By: Richard E. Boyatzis, Ph.D. Case Western Reserve University	Download PDF
Original Reference - Reprinted With Permission Boyatzis, R.E. (2005). Core competencies in coaching others to overcome dysfunctional behavior. In Vanessa Druskat, Gerald Mount, & Fabio Sala, (eds.). Emotional Intelligence and Work Performance. Erlbaum. 81-95.

Coaching, or more accurately executive coaching, is one of the few rapid growth industries of the last few years. First accepted as a practice in executive development, it has spawned tens of thousands of practitioners. As a measure of its popularity, a Google web search on December 12, 2002 revealed over 99,400 web sites using the phrase "executive coaching." HRD and OD internal and external consultants, psychotherapists, psychologists, social workers, teachers, and other professionals (e.g., lawyers, accountants, and nurses), have had business cards printed and promote themselves as executive or life coaches. Even in countries where management training and executive development are awkward because they threaten the public image of a person's competence and authority, such as in Italy (Altomare, 2002), Spain (Serlavos, 2002), and Japan (Voigt, 2002), coaching is the fastest growing sector of the human resource development business.

Coaching is not a new resource for people. In the past, we had professionals to help people develop, but it was hard to get access to them and often had a stigma attached, especially if it was a psychotherapist. Now even Tony Soprano, the fictional head of a crime "family" in the Home Box Office series of the same name, goes to a psychotherapist.

In the past, we had friends as well as others to whom we could turn for the type of personal advice and counsel now provided by executive coaches. Maybe we are too busy to effectively sustain friendships that once filled this void. Or, perhaps, we have become more sensitized to the need for developing our managerial and leadership talent. Therefore, we seek out more help and more specific help than ever before.

Although the practice of executive coaching has expanded dramatically, the writing about it has only expanded at a modest pace. Most of the literature is anecdotal from practitioners or wisdom from those who have provided such services for years (Kilburg, 1996; Kilburg, 2000; Kampa-Kokesch and Anderson, 2001). Except for the use of multi-source feedback (i.e., 360), there has not been a stream of empirical research in any of the aspects of executive coaching.

Research on effectiveness, style, and techniques in athletic coaching and psychotherapy has grown. But in the less formally prepared providers of executive or managerial coaching services, we still do not know much. A review of the few empirical studies and the anecdotal literature reveal confusion in the outcome of coaching. Many explorations focus on effectiveness of the coaching. But few studies actually try to predict behavior change. Many use satisfaction with coaches and the coaching experience, the perception of progress, or help from the person being coached as an indicator of effectiveness (Hall et. al., 1999; Kampa-Kokesch and Anderson, 2001). Unfortunately, these are really measures of the quality of the relationship with the coach, and not necessarily the effectiveness of the coaching. If the latter were the dependent variable of such studies (i.e., attainment of the intended outcome of the process), sustained change in a person's behavior, style, and performance would be measured and predicted.

To offer some insight into what we might find in studying executive coaches, this chapter will review a series of studies of people acting like coaches but in an extremely difficult arena, that of counseling people with alcoholism and substance abuse problems (Boyatzis and Burruss, 1979; Burruss and Boyatzis, 1981). Even thought these studies were conducted a number of years ago, they constitute one of the few empirical studies of competencies trying to predict a life outcome measure. As in the study of executive coaching, studies of counseling and peer counseling in helping people with alcoholism and substance abuse problems, there is a confusion about the dependent variable of the efforts. Maintenance or days of sobriety became the dependent variable and an indicator of effectiveness for many studies. Like the "quality for the relationship" for studies of executive coaching, it is not enough as a measure of effectiveness. A person is sober for some reason- that should be the dependent variable in such studies. One possible reason for sobriety is to become effective at work.

These studies were designed to determine the competencies of alcoholism counselors that predicted the work performance of their clients after they had returned to work. The first study generated a competency model through interviews. The interviews were used to identify competencies as they appeared while working with clients. The second study differentiated a set of competencies of effective versus less effective coaches/counselors with tests. In an extension of the second study, the tests were then used to predict work performance of the counselor/coaches' clients.

What Competencies Might Make a Difference?

