llllllllllllllllllilllllllllllllllll 3 1293 00065 0329 L,— LIEEMBY 1 Micki-raw skate .. I ,T‘ ' .2. ’8 ." .. s Umvcmarg : I ‘ l l \ \ -..-._i This is to certify that the thesis entitled A TEST OF CLIENT-COUNSELOR PERSONALITY MATCHING ACCORDING TO HOLLAND'S THEORY presented by Donald J MacDonald has been accepted towards fulfillment of the requirements for Ph.D. Counseling, Educational Psychology, and Special Education Qcfiaa Major professor degree in Dr. Andrew Porter Date May 4, 1984 0-7639 MSU is an Affirmative Action/Equal Opportunity Institution 'bvisSIBJ RETURNING MATERIALS: _ Place in book drop to LJBRAfiJES remove this checkout from J-I-IE-ll-L your record. FINES will be charged if book is returned after the date stamped below. “egrvsixtn; I My ' -- J r" ‘t A TEST OF CLIENT-COUNSELOR PERSONALITY MATCHING ACCORDING TO HOLLAND'S THEORY by Donald J MacDonald A DISSERTATION submitted to Michigan State University in partial fulfillment of the requirements for DOCTOR OF PHILOSOPHY Department of Counseling, Educational Psychology and Special Education 1984 ©1984 DONALD J MAC DONALD All Rights Reserved ABSTRACT A TEST OF CLIENT-COUNSELOR PERSONALITY MATCHING ACCORDING TO HOLLAND'S THEORY by Donald J MacDonald The purpose of this research was to explore the efficacy of using John Holland's (1973) theory of vocational typologies to match for similarity the personalities of a client and a counselor in their initial contact in one-to-one occupational counseling. As such, the research was part of a line of study in the area of client-counselor matching in general. The research was also a blending of these major lines of investigation that have been only tangentially linked in past studies--client-counselor personality matching for similarities, Holland's personality typologies, and Strong's (1968) Social Influence Models. The research hypothesis was: For subjects with congruent, consistent, and well-differentiated Holland typologies, the more similar an audiotaped counselor's personality was to the subject's personality, the greater would be the subject's preference for the counselor. Subject typologies were identified through the Vocational Preference Inventory. A 6 x 6 Latin Square was used to control for order effects in a repeated measures design where each subject rated for preference each of six different counselor personalities. Data from the subject ratings were analyzed through a three-way repeated measures analysis of variance. The research hypothesis was tested directly through planned comparison _t_-tests. The first planned comparison was statistically significant, Donald J MacDonald indicating subject preferences for counselor presentations that exactly matched the subject's own personality typology. Implications of the results for client- counselor personality matching research are discussed. Suggestions for improving studies on this tapic are also discussed. ACKNOWLEDGEMENTS This dissertation is dedicated to those who contributed so much to its completion. For their patience and input in helping me learn how to become a better researcher, I thank my dissertation committee: Andrew Porter, Director Alex Cade Richard Jolmon, Advisor Gwen Norrell For their cooperation, participation, and unflagging support, I thank the people of: Seattle First Church of the Nazarene Seattle Pacific University For their unconditional emotional and material support, without which I never would have progressed to the point of a doctoral program, I thank my parents: Paul J and Angie Ruth MacDonald For their helpfulness in many valuable ways, I thank my other parents: Leo and Gail Collar Herbert and Belva Kjellberg Jerry and Shirley Phillips For sharing my time and energy with this undertaking, I thank my sons: Jeffrey Leo Collar MacDonald Joshua J MacDonald ii For their love, spiritual guidance, and professional perspectives, I thank two very special women: Constance Kiellberg MacDonald Ginger Phillips MacDonald With the love and grace of God plus the support of those people listed above, I am a more complete person and have finished this project. iii TABLE OF CONTENTS List of Tables List of Figures CHAPTER 1: STATEMENT OF THE PROBLEM Purpose of the Study Importance of the Study Research Focus Overview of Subsequent Chapters CHAPTER II: LITERATURE REVIEW Client-Counselor Relationship Matching Models Personality Factors Values Orientation Cognitive Characteristics Demographic Variables Treatment Methods Client-Counselor Similarity and Attraction Holland's Theory Consistency Differentiation Congruence iv vii viii ll 13 Ill 15 16 21 2‘4 26 28 Matching Applied to Holland's Theory Summary CHAPTER III: METHODOLOGY Research Hypothesis Research Design Subject Selection Demographics of Sample Instrumentation Scripts and Audiotapes Development Study Implementation Pilot Study Statistical Analysis Procedures Summary CHAPTER IV: RESULTS Study Purpose and Design Review Statistical Results Summary of Results CHAPTER V: CONCLUSIONS, DISCUSSION, AND RECOMMENDATIONS Conclusions Discussion and Recommendations Matching Studies Holland's Model Social Influence Model Methodology Final Remarks 29 32 35 35 36 4o 43 at: as 49 51 51 53 54 54 55 57 59 59 61 62 64 67 69 71 Appendix A: VPI Results Profile Sheets Appendix B: SPU Letter Appendix C: Church Letter Appendix D: Counselor Scripts Appendix E: Data Collection Release Form Bibliography vi 73 87 88 90 91 92 LIST OF TABLES Results of Intraclass Correlations Over Holland Typologies Audiotapes Latin Square with Data Three-Way Analysis of Variance Sources of Variation Planned Comparisons for Distance Variables vii 49 55 56 57 LIST OF FIGURES Holland typologies hexagon Latin Square design Arrangement of independent variables Counselor choice rating scale Holland typology identification viii 23 36 37 39 48 CHAPTER I STATEMENT OF THE PROBLEM Purpose of the Study The purpose of this research was to explore the efficacy of using John Holland's (I973) theory of vocational typologist to match for similarity the personalities of a client and a counselor in their initial contact in one-to-one counseling. As such, the research was part of a line of study in the area of client-counselor matching in general, which has been a focus of research interest for three decades. During the same three decades in which the matching studies occurred, Holland investigated how adolescents and adults in the United States made occupational choices. Part of his theory dealt with matching individuals with occupational environments with which they were similar. The present study sought to extrapolate Holland's findings on individual- vocational environment matching to client-counselor matching for personality similarity at their initial contact. Subjects were selected to represent all of Holland's six personality typologies. The subjects listened to audiotapes representing six different counselor personalities which also corresponded to Holland's personality typologies. Subjects then rated how much they liked or disliked each counselor presentation. Importance of the Study It is believed that practical benefits might result from using Holland's typologies to match clients and counselors on basic personality orientations. 1 Generally, these benefits would be gains in counseling outcomes and in counselor training. Enhancement of counseling outcomes is a major incentive for conducting matching studies. Parloff, Waskow, and Wolfe (1978) summarized this point: . . . all matching research implies that one is attempting to discover combinations of therapists and patients of an optimal degree of similarity and dissimilarity on a wide array of psychological dimensions that will enhance the probability of positive therapeutic changes in patients (p. 265). Once any such optimal combinations are identified, they can be predicted and manipulated to more systematically bring about client gains. An effective technology of client-counselor matching should, in turn, improve the capability of the mental health delivery system to provide better counseling services. Ideally, matching would speed up identification of counseling relationships that have a strong chance of being productive, thereby reducing time and resources that might be spent by counselor and client deciding whether or not they can work together. A personality matching system might also help to clarify occurrences of client psychological deterioration, premature termination, and retarded progress of counseling. Bergin and Lambert (I978) documented these events and reported hypotheses about why they happen. It is possible that client-counselor mismatching accounts for some of these difficulties, said Bergin and Lambert, and steps should be taken to rectify such problems. Counselor education is another area that might benefit from a client- counselor personality matching system. Such a system might be an important part of counselor trainee selection where, for example, particular types of trainees could be admitted for preparation to serve matched client groups. This system could augment training as well by delineating to trainees their general personality "strengths" and "weaknesses" in working with various client personality types. A matching system based on Holland's (I973) personality typologies would have the basic counseling outcome and training benefits just mentioned plus a few more. Holland's typologies are part of a larger theory which might be applied to matching studies in general, an area that could use strong theoretical organization. In addition, Holland's theory is supported by three popular assessment instruments which might be used in matching counselors and clients. Research Focus The current project was stimulated by Bruch (1978), who conjectured that personalities of counselors and clients could be identified and intentionally paired through the use of Holland‘s (I973) typologies. Bruch reasoned that since effective predictions could be made about optimal goodness-of-fit between individuals' personalities and environments or settings in which they might be employed, then the theoretical principles and research were applicable to the general counseling settings as well. According to Bruch, a counseling relationship is an environment that requires the same client decisions as to selection, satisfaction, and maintenance as vocational settings. Results of empirical studies documented by Holland and others in the vocational development area (Osipow, 1983; Weinrach, 1979) suggest that Holland's concept of similarity between a person's personality and an occupational environment is much like the idea of client-counselor personality matching for similarity. First, Holland's typologies are personality types, as are the independent variables in many matching studies. Second, the typologies are personality constructs based on general, measured, vocational interests and choices. Many personality matching studies also use generalized personality definitions and measurements. Third, it is predicted that greater similarity between typologies and occupational settings will produce more positive career outcomes, due to greater attractiveness and rewards of the career. Client- counselor matching studies often predict more positive results with greater client-counselor personality similarity, as similarity is thought to increase client-counselor attraction and mutual reward. While the application of Holland's (1973) research to client-counselor matching can be argued logically, such an application lacks an empirical basis. Bruch (1978) identified just one published (Cox 6: Thoreson, 1977) and one unpublished (Bruch 6c Gilligan, 1978, cited in Bruch, 1978) paper to support his contention. A review of major counseling periodicals for the present study revealed only two studies (Cox 6: Thoreson, I977; Kivlighan, Hageseth, Tipton, 6c McGovern, 1981) that pertained to matching according to the Holland theory. From the theory of Holland and from some evidence that client-counselor similarity constitutes an influential dimension in the establishment and maintenance of a helping relationship, it seems that the indirect evidence in favor of using Holland's typologies to match counseling participants is strong. Further tests, thought, are necessary in order to move beyond conjecture. The research here provides direct testing of the suggestion that Holland's personality typologies are useful in predicting client attraction to different counselors in a simulated first session for career counseling, according to the amount of personality similarity between client and counselor. Overview of Subsequent Chapters The following chapters detail the research hypothesis, rationale for the hypothesis, the ways in which it was tested, the results of those tests, and implications of the results. Chapter II reviews major articles and books pertaining to client-counselor attraction, different matching models, Holland's theory, and past attempts to integrate Holland‘s theory with matching. Chapter III specifies the methods and reasons for this research project's design and statistical analyses and the ways in which both facets were put into practice. Description of the sample and how subjects were obtained is part of the chapter. Description of the dependent measures and methods for their interpretation are included. The matching procedures investigated are described, as are the results of a pilot study used to check the efficacy of the procedures. Results of the data analyses are presented in Chapter IV. Chapter V interpretes the results in light of the purposes of this particular study and other studies on matching. Results are also synthesized to explain their meanings and to develop suggestions for further research in this tapic area. CHAPTER II LITERATURE REVIEW Client-Counselor