhD, director of the Center for the Study of Family Violence and Sexual Assualt at Northern Illinois University. Capable assistance with data collection was provided by Jennifer Guimond, Patricia Martens, Jeff Martens, Paul Rodriguez, and Jason Jiandani. Correspondence concerning this article may be sent to Christopher Dyslin, Ph.D., L.P., Dept. of Intercultural Studies and Social Science, Crown College, 8700 College View Dr., St. Bonifacius, MN 55375. Email: [e-mail] Portions of this research were funded by a grant from the G. W. Aldeen Memorial Fund at Wheaton College (IL). Correspondence regarding this article may be sent to Mark R. McMinn, Department of Psychology, Wheaton College, 501 College Avenue, Wheaton, IL 60187. Email: M. After July 1, 2006, correspondence should be directed to Mark R. McMinn, George Fox University, Department of Psychology, 414 N. Meridian St., Newberg, OR, 97132.

TABLE 1 Overall Likelihood of Referral Scores for Clergy in Each of the
Four Conditions

                                    Counseling Method
Identity                Scripture and Prayer  Interpersonal Skills

Biblical Counselor      24.2 (5.4)            22.6 (5.2)(5.4)
                        n = 26                n = 27
Christian Psychologist  23.0 (6.0)            19.9 (5.9)
                        n = 32                n = 26

Note. Scores are reported as means, with standard deviations in
parentheses. Scores are the sums of six items, each rated on a 5-point
Likert scale. A 2 x 2 ANOVA revealed a main effect for Counseling
Methods, F (1, 107) = 4.8, p < .05.

TABLE 2 Clergy Ratings on Likelihood of Collaboration

Referral Ratings Scores

Someone you know is recovering from sexual abuse         4.0 (1.1)
Someone you know needs help for schizophrenia            3.9 (1.1)
Someone you know needs help for an addiction problem     3.8 (1.1)
Someone you know needs help for depression               3.8 (1.0)
A couple you know needs help for their relationship (a)  3.6 (1.2)
You yourself need help (b)                               3.4 (1.2)

Note. Ratings are reported as means (standard deviations). All items are
rated on a 5-point Likert scale, ranging from 1 (definitely would not
refer) to 5 (definitely would refer).
(a) = this item is significantly lower than the previous item, p < .05
(b) = this item is significantly lower than the previous item, p < .01

Table 3 Clergy Importance Ratings for Factors Related to Making a
Referral

Referral Ratings Scores

The counselor is a Christian                                   4.8 (0.6)
The counselor uses scripture and prayer in counseling (c)      4.6 (0.7)
The counselor has a reasonable fee (c)                         4.1 (0.9)
The counselor has formal psychological training                4.0 (1.1)
The counselors shares your theological beliefs                 4.0 (0.9)
The counselor has formal theological training (c)              3.6 (0.9)
The counselor keeps you informed about progress in counseling  3.3 (1.1)

Note. Ratings are reported as means (standard deviations). All items are
rated on a 5-point Likert scale, ranging from 1 (not at all important)
to 5 (extremely important).
(c) = this item is significantly lower than the previous item, p < .001.

TABLE 4 Overall Likelihood of Referral Scores for Psychologists

Liberal/Conservative Affiliation
Education         Unitarian   Baptist

Doctoral Degree   16.7 (5.3)  14.9 (6.2)
                  n = 32      n = 26
Seminary Trained  14.4 (5.2)  12.1 (5.5)
                  n = 27      n = 27

Note. Scores are reported as means, with standard deviations in
parentheses. Scores are the sums of six items, each rated on a 5-point
Likert scale. A 2 x 2 ANOVA revealed a main effect for Education,
F (1, 108) = 6.0, p < .05.

TABLE 5 Psychologist Ratings on Likelihood of Collaboration

Referral Ratings Scores

Client has spiritual questions                   2.9 (1.1)
Client has no support network                    2.8 (1.2)
Depressed client cannot afford psychologist (b)  2.5 (1.3)
Client wounded by religion in the past           2.3 (1.1)
You yourself have spiritual questions            2.2 (1.3)
You yourself need help for a problem (b)         1.9 (1.2)

Note. Ratings are reported as means (standard deviations). All items are
rated on a 5-point Likert scale, ranging from 1 (definitely would not
refer) to 5 (definitely would refer).
(b) = this item is significantly lower than the previous item, p < .01

TABLE 6. Psychologist Importance Ratings for Factors Related to Making a
Referral

Referral Ratings Scores

The minister has formal counseling training               4.3 (0.9)
The minister demonstrates respect for psychology          4.2 (1.0)
The minister has an open and inclusive view of religious  4.1 (1.2)
  faith
The minister is highly educated (a)                       3.8 (0.9)
The minister shares your religious beliefs (b)            2.0 (1.1)

Note. Ratings are reported as means (standard deviations). All items are
rated on a 5-point Likert scale, ranging from 1 (not at all important)
to 5 (extremely important).
(a) = this item is significantly lower than the previous item, p < .05
(b) = this item is significantly lower than the previous item, p < .01