Rogers (1951, 1961) articulated a troika of characteristics that seemed critical for effective "helping" behavior emerging from his work on psychotherapy: empathy, unconditional positive regard, and genuineness. Truax and Carkhuff (1967) and Carkhuff (1969) continued to expand, clarify, and develop these concepts and methods. Although developers or proponents of various approaches to psychotherapy would advocate modifications to this list (e.g., Perls emphasized giving voice to the unspoken and Ellis emphasized pragmatism), the quest for effective helping behavior always returned to characteristics of the helper. Later research on psychotherapy and counseling would suggest that the characteristics of the helper was more important in determining differences in effectiveness than differences in approach to psychotherapy or schools of thought in which the person was trained (Emrick, 1974).

A variety of studies have pointed to the importance of empathy in the effectiveness of those in helping relationships, such as coaches (Bohart and Greenberg, 1997; Hall and Bernieri, 2001). Other studies would offer a single characteristic thought to be important, such as Gendlin's focusing ability (Gendlin et. al., 1968).

Conceptual synthesis of competency studies on various roles within the human resource field suggest that a set of competencies that constitute emotional intelligence as well as several cognitive competencies play an important part (Spencer and Spencer, 1993; Goleman, 1998; Bortoman et. al., 1998). But again, we are left wondering if empirical and predictive testing of such hypotheses would show support for these beliefs or send researchers in another direction in trying to explain why some coaching and coaches seem to be effective in helping others and some are not.

Like that of executive coaches will become, the competency of counselors periodically emerges as a concern. In the alcoholism and substance abuse fields, it emerged from three sources. First, critical reviews of research on the effectiveness of treatment for alcoholics suggested that particular treatment techniques utilized did not account for differences in outcome measures, such as the behavior of patients following treatment (Emrick, 1974). Second, evaluation literature in the field of psychotherapy suggested that characteristics of therapists account for substantial differences in outcome measures (Whitehorn and Betz, 1954; Truax and Carkhoff, 1967; Glass, 1976; Burruss, 1977). Third, there was a movement, supported by the government, to establish a national body to certify alcoholism counselors and establish credentials which indicate the attainment of a professional level of competence in conducting therapy and counseling alcoholics.

The latter will emerge with greater frequency in the coming years. A need for some form of quality assurance has already appeared in the form of an association based certification of executive coaches. The International Federation of Coaches currently has the most active and widely known program for certification other than consulting firms or authors of programs privately "certifying" practitioners in their proprietary methods or techniques.

Methods

Predicting Client Work Performance

For an alcoholism counselor in the US Navy, the most objective measure of their effectiveness is the work performance of their clients following treatment. A work performance measure can be considered conservative (i.e., more difficult on which to show change) and a more difficult treatment goal than abstinence because it requires changes in the client's behavior, ways of dealing with others, as well as changes in his/her drinking behavior.

Boyatzis and Burruss (1979) collected the work performance data independently of the counselor competency data. Their method and results will be reviewed in depth in this chapter. First, a list of all clients admitted to US Naval alcohol rehabilitation facilities in the continental United States was produced. The clients were listed by facility. Second, the list was sent to each facility. Every counselor currently working at each facility was asked to complete a form. If they had worked as a counselor at the facility, they would circle the name of each client with whom they had worked. Each client spent a substantial amount of time with a particular counselor in individual and group therapy, meetings, and other activities.

The social security numbers of clients were used by the U.S. Navy Bureau of Personnel to identify each client's work performance for the six months following their treatment. This measure was a rating on a three-point scale: (a) "1" means that the person received a general discharge, an undesirable discharge, an honorable discharge but was considered unsuitable or unfit, was still on active duty but had a recommendation against reenlistment, or had a medical discharge with a recommendation against reenlistment; (b) "2" means that the person received a medical discharge but no recommendation against reenlistment; (c) "3" means that the person was still on active duty six months after treatment and had not received a recommendation against reenlistment, had received an honorable discharge with no recommendation against reenlistment, or was released to reserve status or to non-disability retirement.

Of the 65 counselors in the system, only 47 were available to be tested at the time of the studies (Boyatzis and Burruss, 1979; Burruss and Boyatzis, 1981). The minimum number of client records required for inclusion was 10. The average number of clients per counselor was 49, with the maximum of 167. The resultant sample was 45 counselors and 2,212 clients.