Relationship Several attempts have been made over the past three decades to enhance the establishment and maintenance of one-to-one client-counselor relationships. The importance of the relationship has been consistently supported by writers who have clear-cut theoretical orientations that differ from each other (Hosford, 1969; Menninger, 1958; Rogers, 1951) as well as by more eclectic writers (Carkhuff, 1972; Scheid, 1976; Venzor, Gillis, 6c Beal, I976; Wolbert, 1977). A positive bond between client and counselor has been indicated as a key element in client gains with all types of major counseling issues (i.e., academic, personal, and vocational concerns). According to Jerome Frank and his colleagues (Frank, Hoehn-Saric, Imber, Liberman, 6: Stone, 1978), the therapeutic relationship is "one of the more important prerequisites for the success of all psychotherapies . . . " (p. 181). A negative client-counselor relationship, on the other hand, is seen as a large contributor to either a lack of progress or deterioration in client functioning in daily operations (Bergin 6t Lamberg, 1978; Carkhuff, I972; Strupp, Hadley, 6c Comes-Schwartz, 1977). As a result, counseling researchers have sought ways to improve the therapeutic relationship. Goldstein (1980) defined the therapeutic relationship as " . . . feelings of liking, respect, and trust by a client toward the helper from whom he is seeking assistance, combined with similar feelings of liking, respect, and trust on the part of the helper toward the client" (p. 20). Important elements of the 6 definition in conjunction with this study are those of the client liking, respecting, and trusting the counselor. These three elements seem more likely to be part of positive counseling relationships where counselor and client believe that they have some major similarities with each other (Bent, Putnam, Kiesler, 6c Nowicki, 1976; Berzins, 1977; Huston 6c Levinger, 1978; Parloff et al., 1978; Shertzer 6: Stone, 1980). Greater client-counselor similarity seems to be associated with greater interpersonal attraction, perhaps due to experienced and/or anticipated rewards through a therapeutic relationship. Matching Models Efforts at improving client-counselor relationships have often taken the form of matching the two people along dimensions thought to be particularly relevant to continued association and to client progress. Matching is a broad area that can be broken down into five sub-areas: (a) personality factors, (b) values orientation, (c) cognitive characteristics, (d) demographic variables, and (e) treatment methods. It has already been mentioned that this research project focused on matching in terms of personality factors. Review studies in the areas of cognitive characteristics, values orientation, demographic variables, and treatment methods should better place the personality factors matching studies in context of the whole matching research emphasis. Further, it will be shown that values orientation and cognitive characteristics matching research results have ramifications for personality factors matching research, whereas demographic variables and treatment methods results are less applicable. Overviews of matching research results on demographic variables and treatment methods are included, though, in order to give the reader a complete picture of the work that has been done in the area of matching. Personality Factors Consideration of personality matching research must start with a definition of the term "personality." Deriving an exact definition of personality is beyond the scope of this research, as it has eluded scholars who write entire texts on the subject. Nevertheless, some basic criteria will be outlined to provide a basis from which to judge what information and explanations apply to personality matching and what do not. The criteria used here are extrapolated from Corsini (1977) and from Shertzer and Linden (1979). Both sources discussed personality in terms of behavior patterns and explanations about why the behaviors might have occurred. The behavior patterns describe and generally predict how a person would act in various situations. The explanations draw upon models of how a person's attitudes, emotions, thoughts, and values are organized, plus how that organization is manifested in behavior patterns. In order to have greater control over possible sources of confounding, only verbal aspects of personality will be investigated in this study. Personality can appear as well, though, through nonverbal behavior patters such as walking. One of the two most notable approaches to personality matching is Carson's (1969) Interpersonal Transactions model. Carson proposed that client— counselor interactions be classified as complementary (i.e., rewarding for both people) or anticomplementary (i.e., mutually discomforting or punishing). Carson held that a client could be assessed to determine her/his habitual style(s) of behavior towards other peOple. It would then be possible to have the client work with a counselor whose behavioral style was complimentary so that the two were more likely to feel rewarded and, therefore, more likely to continue their relationship. Once the therapeutic relationship seemed on solid ground, though, it was recommended that the counselor intentionally behave in anticomplimentary ways in order to encourage client movement away from one or more dysfunctional behavioral styles and toward more effective behavioral styles (Carson, I969; Dietzel dc Abeles, 1975). Evidence to date indicates that the Interpersonal Transactions model may be a productive approach to matching research. In his extensive review of matching studies, Berzins (1977) cautiously endorsed the Interpersonal Transactions approach on the basis of its (a) limited but generally supportive research foundation, (b) emphasis on multiple aspects of client-counselor personality interactions rather than only a single aspect or global, loosely defined aspects, and (c) explicit tie with a theoretical foundation. The line of research just discussed has some importance for a study on personality matching through Holland's (1973) typologies. First, the three positive features that Berzins (1977) highlighted are useful for comparisons with the matching model being evaluated in this study. A second consideration is that Carson's (1969) terms "complimentary" and "anticomplimentary" are not synonymous with this project's terms "similar" and "dissimilar." The Interpersonal Transactions model labels , client-counselor interactions complimentary when they are dissimilar and rewarding for both people. Anticomplimentary interactions are also dissimilar but are unrewarding for the client and the counselor. Unlike this and many past matching studies, the Interpersonal Transactions approach has client-counselor personality dissimilarity as a necessity in order for positive counseling outcomes. Such a fundamental difference on the similarity-dissimilarity dimension exemplifies the lack of clarity and consensus in the whole matching research area (Berzins, 1977; Parloff et al., 1978). The second of the two most notable approaches to personality matching is the A-B model. Beginning in 191111, Whitehorn and Betz (1960) sought ways to 10 identify and predict which psychiatrists were most effective in helping diagnosed schizophrenic patients. The most effective or "A" psychiatrists were described as more understanding of the personal meanings of patients' actions, more likely to develop treatment plans tailored to individual patients, and were more active in conducting treatment than were the less effective or "B" psychiatrists (Cox, 1978; Parloff et al., 1978). Further studies (Whitehorn dc Betz, 1960) suggested that 23 items from the Strong Vocational Interest Blank were useful in distinguishing Type As from Type B5. A few years later, it was concluded that Type A psychiatrists were more effective than Type B psychiatrists in helping diagnosed schizophrenics and that Type Bs were more effective in helping patients diagnosed as neurotic (Cox, 1978; Parloff et al., 1978). The A-B approach to client-counselor personality matching has been extensively researched. A number of reviews of A-B research appeared during the past decade. Evaluations of the A-B research efforts have all been negative. Bergin and Suinn (1975) were the most pessimistic: "Considerable activity persists on the A-B variable but this area is plagued by ambiguities and poor studies . . . . It seems increasingly likely that the A-B measure will go the way of others that have such major difficulties" (p. 520). Cox's (1978) evaluation agreed with Bergin and Suinn (1975). He identified primary problems with A-B research as being (a) ambiguous and subjective definitions of the A and B types, (b) questionable criteria for assessing patient improvement or lack of improvement, (c) a lack of suitable cross-validation research, especially in light of the numerous studies on the topic, and (d) methodological errors in deriving the 23 item A-B scale from the Strong Vocational Interest Blank. An evaluation by Razin (1977) was least negative inasmuch as he concluded that the A-B approach had identified a consistent general pattern of personality ll differences between the A and B psychiatrists. Razin stated that the A-B distinctions were, however, weak predictors of actual counseling process and outcomes and that more specific variables were stronger predictors. Results of the A-B approach to personality matching have a number of implications for the Holland typology approach investigated here. First, the model is based on therapeutic interventions with people diagnosed as schizophrenic; generalizing any principles to this study of non-schizophrenics would be limited. Second, clear-cut data-based criteria for defining different personality types were lacking. Loosely defined types and/or those based on a priori assumptions may create more problems than they solve. Third, even if the personality types are data-based and clear-cut, general personality type definitions might be less useful than more specific definitions in identifying relationships between the definitions and research outcomes. Personality matching research using Holland's (1973) typologies will be reviewed at the end of this chapter. Values Orientation Major recent reviews of matching research (Berzins, 1977; Parloff et al., 1978) identified few studies in client-counselor matching according to values. Both reviews also indicated that results so far were inconclusive. The focus of the studies that were reviewed was to assess various client and counselor values through one or more standardized instruments before counseling began and then to see what happened. Some results of post hoc analyses suggested that similarity of counselor and client values were positively related to counseling process (e.g., continuation of the counseling relationship) or to outcome. Other post hoc results indicated that complimentary (as defined for the Interpersonal 12 Transactions model) client and counselor values were positively related to counseling process; no outcome studies were mentioned. Some values-oriented studies were not discussed by either Berzins (1977) or Parloff et a1. (1978). For example, Cheney (I975) predicted that males who had been jailed for alcohol intoxication would be more attracted to an audiotaped male counselor whose values toward alcohol usage were similar to their own than to an audiotaped counselor with dissimilar values. The results failed to support his prediction, as no statistically significant differences appeared in attraction ratings for similar and dissimilar counselors. Lewis and Walsh (1980) explored the reactions of college females to an audiotaped female counselor whose values on premarital sexual intercourse were either similar or dissimilar to their own. The counselor with similar values was rated and more attractive and trustworthy than the dissimilar counselor. These two studies on initial attraction to an audiotaped counselor are consistent with the mixed results reported by Berzins (1977) and Parloff et a1. (1978). The status and direction of values matching research appears to be quite uncertain at a time when the influence of values on counseling is considered to be an important area for continued research (Brammer 6: Shostrom, I982; Parloff et al., 1978). The personality factors matching research seems to have the same difficulties as values research but to a lesser degree. That is, both areas have weaknesses in (a) theoretical organization, (b) cross-validation, and (c) clarity and specificity of definitions (i.e., the criteria used to define different types of personality or different types of values). Conceptual contributions to the current study from the values orientation approach are insignificant. The values approach is useful, though, from the standpoint of its reliance on the similarity assumption. A more important 13 connection is in the criteria discussed earlier for what may be called "personality." Values can be regarded as explanations of behavior patterns (Corsini, 1977; Shertzer 6: Linden, 1979). As such, then, they are at least somewhat related to personality factor approaches to matching. Cognitive Characteristics Research on client-counselor matching through cognitive characteristics relies upon the assumption that greater similarity of client-counselor cognitions is related to increased effectiveness of counseling outcomes. When taken together, the few studies that have been conducted indicate a positive relationship between some facets of client and counselor cognitive similarity and counseling outcomes. These studies also indicate a positive relationship between some facets of client-counselor