Discovering Key Competencies

Boyatzis and Burruss (1979) used a two-step process to identify the competencies of counselors in the first study. In the first step, an extreme case design was used to inductively discover the competencies that were likely to have an impact on client work performance following treatment. Nominations of those viewed as "outstanding" were collected about counselors who had worked more than 2 years in the system and were available for interviews and testing. A sample of "outstanding" counselors was identified as those receiving multiple nominations. A comparable sample of people not nominated by anyone was randomly chosen from the remaining counselors.

The Behavioral Event Interview (Boyatzis, 1982; Spencer and Spencer, 1993), a variation on the critical incident interview (Flanagan, 1954), was conducted with 26 counselors. Eight competencies were found to differentiate the outstanding from the "average" counselors. They were: (1) Optimism about people's ability to change, considered part of the Self-Management Cluster of EI competencies; (2) Initiative (called efficacy in the original studies), considered part of the Self-Management Cluster of EI competencies; (3) Pattern Recognition, considered part of a Cognitive Cluster of competencies; (4) Client Awareness, considered a part of the Social Awareness Cluster of EI competencies; (5) Accurate Self-Assessment (called a desire for personal growth in the original studies), considered part of the Self-Awareness Cluster of EI competencies; (6) Ego Maturity, considered a part of the Self-Management Cluster of EI competencies; (7) Empathy, considered a part of the Social Awareness Cluster of EI competencies; and (8) Emotional Self-Awareness (called genuineness in the original studies), considered part of the Self-Awareness Cluster of EI competencies.

Tests were selected or developed for each of these competencies. Since several of the competencies were thought to exist within a person at multiple levels (i.e., trait and skill levels of the competency), several tests were chosen to assess Initiative, Accurate Self-Assessment, and Empathy.

Optimism about people's ability to change: The Scenarios Test was designed specifically for this study. It consisted of eight scenarios of counseling situations adapted from actual events in the interviews. The person is asked to imagine himself in each situation and to select, in order of preference, two of ten possible responses that they would be likely to make. A subset of responses for each category was theoretically designed to assess Optimism, Initiative, and Accurate Self-Assessment. These were weighted by multipliers of 2 and 1 according to whether they were first or second choices respectively. This test required approximately 40 minutes to administer.

Initiative: A trait level of the initiative competency was assessed with an operant test by coding Cognitive Self-Definition from the Picture Story Exercise according to the coding system developed by Stewart and Winter (1977) to reflect whether a person habitually thinks of themselves in terms of causes and outcomes or whether a person sees the self as an ineffective victim of events which have an unknown cause. The Picture Story Exercise is a modified version of the Thematic Apperception Test (TAT) originally developed by Murray (1943). A person is asked to spend about five minutes writing a story in response to each of six pictures. People coding this measure had reliability with expert scorers above the 90 percent level.

To address possible method variance issues, a second trait level measure of Initiative was selected. The Nowicki-Strickland Locus of Control Scale is a forced-choice, respondent measure consisting of 40 questions was also used as another trait level respondent measure of Initiative. A high score indicates an external orientation (Nowicki and Strickland, 1973).

To assess the skill level of the Initiative, a scale in the Scenarios Test was developed (see above description of the Scenarios Test).

Pattern Recognition: The Test of Thematic Analysis is a measure of the cognitive competency called Pattern Recognition (Boyatzis, Cowen, and Kolb, 1995). This is thought to be essential in diagnosing problems and understanding others. Presented with two different groups of information, the test-taker is asked to describe the differences and similarities in any manner that he/she likes. The two groups of information used in this study were six counseling situations derived from the interviews. A person can receive a score from -3 to +6. The more a person demonstrates critical diagnostic thinking, the more positive the score will be (Winter and McClelland, 1978).

Client Awareness: In addition to one scale in the Scenarios Test, the Programmed Case was also used to assess Client Awareness. It is a series of 21 episodes of a true life history. Each episode is accompanied by a set of four alternative future episodes of which only one actually happened. The person being tested is asked to guess which of the four alternatives occurred. A specially designed answer sheet indicates whether he is right or wrong after each guess and they are instructed to keep guessing until he/she finds the correct alternative. An important aspect of this instrument is that the person has more information as a result of each guess and, therefore, should have a better understanding or "feeling for" the biographical character. To assess Client Awareness, a counselor's responses on the programmed case were scored according to their overall Performance (i.e., the standardized score of 21 minus the number of guesses made before finding the correct one).

Accurate Self-Assessment: Besides a scale from the Scenarios Test, the Helping Resources Inventory was used to assess a counselor's willingness and resourcefulness in using help to work with clients. The Helping Resources Inventory was adapted from the Activities Questionnaire used by McClelland (1975). The respondent was asked to rate, on a scale of ten, "how likely" he would be to consult each of 11 different sources of help (e.g., minister, psychologist, AA sponsor, friend, supervisor, etc.). The mean response per source was then taken to be an index of the person's Willingness to Seek Help. The person was then asked to indicate whether he has actually sought help from each of those same 11 sources within the past month, six months, or year. The mean response to this question, weighted in the direction of recency, was used as an indicator of the person's actual Use of Help. This questionnaire required approximately ten minutes to administer.

Ego Maturity: To assess ego maturity, the Picture Story Exercise was also scored for Stages of Ego Development, developed by Abigail Stewart and appeared in McClelland (1975). Persons coding the stages have reliability with expert scorers above the 90 percent level. The stories were scored for imagery in four stages. These stage scores were then combined into a single measure with the following formula (as described in McClelland, 1975): [Stage I+2(Stage II)+3(stage III) +4(stage IV)] divided by [Stage I + Stage II + Stage III + Stage IV].

Empathy: The Picture Story Exercise was also used to code a trait level of the Empathy competency. The stories were scored by trained scorers, who have scoring reliability with expert scorers above the 90 percent level, according to the system described in Atkinson (1958) for Need for Affiliation, and that developed by Winter (1973) for Need for Power and Activity Inhibition (McClelland, et. al., 1977). These three scores were adjusted for number of words written with a correction factor developed by Winter (1979). The standardized power score was subtracted from the standardized affiliation score and the difference was multiplied by the square root of the AI score plus one to construct a single variable. The variable, called the Caring Motive Profile, reflects a greater concern for close relationships to people than for having impact on them, while having a high degree of impulse control.

To assess the skill level of the Empathy competency, the Programmed Cases was used in a different way than earlier described. An Improvement Score was calculated as the increase or decrease in the accuracy of their guesses on the later half of the episodes versus the first half (see above description of the Programmed Case test). The Improvement score was adjusted for level of accuracy of early responses.

Emotional Self-Awareness: A modified version of the Focusing Ability, as originally designed by Gendlin, Beebe, Cassens, Klein, and Oberlander (1968), was used to assess Emotional Self-Awareness. Prerecorded instructions asked the person to first relax for a moment and then focus on some meaningful personal problem. The instructions continued at brief intervals to guide the person through an exploration of his/her feelings concerning that problem over a ten minute period. At the end of the ten minutes, the person was asked to answer, into an individually held tape recorder, specific questions about the experience. Two judges independently rated 43 of the 47 recordings on a four-point scale from "Definitely Did Not Focus" on feelings to "Definitely Did Focus" with an interrater reliability of r = .79. (The first four recordings were used to establish initial agreement on the scale definitions.) Disagreements were discussed and the final score was agreed. This score is referred to in this report simply as Focusing Ability. It was intended to be a measure of Emotional Self-Awareness because the "focusing" a person may demonstrate during this exercise is one of identifying their feelings and delving deeper into what they are and describing them. This instrument required approximately 20 minutes to administer.

Variables Other than Competencies

In addition to the above measures of counselor characteristics, the Ward Atmosphere Scale (WAS) was administered to assess the climate of the treatment facility. This instrument, developed by Moos (1974), is a 100-item questionnaire concerning important characteristics of the social environment. The ten subscales were combined by summing eight of them and subtracting "Anger and Aggression" and "Staff Control", which were scored in the opposite direction. The resultant variable, Climate, was intended to account for impact of treatment facility on client outcome.