cognitive dissimilarity and counseling outcomes (Berzins, 1977; Parloff et al., 1978). Such results are not necessarily contradictory, however, as the similar and dissimilar facets are regarded as qualitatively different from each other and are, therefore, indicative of separate aspects of the therapeutic relationship. A problem is the definition of "cognitive characteristics." Agreement seems to exist for a broad definition based on the ways in which an individual understands or describes life experiences. More specific definitions, though, vary considerably (Parloff et al., 1978), including measured intelligence, conceptual complexity, conceptual stage (a la Jean Piaget), and George Kelly's Theory of Personal Constructs (Berzins, 1977; Parloff et a1.) The client-counselor cognitive matching research has features in common with the A-B personality factors research. First, it assumes that client- counselor similarity is preferable. Second, the cognitive matching approach has Ill difficulties in definition. Third, cognitive studies are weakly linked to broader theoretical frameworks and few studies are related to each other. The cognitive matching approach is also like the values orientation approach in some respects. First, it appears to be an area with some potential, but few empirical studies have been conducted. Second, definitions of major constructs are vague. Third, it can be regarded as part of the broader personality area of research, since cognitions entail explanations of behaviors (Corsini, 1977, Shertzer 6t Linden, I979). Demographic Variables This area of matching is broad, with mixed results. Numerous studies have been published on the client-counselor matching variables of socioeconomic status, race, age, gender, marital status, and educational level. These studies usually assume that matching for similarity is positively related to effectiveness of counseling process and/or counseling outcomes. For the most part, demographic variables have proven to be of little influence on either counseling process or outcome (Berzins, 1977; Higgins 6: Warner, 1975; Kruaskopf, Baumgardner, 6t Mandracchia, 1981; Parloff et al., 1978; Wright 6: Hutton, 1977). It would be incorrect to conclude that demographic variables are insignificant. As Berzins noted, many studies examined one, two, or three of these variables at a time, when it might instead be more useful to analyze simultaneous influences of more variables. In addition, demographic variables might be highly influential with particular subgroups (e.g., male clients who have been incarcerated for 10 years paired with younger female counselors) (Parloff, et al., 1978). It is unlikely that matching on demographic variables has many implications for personality matching research. Demographic variables are 15 unrelated to the construct of personality as defined earlier (Corsini, I977; Shertzer 6c Linden, I979). Demographic variables describe external aspects of people over which they have little or no control (e.g., race) or the variables describe more long-term aspects that fluctuate to a small degree (e.g., marital status). By contrast, personality factors appear through shorter-term, more voluntary behaviors. Another difference between demographic variables and personality factors is clarity and specificity of definitions. Demographic variables can be defined more exactly. A corollary of this point is that demographic variables can be measured more exactly than personality factors (Shertzer 6: Linden, I979). If personality factors could be measured as specifically as demographic factors, it is possible that the personality matching lines of research would be more advanced. Treatment Methods Matching according to treatment methods consists of identifying client variables expected to influence counseling process and/or outcomes and pairing the variables with counseling methods or styles believed to be especially helpful in bringing about positive results. Client-treatment matching requires counselors to tailor their counseling methods to their clients (Berzins, 1977; Heck 6: Davis, 1977; Thoreson 6t Coates, 1978). Client variables often used are symptomatic behavior patterns, diagnosis, prognosis, presenting problems, and expectations for counseling. In his review, Berzins (1977) stated that few client-treatment matching studies had been conducted and that the ones reviewed provided limited support for the approach. Berzins criticized client-treatment matching studies as a whole on the grounds that they (a) too seldom defined different client matching- variable groups in terms specific enough to demonstrate that the groups were l6 truly different and (b) failed to account for possible interaction effects between counselors‘ personalities and the treatment methods used. Ramifications of treatment methods matching research for the present personality matching research are limited. First, the approach gives little attention to client personality factors, emphasizing instead more extrinsic client factors. Second, the study presented here has no treatment component. Client-Counselor Similarity and Attraction A number of studies that do not deal with matching per se support the belief that compatibility or similarity between client and counselor in one or more facets of personality will increase client attraction to the counselor, which in turn will enhance their working relationship. These studies will be summarized. Their suggestions for client-counselor matching for personality similarities will be considered as well. Much of the research on counselor attractiveness has been based upon earlier research in social psychology. Huston and Levinger (1978) concluded that attraction between two people was much more likely when the people considered themselves to be similar to each other in some way(s). Huston and Levinger identified the following reasons why similarity might be positively related to attraction: (a) similarity implies higher chances of the partner's being compatible in the future; (b) similarity encourages a person to be more relaxed since her or his partner's apparently comparable behaviors, feelings, or thoughts validate the first person's experiences; (c) similarity raises a person's self- esteem; and (d) similarity is tangibly rewarding (e.g., two people exchange gifts because they like each other). In order for similarity to facilitate attraction, it is necessary to have an absence of apparent threat due to being alike. Huston and Levinger identified the following common examples of similarity-as-threat: 17 (a) anticipated competition for limited resources, (b) a sense of loss of uniqueness or individuality, and (c) reminders of personal inadequacies. Should it be demonstrated in the present study that client-counselor personality matching for similarity is associated with greater client attraction to the counselor, then Huston and Levinger's suggestion of possible links between similarity and attraction would be useful. During the past 15 years, much of the research on counselor attractiveness has been in response to a publication by Strong (1968). Borrowing ideas from social psychology research, Strong's Social Influence Model defines counseling as a reciprocal interaction between client and counselor. The counselor's influence with the client in the process of counseling and, in turn, the counselor's influence in encouraging positive client change depends largely upon the client believing the counselor to have the qualities of attractiveness, expertness, and trustworthiness (Corrigan, Dell, Lewis, at Schmidt, 1980; Strong, 1968). A number of the overt cues on which counselor attractiveness, expertness, and trustworthiness are based are related to the personality definition criteria discussed. earlier. These overt cues involve verbal and nonverbal counselor behaviors plus probable client explanations about why the counselor acts as 5/ he does. Counselor attractiveness was defined by Schmidt and Strong (1971) as the client liking and admiring the counselor to the point of wishing "to gain his approval, and . . . to become more similar to him" (p. 3‘18). Social Influence research on attractiveness has been based upon this definition. The Social Influence definition of attractiveness sounds much like the rationale for client- counselor matching for similarity, insofar as the attractiveness definition and the similarity matching rationale both emphasize striving for similarity between 18 client and counselor. As a result, the attractiveness dimension is reviewed in some depth here. Hackman and Claiborn (1982) argued that the Schmidt and Strong (1971) definition of counselor attractiveness was too simplistic. Their study supported an "attributional" explanation of counselor attractiveness and influence. An attributional explanation says that counselors who are similar to their clients will be more influential in encouraging behavior change when they disagree with their clients. Client-counselor agreement when they are similar encourages status quo in client thoughts and actions. Conversely, counselors who are dissimilar to their clients will foster more client change when they agree with their clients. Disagreement when client and counselor are dissimilar encourages clients to make no changes. The discussion of Hackman and Claiborn (I982) sounds like the complimentary-anticomplimentary explanation of Interpersonal Transactions researchers (Carson, 1969; Dietzel 6c Abeles, 1975) whose work in personality matching supported the idea of counselors purposely acting contrary to clients' preferences in order to better encourage client changes. Behaviors that appear to be related to counselor attractiveness are (a) moderate frequencies of counselor self-disclosures about personal attitudes, beliefs, experiences, and feelings (Daher 6t Banikiotes, 1976; Nilsson, Strassberg, 6: Bannon, 1979); (b) counselor self-disclosures that indicate similarities between counselor and client (Daher 6c Banikiotes, 1976; Hoffman-Graff, 1977); and (c) frequent changes in counselor nonverbal behaviors (e.g., body positions and smiling) (Claiborn, 1979; Strong, Taylor, Bratton, 6t Loper, 1971). All of these variables could be considered aspects of personality. All of the variables are also important considerations when counselor attractiveness is part of a research focus. For this particular study, the indication of similar client-counselor self- disclosures positive affecting counselor attractiveness is directly relevant to the 19 research topic. The other findings, however, must be regarded for the influence they might have on client perceptions of counselor attractiveness and similarity. Dowd and Pety (1982) investigated college male and female subjects' reactions to an audiotaped counselor. They found that the counselor's matching her or his spoken vocabulary to the subject's vocabulary had no significant impact on subjects' ratings of counselor attractiveness or on subjects' willingness to see the counselor themselves. Subjects did indicate interest in seeing the taped counselor nevertheless, which was related to unknown factors other than verbal matching. Dowd and Pety interpreted these results as indirect evidence in favor of including counselor nonverbals in research on attractiveness. Bodden and Winer (1978) and Heppner and Heesacker (1982) found that clients' ratings of counselor attractiveness, expertness, and trustworthiness increased or decreased over the course of approximately seven to eight counseling sessions, depending upon what experiences the client had with a counselor. This outcome suggests that counselor attractiveness might be intentionally altered by the counselor for the same reasons as those offered by Interpersonal Transactions researchers (Carson, I969; Dietzel 6c Abeles, 1975). Corrigan et a1. (1980) and Heppner and Heesacker (1982) reported that all counselors were rated attractive before counseling actually began. They interpreted this to be an indication of the counselors' initially having "legitimate status," which Corrigan et al. defined as the "receiver's (i.e., client's) perception of the source (i.e., counselor) as having legitimate authority by virtue of social role or position" (p. 396). It is possible that legitimate status might be a factor to consider in this research project inasmuch as only initial impressions of subjects are checked. Attempts to relate demographic variables to counselor attractiveness in a Social Influence Model context are limited and inconclusive. Evidence to date 20 suggests that demographic variables have little influence on attractiveness (Corrigan et al., 1980; Heppner dc Dixon, 1981). Results of studies matching subjects and counselors on demographic variables corroborate these tentative Social Influence Model findings. Efforts to relate pre-counseling information about counselors to their attractiveness have also been inconclusive. Goldstein (1971, 1980) conducted research and summarized research supporting the idea of increasing counselor attractiveness by providing subjects with pre-counseling information about counselors' personality, experiences, or treatment methods. Literature reviews by Berzins (1977) and Manthei (1983), however, concluded that enough favorable and unfavorable research results existed to withhold judgment about effectiveness. The tentativeness of these results has implications for research that uses pre-counseling information as a method for eliciting subjects' evaluations of counselor attractiveness. Even though research has established that subjects are attracted to counselors in the initial stage of counseling, it cannot be determined that providing pre-counseling information is an effective way to enhance attractiveness. Applications of research