Results

Counselors were classified as "superior" or "average" on the basis of a median split on the mean work performance rating of the clients with whom they worked. The "average" counselors had mean client work performance ratings ranging from 2.20 to 2.56; the percentage of clients with a rating of "3" ranged from 64 percent to 78 percent with an average of 70.5 percent. The "superior" counselors had mean client work performance ratings ranging from 2.60 to 3.00; the percentage of clients with a rating of "3" ranged from 75 percent to 100 percent (one of the "superior" counselors had a percentage of 75 percent; the rest had 80 percent or better) with an average of 87.2 percent.

T-tests were computed on each of the measures comparing average and superior counselors. The results, shown in Table 1 (Table 2 from Boyatzis and Burruss, 1979, page 25) are that superior counselors had significantly higher scores on the Thematic Analysis score, the Programmed Case: Improvement score, and the Focusing score. They also had near significantly higher scores on the Caring Motive Profile, Cognitive Self-Definition score, and the Programmed Case: Performance score. The Use of Resources was statistically significant, but opposite to the predicted direction; average counselors showed higher scores than superior counselors. All other differences were non-significant.

 

Table 1. Comparison of 15 Superior and 14 Average Counselors
Variable	Average Counselors	Superior Counselors	t	Sig. Level
Optimism: Scenarios/Opt	2.29	2.53	.32	ns
Initiative: Cognitve Self-Def. Nowicki-Strickland Scenarios/Initiative	-6.65 16.86 .57	-4.30 15.67 .80	1.56 -1.08 .37	.066 ns ns
Pattern Recognition: Thematic Analysis	.43	.93	2.01	.025
Client Awareness: Programmed Case/Perf.	55.36	55.85	1.45	.079
Empathy: Caring Motive Programmed Case/Imp.	-.92 2.34	.32 15.76	1.63 1.68	.058 .054
Accurate Self Assessment: Scenarios/ASA HRI/Seeking Help HRI/Using Help	7.29 6.08 2.23	5.00 6.53 2.42	-2.16 .72 .69	.02 ns ns
Ego Maturity: Ego Dev.	2.43	2.44	.08	ns
Emotional Self-Awareness: Focusing	1.71	2.64	2.55	.005
*One-tailed significance levels are reported

On the basis of these t-tests, the six measures indicated that four of the EI competencies and the cognitive competency could be considered valid differentiators of superior versus average counseling performance. They were: (1) Empathy at both the trait and skill levels as measured by the Caring Motive Profile and the Programmed Case: Improvement score, respectively; (2) Initiative as measured by Cognitive Self-Definition; (3) Pattern Recognition as measured by the Thematic Analysis score; (4) Client Awareness as measured by the Programmed Case: Performance score; and (5) Emotional Self-Awareness, as measured by the Focusing Ability.

The findings reveal that one or more competencies from each of the four clusters assessed were needed for effectiveness. These competencies were from the Self-Awareness, Self-Management, and Social Awareness Clusters of EI and the Pattern Recognition of the Cognitive Cluster.

A discriminant function analysis (DFA) was run on the six measures and the derived function used to assign counselors by groups. It was chosen as the multivariate statistical routine to perform this function due to the categorical nature of the client outcome data. The results are shown in Table 2 (Table 4 from Boyatzis and Burruss, 1979, page 29). The combination of the six measures yielded a high canonical correlation (r=.677) and correctly classified 83 percent of the counselors (correctly classified 87 percent of the superior counselors). A DFA was also run on the sum of the standardized scores on each of the measures, yielding an even higher canonical correlation (r=.898) and correctly classified 83 percent of the counselors (correctly classified 80 percent of the superior counselors). A third DFA was run on the number of competencies possessed by a counselor. This was computed by giving a counselor a +1 for each measure on which his standardized score was positive, and 0 for each measure on which his standardized score was negative. The sum was computed as the number of competencies possessed by the counselor. This DFA yielded a high canonical correlation (r=.633) and correctly classified 83 percent of the counselors (correctly classified 87 percent of the superior counselors).

Table 2. Discriminant Function Analysis of the Model
Variables	Canonical Correlate	% Counselors Correctly Classified (n=29)	% Superior Counselors Correctly Classified (n=15)
Caring Motive Profile Programmed Case:Improv. & Perf. Focusing Thematic Analysis Cognitive Self-Def.	.667	83%	87%
Sum of Standard Scores on all 6 Variables	.898	83%	80%
Number of Competenices Possessed	.633	83%	87%

The Impact of Situational Variables

Since the counselors worked in different treatment facilities, it was possible that organizational climate of the facility may affect their performance. The correlation between the Moos Climate score and client outcome (mean performance rating) was not significant (r=-.106, n=29). A t-test of the climate in which superior counselors worked versus the average counselors worked was also not significant, means of 50.00 and 49.93 respectively (t=.03, n=29). Another situational characteristic which could possibly have an impact on counselor effectiveness was the age of their clients. The correlation of mean work performance rating with the percent of clients over 25 years old was not significant (r=-.038, n=29). Average counselors had a mean of 51.06 percent of their clients over 25, while superior counselors had a mean of 49.66 percent of their clients over 25. This difference is not significant (t=-.39, n=29).

As a further check, correlations were computed of the Moos Climate score and percent clients over 25 against each of the competency measures. None of the correlations with percent clients over 25 was significant. None of the correlations with the Moos Climate score was significant.

Predicting Client Work Performance

A multiple-regression was conducted with the average client work performance rating per counselor as the criterion available (Work Performance) and the six competency test scores as the predictor variables. Since days of a client's active service after treatment and the work performance measure were related (r = .378; df =43; p<.01), it was decided to control for the confounding effects of unequal time periods between treatment-discharge and service-discharge dates. A seventh independent variable was included, representing the average number of days between treatment and discharge from the service for each counselor's sample of clients.

The multiple regression showed a significant multiple correlation (R=.562) in predicting Work Performance from the competencies, as shown in Table 3 (Tables 1 and 3 from Burruss and Boyatzis, 1981, page 9 and 12). A test of curvilinearity (F=.474; p=ns) demonstrated no significant deviation from linearity. A detailed analysis of the individual contribution of each variable revealed, however, that the primary contributors to that relationship were Days of Service, Empathy as measured by the Programmed Case: Improvement score, and Emotional Self-Awareness, as measured by the Focusing measure. Since Days of Service was stepped into the equation first to eliminate its contaminating effects on predicting Work Performance, the data suggest that two competencies, Empathy and Emotional Self-Awareness, are more important than others in predicting the effectiveness of counselors.

Table 3. Predicting Work Performance: Summary Table (n=45)
Analysis	DF	Sum of Squares	Mean Squares	F
Regression	7	.426	.061	2.437
Residual	37	.924	.025
Multiple R=562, R2 = .316, Standard error = .158
Variable	Multiple R	R2	R2 Change	Simple R	Beta	F
Days of Service	.378	.142	.143	.378	.459	9.567
Caring Motive	.380	.144	.001	-.025	-.015	.011
Prog.Case: Perf.	.380	.144	.000	-.025	-.002	.000
Prog.Case: Impr.	.506	.256	.112	.307	.331	5.117
Cog.Self-Def.	.507	.257	.000	.192	-.049	.094
Focusing	.554	.306	.050	.174	.254	3.113
Thematic Analysis	.562	.316	.009	-.053	-.101	.495

Discussion of Results

The superior counselors appeared different from the average counselors. This does not imply any difference in the dedication or compassion of the individual counselors, but reflects a difference in their effectiveness. Given the relatively high effectiveness of the Navy alcoholism treatment programs as a whole (Bucky, 1977), the differences discussed do represent the difference between superior and average counselors and not between adequate and inadequate counselors.

Two competencies, Empathy and Emotional Self-Awareness, distinguished effective counseling and predicted work performance. Another competency, Pattern Recognition, appeared to contribute to the overall effectiveness of a counselor, but did not add unique or distinctive capability as reflected in the regression analysis. Two others, Initiative and Client Awareness, also appeared to contribute to the overall effectiveness, but not as strongly and did not add unique variance in the regression analysis.