results from the Social Influence Model to the current study must be done carefully. The most frequently mentioned limitation about the social influence line of research is its strong reliance on analogue studies (Corrigan et al., 1980; Heppner 6t Dixon, 1981; Manthei, I983). Analogue studies in counseling are those that replicate parts of counseling in laboratory settings. These settings provide greater control over independent, dependent, moderating, and confounding variables. Analogue settings provide conditions where subjects can undergo experiences that might be questionable ethically in field settings (e.g., having subjects experience an untested counseling technique 21 that might have unknown aversive side effects). Analogue settings are also more practical than many field settings (e.g., testing 30 volunteer subjects in two days rather than waiting weeks or months for 30 counseling center clients to agree to be subjects). The greatest limitation of analogue research is its lack of clear generalization to actual counseling settings. Eventually what is learned in the laboratory must be tested in vivo for usage. Manthei (1983) raised a concern which was part of the analogue study generalization issue. He pointed out that volunteer subjects who have little interest in or expectation of entering counseling at the time of the study might respond differently to a counselor or counselor surrogate than would an actual client. How a possible difference between clients and non-clients might affect their ratings of counselors was not discussed. A second criticism of most social influence studies is that they are too brief and limited to subject evaluations after only initial contact with the counselor (Corrigan et al., 1980; Heppner 6: Dixon, 1981). Research showing that client impressions of counselor attractiveness, expertness, and trustworthiness change over the course of counseling supports this criticism (Bodden 6c Winer, I978; Heppner 6c Heesacker, 1982). Holland's Theory The purpose of this research was to explore the use in client-counselor matching research of Holland's theory of career choices (Holland, 1973; Holland 6: Gottfredson, 1976). Holland's (1973) approach has also been regarded as a theory of personality, since he held that personality manifested itself through different aspects of life, including vocational interests and choices (Osipow, 1983; Weinrach, 1979). Holland identified six personalities or typologies which he labeled as artistic, conventional, enterprising, investigative, realistic, and social. 22 His definitions of these six typologies are paraphrased in the following paragraphs. Artistically oriented people regard themselves as emotionally expressive, creative, nonconforming, and independent. They prefer activities in the visual or musical arts, wherein physical objects, symbols, or humans are manipulated to create some product. These people prize aesthetics highly. Conventional people find computational or clerical tasks, with explicitly defined rules of process, most satisfying. They are usually orderly, conforming individuals, who value business, organizational, and economic gains. Enterprising persons seek situations where being popular, verbally assertive, gregarious, and taking leadership roles are important. Verbally manipulating others so as to attain organizational or economic gains is common. Achievement and recognition are valued highly, while analytic and more individually oriented endeavors are eschewed. Investigative individuals tend to have creative, analytic, and intellectual interests, being prone to pursue mathematical or scientific careers and hobbies. They often report themselves as being scholarly, confident, and lacking leadership capabilities. Scientific philosophy and methods are esteemed by these persons. Realistic people prefer physically demanding activities that are frequently performed outdoors. Their occupational and avocational interests revolve around tools, machines, and animals. They often find social interaction as an end in itself to be unrewarding. More concrete objects and ends (e.g., money) are valued higher. M people enjoy understanding and helping others. They prefer settings where they can work with people, especially in helping, training, or informing 23 them. Social individuals usually value socially and ethically oriented issues and activities, while avoiding mechanical or scientific ones. The relationships among the typologies are illustrated in Figure l. The figure displays a hexagonal arrangement of the six types, which was indicated by correlational studies (Osipow, 1983). Greater similarities and compatibilities of the personality typologies are shown by greater proximity on the hexagon and by the correlational numbers that appear on the lines connecting the typologies with each other. Realistic AL Inveatfgtive I: I/ .- g ”‘6 a ’ 1 ‘\ I I \ I . ' lfi“ .. O l ’ ‘\ ': I ‘5 1 .' \ I o \" O ;6 r‘ H :0 | a '- .° 1 ‘ O .9 o. | . “ .3 o0.000.000.3300000‘101000000.......’.........; Artistic : .0 o. 3 s .24, u .0. O. s': '. ” I . :“s I’ '. ~~ ’v . x’ “.38 ° . . I’ “ I' \‘ Figure 1. Holland typologies hexagon (from Holland, 1973, and Weinrach, I979). 211 The six typologies just described are usually identified and defined through one of the Holland-based interest inventories—the Self-Directed Search inventory, the Strong-Campbell Interest inventory, or the Vocational Preference inventory. As all three inventories indicate, no pure types exist; all people possess elements of each. According to Holland (1973), the personality typologies develop from learning experiences. Individuals learn from their physical and social environments through direct experiences and/or through observing others. As people interact with their environments, they learn that some activities will likely be more rewarding than other activities. People also learn that they have greater aptitudes and acquired skills for some activities than for other activities. Environments providing more rewards and expectations of rewards have greater impact on shaping personalities of people than do less rewarding environments (Bandura, 1978; Holland, 1976; Drumboltz, Mitchell, 6: Jones, 1976). Cumulative experiences through the learning processes described gradually lead an individual to one or two of the typologies dominating and characterizing most individual choices (Holland, 1973). Congruence, consistency, and differentiation are three concepts of Holland's theory that have been positively related to satisfaction, achievement, and stability in occupational environments (Frantz 6t Walsh, 1972; Holland, 1973). The three concepts will be defined in the following paragraphs. Research on the importance of the concepts for occupational selections will be summarized as well. Consistency The term consistency refers to the similarities of personality types compared to other personality types in Holland's hexagon (Figure I). The distal 25 arrangements and the correlations on the hexagon indicate the amount of consistency between types. Thus, someone with a typology code that is, in order of predominance, social-artistic-investigative would be more consistent than a person with a code of social-conventional-investigative. Holland (1973) regarded high consistency to be important for effectively predicting matching typologies with satisfying occupational environments and for satisfaction in occupational environments once a match occurred. Persons with more consistent typologies should be able to identify preferred environments easier than less consistent types. This should be the case since the more consistent people could focus their attention on environmental characteristics that are compatible. Less consistent individuals, however, would have their attention divided by the somewhat contradictory characteristics of divergent environments. The recent research literature presents mixed findings on consistency. Two supportive studies will be discussed, then research with negative results will be discussed. A seven year study by O'Neil, Magoon, and Tracey (1978) on investigative types, using the Self-Directed Search instrument, identified a moderately predictive relationship between consistency and job entry, consistency and vocational ideas, and consistency and graduate education major choices. Thus, subjects with consistent personality profiles were predicted with more accuracy to enter careers, have occupational preferences, or enter graduate majors compatible with investigative types. Although the study did not focus on satisfaction per se, it can be inferred that at least for this one of Holland's six types, satisfaction was implicit in the three areas checked. Barak and Rabbi (1982) found a positive relationship between staying in college and consistency regardless of personality typology or academic major. 26 Results of this research supported the idea that consistency is directly related to predictions of stability in academic majors. Again, attraction or satisfaction must be inferred from the results, since it was to be measured directly. In addition, the authors defined the construct through subject self-reports rather than using an inventory; it may be that the different procedure for defining consistency produced a meaning in the Barak and Rabbi study that was different than an inventory-based meaning. Unlike Holland (1973), Hughes (1972) reported finding no relationship between males' career satisfaction or dissatisfaction and their consistency codes on the Vocational Preference inventory. Nafzinger, Holland, and Gottfredson (1975) noted similar results with the Self-Directed Search inventory. Erwin (1982) defined the construct through an instrument other than one that was directly tied to Holland's theory and observed no link between consistency and college major choices, changes in college major, college achievement, and standardized aptitude test scores. Spencer (1982) actually encountered a slightly negative correlation between consistency ratings from the Vocational Preference inventory and inventoried job satisfaction from a sample of university professionals. Perhaps findings such as these have discouraged interest in the consistency construct. For example, consistency is absent from Osipow's (1983) most recent review of Holland's theory, even though an earlier edition (1973) presented the construct as an important one for Holland‘s theory. Differentiation The greater the differentiation in a person's typology, the more distinct it is. Holland (1973) defined personality orientation according to the one typology the person most resembled. If the accumulated data indicated, for instance, that an individual with a social-artistic-investigative profile clearly preferred social 27 activities and environments over the other five types, then the person is said to have a well-differentiated, as well as consistent, profile. Conversely, a social- artistic-investigative individual with interests of the artistic, investigative, conventional, enterprising, and realistic areas closely approaching the social preferences in importance would be identified as having a weakly-differentiated but consistent profile. Holland (1973) considered a well-differentiated typology to be important for effectively predicting matching typologies with vocational environments. It should be easier to identify compatible environments accurately for a well- differentiated type than for a poorly-differentiated type. In addition, a well- differentiated type ought to be more satisfied with the environment chosen. In recent studies on differentiation, the utility of the dimension appears to be quite limited as a predictor of attraction and/or satisfaction. Nafziger, Holland, Helms, and McPartland (1974) used the Self-Directed Search inventory and found differentiation unrelated to satisfaction with choice of college major. Lowe (1981) conducted a longitudinal study of more than 20 years and noted no relationship between differentiation as measured by the Vocational Preference inventory and career decisions. Further, he asserted " . . . it is likely that the failure to find predicted correlations results from either inadequacies in the measurement of the differentiation construct or from inaccurate assumptions about the nature of the relationship between differentiation and decidedness scores" (p. 3148). As was the case with consistency, Osipow's (1973) discussion of the dimension as an important one for Holland's theory in 1973 was omitted 10 years later in an updated review of the theory (Osipow, I983). 