Application of these results to executive coaching would confirm what most coaches, and their clients, know-- Empathy or sensitivity to the client is the key characteristic of the effective coaches. Coaching, like counseling, cannot proceed without listening to and understanding the client, his/her issues, problems, and situation at work and at home. If the client is viewed as merely a "problem bearing platform," the coach will focus on the problems and tasks not the person. In the process, he/she may miss underlying issues or factors contributing to sustaining current behavior and impact. The coach must be sensitive to changes in the client and their process of change to tailor their comments and suggestions to that person's needs at that point in time.

This brings us to the second critical competency revealed in this study, Emotional Self-Awareness. A coach, like a counselor, cannot focus on their client and understanding their situation if the coach is preoccupied with his/her own challenges. Similarly, awareness of transference, counter-transference, and projection must be an integral aspect of the functioning of an executive coach. They must be able to separate their own feelings and values from those of the client. This is not to say that the coaches feelings and values are not important, they are. But the coach has the responsibility to be able to identify and manage his/her feelings and reactions. Whether the coach uses this information as part of understanding the client or as a vehicle for suspending their own needs and anxieties, managing one self is difficult if not impossible without a high degree of self-monitoring, or Emotional Self-Awareness.

The analyses strongly suggest that counselor competencies account for the majority of variance in effectiveness with clients. Situational factors do not seem to account for variation in client outcome. Although this study did not include a control for type of treatment technique, it is important to indicate that all of the Navy programs were heavily based on practices and procedures of Alcoholics Anonymous and peer coaching. A variety of other techniques are used in their programs, such as psychodrama, education, group and individual therapy, and so forth. Three of the fourteen average counselors were civilian (21 percent) and five of the fifteen superior counselors were civilian (33 percent), suggesting no differential impact coming from civilian status. Five of the fourteen average counselors (35 percent) and six of the fifteen superior counselors (40 percent) were not trained in the Alcohol Treatment Specialist program, suggesting no differential impact from the Navy's counselor training program.

It is important to note that the "success rate" of the average Navy alcoholism counselors (70.5 percent) is almost identical to the 67 percent found in most psychotherapy outcome studies conducted (see Burruss, 1977). This suggests that these counselors are doing at least as well as any other treatment group. The considerably higher "success rate" of the superior counselors (87.2 percent), on the other hand, highlights the efficacy of increasing the number of counselors who possess and utilize specific competencies as a way of improving the overall effectiveness of counseling.

Summary and Implications

The results showed that two competencies appeared to have a substantial and significant impact on a counselors' effectiveness. To be effective as a counselor, and by extension an executive coach, a person must be sensitive to others. To be sensitive to others, he/she must be sensitive to themselves. These critically important competencies were: Emotional Self-Awareness from the Self-Awareness Cluster of EI competencies; and Empathy from the Social Awareness Cluster of EI competencies. Both of these were significant at the skill level. This implies that training or developing these competencies may be more feasible than if they were at the trait or motive level of personality.

Regardless of the organizational climate of the various facilities in which they worked, type of training received to prepare for their role, and the age demographics of their clients, these characteristics, these competencies explained why some counselors were more effective than others.

To extend these findings into the arena of executive coaching, we can infer that a similar set of competencies would help us to understand why some coaches are more effective than others. Specifically, we can also hypothesize that the most critical competencies will come from a wide variety of the clusters, and that Emotional Self-Awareness and Empathy are most likely to be two of the competencies critical to coaching effectiveness. We need a parallel study to those reviewed here conducted with executive coaches and their clients to determine the precise competencies and their impact. We need to expand the sample size to generalize to all forms of executive and life coaching.

REFERENCES

Altomare, P. (2002). Personal communication from a senior consultant at the Hay Group. Milan, Italy.

Atkinson, J. W. (Ed.)(1958). Motives in fantasy, action, and society. Princeton, NJ: Van Nostrand.

Bohart, A.C. & Greenberg, L.S. (1997). Empathy reconsidered: New directions in psychotherapy
Washington D.C.: American Psychological Association.

Boyatzis, R.E. (1982). The competent manager: A model for effective performance. NY: John Wiley.

Boyatzis, R.E. & Burruss, J.A. (1979). Validation of a Competency Model for Alcoholism Counselors in the Navy. Final Report on Contract Number N00123-77-C-0499.