28 Congruence When people actually choose environment types compatible with their own typologies, congruence is said to exist. What consistency is to intrapersonal similarity, congruence is to person-environmental compatibility. Someone with a conventional typology, for example, would be more congruent working as an account (classified as conventional-enterprising-social) than as a freelance journalist (classified as artistic-social-enterprising). As defined by Holland (1973), the congruence construct differs from the personality criteria adopted for this study. Congruence itself is a match between a typology and a vocational environment. This matching is not part of personality per se. Rather it is the process whereby personality as a typology fits with an occupational environment. Holland regarded congruence to be important for satisfaction in an occupational environment. It could also be generalized that making congruent choices predicts greater likelihood by making congruent selections in the future. Research on the construct of congruence has yielded a number of supportive findings. A series of studies by Walsh (197A; Walsh 6: Lewis, 1972; Walsh 6: Russell, 1969) indicated greater satisfaction with college major choice and with life in general in college students whose Vocational Preference inventory or Self-Directed Search inventory profiles were congruent with their selection of college majors, over those whose scoring patterns were incongruent. Morrow (1971) obtained similar results in a similar study. Osipow (1983) summarized favorable and unfavorable findings on congruence, concluding that "congruence is, in fact, one of the principle . . . outcomes associated with Holland's theory" (p. 86). One of the few unsupportive studies on congruence was conducted with college women by Spokane and Derby (1979). They functionally defined 29 congruence as an exact match between a subject's highest typology score on the Vocational Preference inventory and her college major. They noted no relationship between congruent and incongruent Vocational Preference inventory scores and measured satisfaction with college life in general. In order to have a meaningful test of Holland's theory in initial client- counselor matching, it was considered necessary to incorporate all three of the dimensions Holland (1973) said were part of environment selection—congruence, consistency, and differentiation. That is, omitting one or two of the dimensions would leave the study open to criticism that the subjects were incompletely matched because they did not clearly fit all of the criteria that Holland considered vital (Bruch, 1978). Matching Applied to Holland's Theory Bruch (1978) published an article in which he suggested that Holland‘s (1973) typologies could be effectively used to pair clients with counselors who had similar personality types. Only one published study (Cox 6: Thoreson, 1977) on matching through typologies was discussed by Bruch. That study and a study on client-treatment matching (Kivlighan et al., 1981) are reviewed in the following paragraphs. These were the only published studies found that related client- counselor personality matching to Holland's theory. The Cox and Thoreson study involved 1M1 lower division undergraduates of both sexes from general and experimental psychology courses. All subjects took the Vocational Preference inventory in order to identify their typologies. Those with well-differentiated profiles listened in small groups to three 3-minute audiotapes depicting an investigative typology, a conventional typology, and a social typology counselor. Subjects then rank ordered the counselors from most preferred to least preferred. A one-sample, non-directional chi-square analysis 30 was not significant. The social counselor type was the single most preferred presentation, regardless of subject typology. It is unclear from his article why Bruch (1978) hypothesized the possibility of client-counselor personality matching when the single published study on the topic was negative. Apparently he believed that the positive research on Holland's (1973) theory in the vocational area plus the results of his own unpublished study (Bruch 6c Gilligan, 1978, cited in Bruch, 1978) were sufficient bases for further matching model development. Another procedure Cox and Thoreson (1977) used was to have subjects complete an adjective checklist created out of terms Holland used to describe occupational environments. Terms that subjects checked to describe themselves were then compared to their individual typologies from the Vocational Preference inventory for similarity. Chi-square analysis indicated that artistic, enterprising, investigative, and social type subjects preferred environments similar to their own typologies. Conventional and realistic subjects demonstrated no preferences in any direction. Cox and Thoreson interpreted their results from the rank ordering and the adjective checklist as unsupportive overall of client-counselor personality matching for personality similarity. Most subjects apparently considered the social typology counselor to be most preferred. Cox and Thoreson inferred that the results endorsed counselors having more social type characteristics at the outset of counseling. Bruch (1978) criticized Cox and Thoreson on the basis of their excluding the concepts of congruence and consistency. Bruch and Holland thought that congruence, consistency, and differentiation all must be present in subjects in order to have a full test of Holland's theory. 31 Other short-comings are apparent as well. First, like much of Holland's earlier work, Cox and Thoreson (I977) relied upon a chi-square test of independence; this statistic is less powerful than a parametric one, such that significant differences might have been overlooked. Second, a one-tailed test which could be derived from the development of Holland's (1973) theory would have been more sensitive to hypothesized changes than the two-tailed test actually used. Third, computational errors were found in two of the six chi- square computations that were reported on subjects' counselor preferences. While none of the identified errors would have altered the conclusions of the authors, their presence suggested the possibility of inaccuracies in other parts of the computational or data reporting processes that could have changed the statistical outcomes. Finally, only three of the six typologies were represented on the audiotapes, yet all six types of subjects were included. It could be that the cumulative result of these factors—excluding congruence and consistency, using less powerful statistical tools, using nondirectional hypotheses, and possibly committing computational errors limited typologies representationuoccluded potentially important relationships. Kivlighan et al. (1981) studied occupational counseling groups comprised of male and female undergraduates. Through the Vocational Preference inventory, subjects were categorized as either more people oriented (enterprising and social types) or more task oriented (investigative and realistic types). Forty subjects were randomly assigned to one of two groups, balanced for numbers of subjects, gender, and subject orientation toward people or tasks. One group emphasized vocational counseling through numerous interactions of the participants (i.e., more people oriented). The second group emphasized vocational counseling as an individual problem-solving effort (i.e., more task oriented). Both groups met twice. Analyses of covariance on subject' responses to an instrument (the Career 32 Maturity inventory) designed to check the thoroughness of occupational decision making indicated that the subjects matched with treatments similar to the subjects' orientations (a) demonstrated greater measured career maturity or thoroughness in decision making, (b) displayed more career information seeking, and (c) enjoyed the group sessions more. Results of the Kivlighan et al. (1981) study have limited application to the present research, since their focus was subject-treatment matching rather than subject-counselor personality matching. Counselor personality is part of an environment just as treatment is part of an environment, but treatment and counselor personality are nevertheless different variables. Further, Kivlighan et a1. matched subjects with group counseling treatments solely on the basis of the subjects' highest Vocational Preference inventory typology score. The published report said nothing about congruence, consistency, or differentiation. If Holland's (I973) assertion that congruence, consistency, and differentiation "are needed to apply the theory to practical problems" (p. 37) is correct, then research that disregards these constructs is inadequate. The more recent results from studies on consistency and differentiation suggest that these two dimensions might be deleted from research on Holland's typologies. Results of studies on congruence, however, indicate that it should be included in research on the typologies. Summary Research results reviewed pertained to (a) client-counselor relationships, (b) the various major approaches to matching, (c) the relationship 1 between client-counselor similarity and attraction, especially according to the Social Influence model, (d) Holland's (1973) theory, and (e) attempts to use Holland's typologies in matching. Each section also considered implications of the 33 research results for the current project's study on client-counselor matching of personality similarity at initial contact. A definition of the therapeutic relationship was discussed in terms of client-counselor similarity, attraction, and reward potential. Two models for client-counselor personality matching were also discussed. The interpersonal transactions approach had enough empirical support to make it a productive line for further research. The history of the A-B approach was traced, with the conclusion that cumulative research results do not support the model. Review of client-counselor attraction was organized primarily along Strong's (1968) Social Influence model, since it provided a conceptual framework and a well-developed line of research. Implications of the attractiveness aspect of the Social Influence model for client-counselor personality similarity were discussed. Holland's (1973) theory and more recent research related to the theory were reviewed. Ways in which the definitions and research results related to client- counselor personality matching for similarity were identified. Only two published studies that used Holland's typologies in matching were found and reviewed. One study (Cox 6: Thoreson, 1977) failed to support the client-counselor personality matching hypothesis. It was demonstrated, though, that the study contained weaknesses that might have obscured possibly positive results. The second study (Kivlighan et al., 1981) supported the use of Holland's typologies in matching. However, the study used a client-treatment matching for similarities approach, which was different in orientation than client- counselor personality matching for similarity. Inasmuch as client-counselor preferences do occur and apparently have a bearing on counseling process and results, it seems essential that they be 34 Some evidence indicates a facilitative link between client preferences for counselors and development of at least early stages of counseling relationships. If such is the case, then it seems reasonable to assume that a matching model would be helpful. All of the past matching approaches have met with limited success for reasons that are sometimes unclear. Some major possibilities can be identified, though. First, it might be that past matching models are conceptually flawed or are too inadequately developed to enable empirical validation. A second consideration is that the matching studies examined here had research design or methodological problems which compromised their findings. Design weaknesses occur in many matching studies. Reviewers of those studies (Berzins, I977; Parloff et al., 1978) agreed that factorial designs were necessary rather than the usual single factor approach, in order to identify possible interaction effects. Further, Berzins (I977) recommended using multiple dependent measures so as to better identify any patterns of counseling outcomes or processes and to have data from different perspectives. The research that is detailed in the following chapters is meant to address questions and concerns raised by past matching studies. The present research is also a blending of three major lines of investigation that have been only tangentially linked together in past studies—client-counselor personality matching for similarities, Holland's (I973) personality typologies, and Strong's (1968) Social Influence model. It is believed that these efforts will contribute to learning more about the therapeutic relationship in ways that will benefit client and counselor. CHAPTER III METHODOLOGY Research Hypothesis Chapter I reviewed Bruch's (1978) contention that using Holland's (1973) typologies to match clients and counselors on similarity of personality would result in greater client attraction, satisfaction, and duration with a counselor. The review in Chapter I also established that Bruch had little direct research support for his contention. Chapter II reviewed literature which indicated that (a) Holland's theory effectively predicted greater attraction, satisfaction, and duration in occupational environments similar to the subject's personality orientation, (b) client-counselor matching for personality similarity had a long and generally negative history in predicting counseling outcomes and/or processes, and (c) the Social Influence model of counseling supported the idea of client attraction to a counselor who seemed similar to the client. Combined information and inferences from these strands of research suggest that Holland personality typologies might provide a successful basis for client-counselor matching for personality similarity. This project tested Bruch's thesis of greater subject attraction to counselors with Holland's personality typologies similar to the clients' typologies. Specifically, it was hypothesized that for subjects with congruent, consistent, and well-differentiated Holland typology profiles on the Vocational Preference Inventory, the more similar the subject's personality is to the personality of a 35 36 counselor portrayed through an audiotape, the greater the subject's expressed preference will be for that counselor. Research Design In order to test the research hypothesis, 36 subjects were each exposed to six audiotapes. Each audiotape represented a different Holland (1973) personality type. The 36 subjects also represented the six Holland personality types, with six subjects per type. The experimental task required each subject to indicate her/his preference for each of the six counselor audiotapes. A 6x6 Latin Square was used to control for order effects in the repeated measures design (see Figures 2 and 3). Thus, six orders of audiotape presentations were crossed with the six groups of subjects representing the six Holland personality types. Subjects were randomly assigned to audiotape playing order within subject personality type. Subject Primary ijology Choice S A I R C E A ICARES I u i d l i SICARE 1 O l t . a ESICAR p 0 e r d RESICA P e l r a ARESIC y A 1 n CARESI g l l Figure 2. Latin Square design. A = artistic typology; C = conventional typology; E = enterprising typology; I = investigative typology; R = realistic typology; S = social typology. 37 38 After hearing all six audiotapes, each subject indicated on a six point Likert scale (Figure ‘1) the extent to which s/he would want to choose or avoid each counselor; a rating of six indicated the strongest preference. These data were sued to form three dependent variables. The first dependent variable was the rating the subject gave to the audiotape which exactly matched the subject's personality type (i.e., highest interest score on the VPI). The second dependent variable was the mean rating the subject gave to the two audiotapes most similar to but not the same as the subject's exact personality type* (i.e., second and third highest interest scores on the VPI). The third dependent variable was the mean rating the subject gave to the remaining three audiotapes. The repeated measures design in this project controlled differences between subjects when comparing preferences among counselors (Meyers 6c Grossen, I974; Winer, 1962). This procedure reduced the likelihood of confounding variables entering the contrasts of interest and at the same time improved statistical power by removing between-subject difference from the error term of the analysis (Howell, 1982). The Latin Square design controlled for a major source of confounding in repeated measures research: order or sequence effects from multiple treatments (Campbell 6: Stanley, 1966; Kirk, 1972). The primary order effects of concern in this research were primacy and recency memory impact and practice effects. The Latin Square equally distributed order effects across all six subject groups and audiotapes. *A subject qualified as consistent even though her/his second highest or third highest VPI score was not immediately adjacent to the typology with the highest VPI score. If either the second or third highest score was not immediately adjacent to the highest typology, it could be no more than one typology apart from the highest typology. 39 Code Number: Counselor Choice Rating Scale Directions: The six letters under the heading "Recording Letter" represent the six different tape recordings you have just heard. Under the heading "Your Choice" and in line with each of the recording letters are six numerals, indicating the degree to which you would wish to choose or avoid each counselor, given the opportunity to start seeing any one of the six for career-oriented counseling. The meaning of each numeral under "Your Choice" is as follows: 6 = strongly prefer 5 = prefer 4 = slightly prefer 3 = slightly avoid 2 = avoid l = strongly avoid Circle the numeral that best indicates your personal selection for each of the six counselors. These are to be entirely your own choices; do not discuss them with anyone else. Recording Letter Your Choice A l 2 3 £1 5 6 B l 2 3 (l 5 6 C l 2 3 £1 5 6 D l 2 3 4 5 6 E l 2 3 4 5 6 F l 2 3 ll 5 6 If this were an actual counselor choice situation, how interested presently would you be in receiving vocational counseling? 1 2 3 ll 5 Very Very Disinterested Interested Figure 4. Counselor choice rating scale. #0 Subject Selection In order to qualify to be a subject, a volunteer must have had a typology profile that was congruent, consistent, and well-differentiated according to Holland's (I973) basic definitions. Including all three criteria addressed Bruch's (1978) and Holland's recommendation that client-counselor matching research based on Holland's theory have congruence, consistency, and differentiation in order to be truly representative of Holland's theory. One problem in testing Holland's constructs is a lack of functional definitions of congruence, consistency, and differentiation (Spencer, 1982). Nevertheless, functional definitions for them were developed by the investigator for the current undertaking. The definitions were also used as subject selection criteria. These definitions appear in the following paragraphs. Congruence was defined in terms of a subject's dominant typology being compatible with that subject's occupational environment (Holland, 1973). People in vocations fitting their modal typologies qualified as congruent. It should be noted that congruence was not regarded as a facet of personality and was, therefore, irrelevant in examining the possibility of client-counselor matching for personality similarity. Congruence was a selection criterion, however, because Bruch's (1978) interpretation was that congruence had to be present in order for a typology to be valid. Such a stimulation is unclear from Holland It was decided to adopt Bruch's interpretation so that it could not be argued that subjects' typologies were inadequately defined. One criterion for assessing congruence was matching a potential subject's Vocational Preference Inventory profile to occupational listings by typology codes found in Appendices B and D of Holland. If a particular work area was not in either appendix, a second criterion source was the description of typologies and occupational environments in Holland. A person was also accepted as (ll congruent whose occupation was incongruent, but whose career training and/or college academic major was congruent and whose verbalized vocational choice was congruent as well. This accommodated individuals who had taken undesired jobs as a temporary measure or merely as a means of sustenance until more appropriate employment developed. Holland (1973; Holland 6r Gottfredson, 1976) also suggested using academic major as supporting information for making decisions about congruence of occupational choice. Acceptable consistency required that of the three tOp typology preferences expressed by the subject on the Vocational Preference Inventory, at least two must be immediately adjacent on the Holland hexagon (Figure l) and the third no more than once removed from the highest typology score on the Vocational Preference Inventory. For example, a conventional-realistic-social profile would be accepted as consistent, whereas conventional-realistic-artistic and conventional-social-investigative would not. This decision was based on the observation that correlations between types once removed on Holland's hexagon were usually just slightly lower than coefficients for adjacent types, but twice removed types correlated much lower. Holland (1978) stated that in scoring the Vocational Preference Inventory, two or more raw scores points must separate any two of the typologies in order to consider them differentiated. On the interpretive profile sheet for the test (Appendix A), these two raw scores translate into approximately five standard score points. As a result, acceptable differentiation for'this project was defined as a Vocational Preference Inventory score on any one of the six typologies that was five or more standard score points higher than the next highest type(s). Using standard scores rather than raw scores provided more uniformity in defining the differentiation criterion. 42 Obtaining subjects who qualified on the three criteria of congruence, consistency, and differentiation was a major obstacle. Data collection originally began with community college students who were enrolled in identified vocational training programs (e.g., nursing) and in general studies with declared majors. VPIs for the first 24 respondents, however, indicated that only 2596 of them qualified to hear the audiotapes. The greatest sources of disqualification for this sample were lack of congruence or consistency, as each of these variables accounted for 11296 of the nonqualifiers. With such low preliminary returns, it was thought impractical to continue working with that group. A sample of graduate students in a pilot study for this research produced a higher qualification rate, albeit only 30% of the pilot subjects would have qualified. In addition, vocational research indicated that greater occupational maturity was positively associated with more clear-cut measured career interest patterns (Gottfredson, 1979; Holland, 1973; Tolbert, 1980). Therefore, it was decided that a potentially more vocationally mature sample would be sought. That sample consisted of the university and church groups that served as the study subject pooL Thirty-two percent or 47 of the volunteers from the two settings combined produced VPIs that qualified them for the second part of the study. This percentage of qualification was higher than the community college sample, as predicted. The differences in rate of qualification for the three groups tests, though, were negligible: community college students' sample-2596, counseling graduate students' sample (pilot study)-3096, and research subjects' sample-- 32%. The variables presenting the church and university volunteers from qualifying were different than those of the community college volunteers. A lack of sufficient differentiation was the single greatest obstacle for qualifying 43 the church and university volunteers, with 35% of the nonqualifiers falling into that category. The next largest categories for nonqualification were (a) lack of congruence and (b) lack of congruence and differentiation combined; each of these account for 20%. Thirty-six subjects were selected from the #7 qualifiers, with six in each of the six Holland typologies. The Holland typology scores on the VPI for the 117 qualifiers were as follows: artistic--ll, investigative-10, enterprising-J, realistic-J, conventional--6, and social--6. Qualifiers from those typologies with more than six representatives were randomly selected to be subjects. Having equal numbers of each possible typology was judged important in light of the research design used and in an effort to secure a balanced design. Demographics of Sample The 36 subjects were volunteers from the Seattle, Washington, area. Seventeen were faculty or professional staff members at a small, private university. Nineteen were members of a local Protestant church. Each of the subjects was an active member of a Protestant denomination. It is probable that the sample represented a disproportionately high level of religious practice and identification with Protestant theology in comparison with the population as a whole. All subjects volunteered to take part in response to one of two letters that were distributed to everyone on the faculty or professional staff of the university (Appendix B) and to members of the church that was involved (Appendix C). The university letter was distributed just prior to an annual autumn faculty retreat. When an insufficient number of people for the study developed out of the university setting, church members were contacted. The church contact came seven months after the university sample was first sought. All Subjects ranged in age from 21 years old to 62 years of age. The mean age was 39.88 (SD = 10.02). All subjects were racially white. Twenty were males and 16 were females. All were citizens of the United States. All were in income groups with annual earnings ranging from approximately $12,000 to $25,000. The 36 subjects were high school graduates. Thirty held baccalaureate degrees, 11 had earned doctorates, and four were graduates of master's degree programs. The sample chosen was appropriate for the purposes of this research. First, no characteristics of the sample disqualified it as far as the theory or research for Holland (1973) were concerned. Second, the 36 subjects met the criteria set for congruence, consistency, and differentiation. Nevertheless, the demographic variables of education and religious orientation restrict generalizations. Instrumentation The Vocational Preference Inventory (VPI), developed by Holland (1978), was the instrument used to identify subjects' personality types. As a vocational interest measure and a personality inventory, the VPI has well established norms, reliability, and validity, having been in continuous use and revision (Holland, 1973, 1978). Two other measures, the Self-Directed Search inventory (Holland, 1972) and the Strong-Campbell Interest inventory (Campbell, 1977), have been designed to codify examinee responses according to Holland's typologies. The former is a relatively new test and is still being revised, while the latter represents a revision of the Strong Vocational Interest Blank; neither is currently as well established through researcher as the VPI in identifying the Holland personality types. (+5 The Counselor Choice Rating scale (Figure (I) was developed for the study to identify subject preferences for different types of counselor audiotape presentations. Subjects were instructed to circle the number corresponding to their preferences for an audiotaped counselor presentation. They were also instructed to circle the number at the bottom of the page, showing the degree to which they were presently interested in actual career counseling; this rating was included to obtain an assessment of the relevance of the counselor rating task for each subject. Subjects more interested in vocational counseling at the time of the rating might have responded differently than those less inclined. Scripts and Audiotapes Development This study focused on a situation similar to the first few minutes of client- counselor contact. Subjects were to react to six different counselor presentations as if they were meeting the counselor for an initial, individual, career counseling interview. Depiction of an initial session was considered necessary in light of findings that client-counselor relationship changes occur in different stages of counseling (Bodden 6c Winer, I978; Dietzel 6r Abeles, 1975). In addition, it has been suggested that counselor attractiveness might be most influential at the outset of counseling (Bodden 6t Winer, I978; Heppner dc Heesacker, 1982). Hence, the possible attraction of client-counselor personality similarity ought to be easier to detect in the entry phase of counseling. The hypothetical meetings between counselor and subjects were structured as one-to-one sessions in order to be consistent with most past efforts at client- counselor matching research. Simulations of group counseling could also have introduced concerns or questions extraneous to the issue of client-counselor similarity (e.g., the typologies of group members who would also be interacting with the subject). 