Brotman, L.E., Liberi, W.P., & Wasylyshyn, K.M. (1998). Executive coaching: The need for standards of competence. Consulting Psychology Journal: Practice an Research, 50 (1), 40-46.

Bucky, S. F.(1977). The 1976 evaluation of the Navy's Alcohol Rehabilitation Programs. Unpublished study Naval Alcohol Rehabilitation Center, San Diego.

Burruss, J. A. (1977). Evaluating therapists' effectiveness. Unpublished doctoral dissertation, Harvard University.

Burruss, J.A. & Boyatzis, R.E. (1981). Continued Validation of a Competency Model for Alcoholism Counselors in the Navy. Final Report on Contract Number N002 44-80-C-0521.

Carkhuff, R.R. (1969). Helping and human relations: A primer for lay and professional helpers. NY: Holt, Rinehart and Winston.

Emrick, C. D. (1974). A review of psychologically oriented treatment of alcoholism. Quarterly Journal of Studies on Alcohol, 35, 523-549.

Gendlin, E. T., Beebe, J. III., Cassens, S., Klein, M. and Oberlandes, M. (1968). Focusing ability in psychotherapy, personality and creativity. Research in Psychotherapy, 3, 217-241.

Hall, D.T., Otazo, K.L., & Hollenbeck, G.P. (1999). Behind closed doors: What really happens in executive coaching. Organizational Dynamics. 27(3), 39-53.

Hall, J.A. & Bernieri, F.J. (2001). Interpersonal sensitivity: Theory and measurement. Mahwah, NJ: Lawrence Erlbaum Associates.

Kampa-Kokesch, S. & Anderson, A.Z. (2001). Executive coaching: A comprehensive review of the literature. Consulting Psychology Journal: Practice and Research. 53(4). 205-228.

Kilburg,, R.R. (1996). Toward a conceptual understanding and definition of executive coaching.
Consulting Psychology Journal: Practice and Research. 48(2), 134-144.

Kilburg, R.R. (2000). Executive coaching: Developing managerial wisdom in a world of chaos. Washington DC: American Psychological Association.

McClelland, D. C. (1975). Power: the inner experience. New York: Irvington Press.

McClelland, D. C., Davis, W. N., Kalin, R., Wanner, E. (1972). The drinking man. New York: Free-Press.

Moos, R. (1974). Evaluating treatment environments: a social ecological approach. New York: John Wiley.

Murray, H. A. (1973). Thematic apperception test. Cambridge, MA: Harvard University Press.

Nowicki, S. & Strickland, B. R. (1973). The locus of control scale for children. Journal of Consulting and Clinical Psychology, 3 (40), 148-154.

Rogers, C.R. (1961). On becoming a person: A therapist's view of psychotherapy. Boston: Houghton Mifflin.

Rogers, C.R. (1951). Client-centered therapy: It's current practice, implications, and theory. Cambridge: The Riverside Press (republished 1965, Boston: Houghton Mifflin).

Serlavos, R. (2002). Personal communication from a senior Professor at ESADE who was a former consultant at the Hay Group. Barcelona, Spain.

Spencer, L.M. Jr. & Spencer, S. (1993). Competence at work: Models for superior performance. NY: John Wiley.

Stewart, A. J. & Winter, D.G. (1977). Self-definition and social definition in women. Journal of Personality. 42. 238-259.

Truax, D. B. & Carkhuff, R. R. (1967). Toward effective counseling and psychotherapy. Chicago: Aldine Publishing.

Voigt, K. (2002). Job coaches catch on in Asia. Wall Street Journal (European Edition). Tuesday, November 12, 2002, page A11.

Whitehorn, J. C. and Betz, B. J. (1954). A study of psychotherapeutic relationships between physicians and schizophrenic patients. American Journal of Psychiatry. 111, 321-331.

Winter, D. G. (1973). The power motive. New York: The Free Press.

Winter, D. G. (1979). Correcting projective test scores for the effect of significant correlation with length of protocol. Unpublished Technical Note. Boston: McBer and Company.

Winter. D.G. & McClelland, D.C. (1978). Thematic analysis: An empirically derived measure of effects of liberal arts education. Journal of Educational Psychology. 70(1). 8-16.