46 It was judged that presenting the hypothetical meeting as a vocationally oriented one would keep the focus of the responses closer to the content of Holland's 1973) theory, which was the basis of the VPI and the audiotapes. It was also thought possible that subjects might respond differently to interviews portrayed as being chiefly concerned with personal, occupational, or academic issues. It seemed that clarity of the outcomes would be greatest by explicitly associating the subject choice with just one general type of counseling. The audiotapes were based upon six scripts (Appendix D) read by the same male, a local middle school teacher who volunteered to make the recordings and who was uninformed about the nature of the study. Audiotapes were chosen to control for counselor physical appearance, which has been shown by some researchers to be a salient feature in early client-counselor contacts (Carter, 1978; Cash et al., 1975; Cash ,6: Salzbach, 1978) and would be a possible source of confounding. Counselor age might influence preference ratings (Helms 6r Simons, 1977), and audiotapes controlled this factor as well. Counselor gender may or may not influence initial preferences since the literature is mixed on the effects of this variable (Helms 6t Simons, 1977; Johnson, 1978). Nevertheless, any gender bias ought at least to be systematically spread across all ratings by employing the same stimulus tape counselor for all six recordings. Each script was recorded on a separate cassette. This allowed variable orders of presentation and thereby avoided any potential bias from the order of tape presentation. Each script was structured according to guidelines for informed consent procedures for counseling sessions suggested by Stewart, Winborn, Johnson, Burks, and Engelkes (I978). The hypothetical counselor in each script presented his views on the purpbses of career counseling, the responsibilities of client and counselor, the focus of the issues with which they could expect to deal, and 47 limits of their relationship (e.g., length and numbers of counseling sessions and confidentiality). These four issues appeared in each script in the order just listed. Structuring the scripts according to informed content guidelines was useful in two ways. First, the general pattern of organization suggested by Stewart et a1. (1978) facilitated development of a standard format for all scripts. The standard format removed one possible source of confounding wherein subjects might react to different audiotapes partially as a result of the order of information given. A second value of using informed consent guidelines in their wide-spread use in counseling. It has been recommended that all counselors provide clients with informed consent opportunities during the first few sessions (Brammer & Shostrom, 1982; Hare-Mustin, Marecek, Kaplan, & Liss-Levinson, 1979). An informed consent format in the audiotapes provides a situation comparable to what should occur in an initial counseling session. Attempts were also made to control for other potential sources of confounding in the scripts. All six tapes were approximately the same length, being four to five minutes long on the recording, with a mean tape length of ‘1 minutes, 47 seconds. The scripts were written so that the complexity of language was equivalent in all of them as well. Copies of the scripts appear in Appendix D. The intent was for the audiotapes to differ only on the Holland-based traits the counselor used to describe himself, his views on counseling, and the counseling process. Each script emphasized features attributed to its particular typology by Holland (1973). Thus, what raters should have responded to on each stimulus tape was the vocational counseling environment that the hypothetical counselor verbally presented. 48 Three Seattle Pacific University master's level counseling students who had recently completed a vocational counseling class judged the tapes as to the sort of typology each depicted. The raters had been introduced to Holland's theory in the class. In the 30 minutes before they heard the six audiotapes, they also received an instructional overview of Holland's theory from the investigator. The presentation consisted of a verbal summary of the theory, accompanied by overhead visuals; terms used to describe counseling environments in the audiotapes were not part of the overview. Immediately after the instruction, the raters listened to the six tapes. When all of the tapes had been played once, the raters independently wrote their opinions as to which tape represented which typology. Figure 5 displays the form used by the raters. Rater name Date Directions: After each numeral under the "Tape Number" column, place the code letter for the Holland typology you believe it represents; i. e., A for artistic, _C_ for conventional, E for enterprising, _I_ for investigative,_ R for realistic, or _S_ for social. Only one code letter should appear for each— tape. Make your selection after listening to all of the tapes. Please avoid conferring with other raters as this could bias your choices. Tape Number Typology Code I 2 Figure 5. Holland typology identification. 49 The intraclass correlation coefficient (Scott 6: Wertheimer, 1962) was used as an index of interrater agreement (i.e., the extent to which the raters made the same exact judgments). The coefficient 0.80 was adopted as the acceptable minimum level of agreement, since it was a common criterion for interrater agreement in complex rating tasks (Hersen 6c Barlow, 1976). Calculation of the coefficients yielded agreement levels of 1.00 for the three judges with the established criteria (see Table I). With such exact rates of agreement, it could be argued that the tapes clearly depicted what they were meant to present. Table I Results of Intraclass Correlations Over Holland Typologies Audiotapes Tape 1 Tape 2 Tape 3 Tape 4 Tape 5 Tape _6 Rater l j T 1 T T j + + j + j Rater 2* + l + + I + + + 1 i T ' 1 Rater 3 + . + + + + + i R1 = 1.0 R2 = 1.0 R3 =1.o Note: R1 = rater 1; R2 = rater 2; R3 = rater 3; + = rater correctly identifying Holland typology. Study Implementation The investigation began when volunteers were sought in the manner described under the subject selection section. Each subject received a VPI test booklet, seventh edition, a VPI answer sheet, and a research informed consent sheet (see Appendix E) plus instructions on how to complete the forms individually and privately at her/his convenience. All forms were returned to the 50 researcher who stored the informed consent sheets and scored the VPIs with scoring stencils. Those volunteers who qualified for the project by having congruent, consistent, and well-differentiated profiles were contacted by telephone by the researcher to learn if they were willing to hear the audiotapes. A time and place for the listening was set up with people agreeing to continue. All of those who were contacted agreed to hear the tapes. Subjects rated the audiotapes singly or in groups of two to five. Nine people rated them alone, twelve rated them in dyads, six in triads, four were in a foursome, and five rated in a group of five. All audiotape presentations were monitored by the investigator who gave the same verbal instructions and insured that ratings were independent. The tapes were presented and evaluated in quiet surroundings, with only the researcher and subject(s) present. No interruptions or distractions occurred. Each tape was played once, with the actual rating coming after the conclusion of the sixth tape; these tasks were completed in one session that lasted 25 to 30 minutes. A number of differences occurred among the times when the tapes were rated by subjects. Subjects heard the tapes at various locations--private homes, an office, and conference rooms. They heard them at various times of the day, although all were after noon and before 8:00 pm. Once each individual or group finished rating the audiotapes, a debriefing on the study took place. Each subject was also offered an opportunity to receive an interpretation of her/his own VPI results. Lastly, all subjects were asked to refrain from discussing the project with anyone else, in order to avoid possibly influencing subjects who had not yet heard the tapes. Everyone agreed to maintain this confidence. 51 Pilot Study Prior to data collection, a pilot study was conducted to check the procedures. Ten masters' students in counseling from Seattle Pacific University who had not taken a vocational counseling class, volunteered to take the VPI and rate the audiotapes. As a result, it was determined that the study could proceed as originally planned and as described in the study implementation section. Only one change was made to the original plan of the study. It was assumed to be necessary to play the six audiotapes twice in order to ensure subject recall of their contents. Participants in the pilot study, however, reported no difficulties recalling the basic contents of the tapes and identifying their own counselor preferences after hearing the six tapes played once. It was, therefore, deemed unnecessary to play the tapes more than one time. Statistical Analysis Procedures The data were analyzed in three parts. First, the research hypothesis was analyzed through a three-way repeated measures analysis of variance. Second, the anticipated direction of the research hypothesis was checked through planned comparisons. Third, a correlation between the highest typology score and interest in occupational counseling was conducted. The three-way repeated measures analysis of variance provided a comparison of means across the three independent variables plus a check for possible interactions (Howell, 1982; Meyers 6r Grossen, 1974). The independent variables were subject typology, counselor typology, and audiotape playing order. The analysis of variance tested these two null hypotheses: H1:um - (1102 + 1103 )/ 2 H2: 1102 = 1103 where “D1 represented the mean of the first dependent variable (the Likert scale rating given to the audiotape exactly mating the subject's own typology), “D2 52 represented the second dependent variable (the mean rating given to the two audiotapes most similar to but not exactly the same as the subject's own typology), and “D3 represented the mean of the third dependent variable (the mean rating given to the three remaining audiotapes) (see Figure 3). Statistical support for the research hypothesis would reject the null hypotheses, finding instead that: ”Di ’ uDz ’ uD3 Cumulative research on Holland's hexagonal model indicated that individuals with congruent, consistent, and well-differentiated typology profiles would most likely select congruent counselor environments for themselves. Therefore, a one-tailed test of differences was appropriate (Borg 6r Gall, 1983; Glass 6: Stanley, 1970). The alternate hypotheses were: ”1 ‘ “Di ' (“02 I “031',> 0 ”2‘ “02 ' “D3 ’ 0 That is, it would be hypothesized that subjects most preferred a counselor whose personality type is the same as the subject's. It was further hypothesized that subjects would prefer counselor personality types more similar to, although not identical to, their own personality types over dissimilar counselor personalities. A Pearson product moment correlation was used as an index of the size of the linear relationship between how interested each subject was in receiving actual vocational counseling and the Likert rating given to the counselor tape representing her/his most preferred typology from the VPI (see Figure 4). A high correlation would suggest a link between current readiness or desire for occupational counseling and preference for the hypothetical counselor. 53 Summary In Chapter III, the purpose of the study was integrated with results of prior research on client-counselor matching and a research hypothesis was stated. This hypothesis was "for subjects with congruent, consistent, and well-differentiated Holland typologies, the more similar an audiotaped counselor's personality was to the subject's personality, the greater would be the subject's preference for the counselor." A Latin Square with one subject per cell was used to obtain a balanced design and to control for numerous potential sources of confounding. A three- way repeated measures analysis of variance was used to test the main and interaction effects associated with the three independent variables. Two planned comparisons were used to test the direction of preference indicated in the research hypothesis. A Pearson product moment correlation was used to test the relationship between vocational counseling readiness and counselor preference. The characteristics of the 36 subjects were identified, as were the methods for obtaining the subjects. The development of the six stimulus audiotapes was described. The dependent measures from the Counselor Choice Rating Scale were also described. Finally, the procedures through which the study and a pilot study were implemented were described. CHAPTER IV RESULTS Study Purpose and Design Review This study investigated the utility of Holland's (1973) typologies in client- counselor matching for personality similarities. It was believed that Holland's (1973) theory of matching individuals' personalities with occupational environments could be applied to client-counselor personality matching (Bruch, 1978). The study directly tested this thesis through setting up a hypothetical situation in which subjects rated verbal presentations of an audiotaped vocational counselor, depicting the first few minutes of contact in a counseling session. A six-by-six Latin Square was used to control for order effects in the repeated measures design that had each subject indicate preferences for each of six audiotapes of hypothetical counselor presentations. Three different statistical procedures were used to analyze data from the counselor choice rating scale (see Figure 4). First, a three-way repeated measures analysis variance was used to test the main effects and interactions among the independent variables. Second, two onetailed planned comparisons _t-tests for dependent samples were used to check the research hypothesis specifically. Third, a Pearson product moment correlation was used to check the possibility of a connection between subjects' present readiness for vocational counseling and their ratings of audiotaped counselors. 54 55 Statistical Results Table 2 presents the data analyzed. In each cell the uppermost numeral is the rating a particular subject gave to the tape matching (i.e., the most similar tape) the subject's own highest typology score on the VPI. The middle numeral is the mean of the subject's ratings of tapes matched with her/his second and third highest VPI scores (i.e., the tapes most similar to but not exactly matched with the subject's own typology). The bottom number in a given cell represents the mean of the subject‘s ratings of the remaining three tapes. Table 2 Latin Square with Data Subject Primary Typology Choice S A I R C E 4 5 5 4 6 3 3 6 2 5 3 6 2.5 4 3 3 4 4.33 3 ICARES A 6 4 5 6 4 5 U 4.5 3 1.5 4.5 5.5 6 D 4.33 4.33 2.66 4.33 4 2.66 I SICARE O 4 5 5 6 5 5 T O 3.5 2.5 4.5 2 3 5 A R 3 4 3.33 5.33 2.66 5.33 P D ESICAR E E 5 5 5 2 5 3 R 3 4.5 2.5 3.5 5 5.25 P 3.66 4.33 4.33 5.66 4 5.5 L RESICA A 6 2 5 3 4 5 Y 3 3.5 4 4.75 4 4.25 I 3.66 4.66 3.33 4.5 5 4.83 N ARESIC G 6 4 4 2 3 5 4 5 5 4.5 3.5 3.5 3 3.66 4.33 3.33 3.33 3 CARESI NOTE: A = artistic; C = conventional; E = enterprising; = investigative; R = realistic; S = social 56 The sources of variation for the three-way analysis of variance appear in Table 3. No statistically significant main or interaction effects were found, using a 0.05 Type I error. Table 3 Three-Way Analysis of Variance Sources of Variation Sources of Variation S_S gf _M_S E 5 2.433 5 0.4870 0.32 D 4.493 2 2.2465 1.48 O 1.393 5 0.2786 0.18 SD 13.220 10 1.3220 0.87 SO 18.530 25 0.7400 0.49 DO 14.647 10 1.4700 0.97 SDO 76.430 50 1.5200 NOTE: S = subject typology; D = distance from most preferred typology; O = order of audiotape playing The analysis of variance was modified by reducing by one half the degrees of freedom in Table 3 when they were used to locate critical values in the _E Distribution Table for the distance (D) variable. This was done to provide a conservative test of the calculated _F values. A conservative test of the D variable was a compensation for not directly testing an assumption that the intercorrelations of the three levels of the D variable were equal. Two planned comparisons specifically checked the study's research hypothesis. Table 4 presents calculations for: 57 The three means being compared were: Xm = 4.47 XD2 = 3.92 XD3 = 3.95 It can be seen in Table 4 that the first contrast was significantly different from zero and in the predicted direction, whereas the second contrast was not. Both were tested with a 0.05 Type I error. Table 4 Planned Comparisons for Distance Variables Planned Comparison _SQ 1 gf _t_ v] > um - (1102 + 1103) 1.61 0.6100 35 2.2600* 112 > “02 — 1103 1.35 000112 35 -0.0187 :2 < 0.05, one-tailed. The possibility of a relationship between subjects' expressed interest in receiving occupational counseling and how they rated the counselor typologies was tested through a Pearson product moment correlation. The 0.0125 coefficient was nonsignificant, with a 0.05 Type I error. The low correlation plus the noncommittal vocational counseling interest ratings (_)_(_ = 2.97, _S_Q = 1.16 on a five point Likert scale where a 5 indicated maximum interest) made by the subjects intimated little bearing of counseling interest on counselor preference. Summary of Results Subjects' ratings of the audiotapes provided some support for the thesis that client-counselor matching for similarity along Holland's six personality types would lead to differential client choices in favor of that similarity. The 58 three-way analysis of variance was conducted to check for main and interaction effects. The first planned comparison--subjects' preferences for counselor typologies most similar to their own typologies—~was statistically significant whereas the second comparison was not. The Latin Square design and statistical analyses for repeated measures provided strong tests of the research hypothesis. The Pearson product moment correlation was nonsignificant. The smallness of the correlation suggested that current readiness for vocational counseling had little influence on counselor preference ratings. CHAPTER V CONCLUSIONS, DISCUSSION, AND RECOMMENDATIONS Conclusions The present research tested the applicability of Holland's (1973) personality typologies to client-counselor personality matching. Specifically, the study tested Bruch's (1978) thesis that clients and counselors could be effectively matched for similarity along Holland's six personality typologies and that clients would be more attracted to counselors with personalities similar to the clients'. Few studies have investigated the possibility of using Holland's typologies in matching. This study built upon those efforts, adding a stronger research design and statistical analyses than previous attempts. These improvements allowed greater control over sources of confounding and more exact tests of the research hypothesis. A planned comparison supported the hypothesis that subjects most preferred audiotaped counselors whose typologies exactly matched their own. A second planned comparison did not support the hypothesis that subjects preferred audiotaped counselors whose typologies were more similar, although not identical, to the subjects' own highest rated typologies. These results are more encouraging than the history of client-counselor matching attempts in general. The results are especially encouraging in light of the brief, hypothetical nature of the client-counselor contact. Subjects' preferences for counselors exactly matching the subjects' own typologies support previous findings of a subject's being more attracted to a counselor whose self-disclosures indicated similarities with the subject (Daher 6: 59 6O Banikiotes, 1976; Hoffman-Graff, 1977). Any or all of the possible reasons for attraction through similarity proposed by Huston and Levinger (1978) might explain the results: (a) similarity suggesting greater chances of future compatibility, (b) similarity fostering greater relaxation, (c) similarity bolstering self-esteem, and (d) similarity being tangibly rewarding. Some of Huston and Levinger's suggestions seem more applicable here than others. Subjects' ratings were based on only a few minutes of information about a hypothetical counselor. Thus, explanations that tie more into actual client- counselor interactions or exchanges are less plausible. The possibility of a tangible reward would not be a factor, as subjects were told they would simply hear an audiotape, not interact with an actual counselor. Impressions of future compatibility also seem unlikely, since subjects were informed of the hypothetical nature of their listening to and rating the counselor presentations. It is possible that subjects felt more relaxed with presentations representing counselor typologies most similar to their own. Subjects might have felt a boost in self-esteem as well. A subject might have experienced either or both situations by perceiving that the counselor presentation most similar to the subject's own personality made sense, was comfortable, or was non-threatening. These and other possible explanations are speculative, of course, as neither relaxation nor self-esteem were measured. A question remains as to why the second planned comparisonflsubjects preferring counselor presentations more similar to, but not exactly matching, the subject's own personalities--was nonsignificant. Many possible reasons exist. Given the possible reasons for perceived similarity leading to attractiveness that were discussed by Daher and Banikiotes (1976), Hoffman—Graff (1977), and Huston and Levinger (1978), the issue might be one of salience of perceived counselor personality similarities. Where an exact match between subject and counselor 61 typologies occurred, the similarities between them might be easy to perceive within four to five minutes. As the similarities become fewer, however, they might be harder to detect in a short period of time. As a result, subjects would notice little difference between counselor typology presentations that were not exact matches. The final statistical analysis was a correlation between subjects' present interest in receiving vocational counseling and how they rated the counselor types matching their own typologies. The low coefficient from this test suggests that no relationship existed between subjects' readiness for occupational counseling and their preferences for counselors. Implications and recommendations for dealing with this result will be presented in the methodology portion of this chapter. Discussion and Recommendations The mixed results of this study have a number of implications. These implications are organized and discussed in four sections. The sections contain suggestions for further research with Holland's (1973) typologies in client- counselor matching for personality vis-a-vis (a) matching studies in general, (b) Holland's theory, (C) the social influence model, and (d) the methodology of this project. Each of the sections will integrate outcomes of this present study with pertinent literature, in order to better understand the former. It was necessary to control for a number of variables that are often influential in research on human behavior. Differential client-counselor choice preferences by gender or race have generally been found to be equivocal but might, nevertheless, have an affect on results (Carter, 19781; Comes-Schwartz, Hadley, 6c Strupp, 1978; Higgins 6: Warner, 1975; Strupp, Fox, & Lessier, 1969). Cumulative research also indicated the reliability and validity of Holland's (1973) 62 theory and instruments across both genders, across major racial groups in the United States, and across divergent socioeconomic levels (Alston, Wakefield, Doughtie, 6r Bobele, 1976; Holland 6: Gottfredson, I976; Salomone 6: Stanley, 1978; Wakefield, Yom, Doughtie, Chang, 6: Alston, 1975). It is likely that these factors have few implications for the results of this study since they were not variables and will not be discussed. Matchifl Studies Finding support for a matching model based on client-counselor personality similarity goes against the trend of results from other personality matching studies (Berzins, 1977; Parlott et al., 1978; Razin, 1977) and on Holland's typologies (Cox 6: Thoreson, 1977). It cannot be determined why other personality matching studies had nonsignificant reuslts, but results of the present research could help provide clarity. Two possible explanations were mentioned in Chapter II: (a) matching models in general were conceptually flawed or required conceptual elaborations and (b) poor research designs and/or study executions were unable to detect significant relationships. Possibilities related to conceptualizations of matching will be addressed in this section. Design and implementation considerations will be discussed in the methodology section of this chapter. It is likely that the basic premise of matching for similarity is still plausible. A quote from Parloff et a1. (1978) in Chapter I identified matching as an effort to find numerous key areas along which client and counselor could be intentionally paired in order to attain the best possible counseling outcome. Sufficient evidence exists to support the idea that matching personalities with similar occupational environments does result in satisfaction as predicted by Holland's theory (Holland, 1973; Osipow, 1983). In addtiion, reviewers of the 63 personality matching research literature agree that enough affirmative findings existed to hold that client-counselor personality matching deserves further investigation (Berzins, 1977; Parloff et al., 1978). They were also unanimous in recommending how research efforts could improve; these ideas are presented in the following paragraphs. Reviewers of matching research labeled the primary obstacle as being a lack of specificity in the conceptual bases of past studies (Bergin 6c Suinn, 1975; Berzins, 1977; Parloff et al., 1978; Phillips 6: Bierman, 1981; Shertzer 6: Stone, 1980; Whiteley et al., 1975). Phillips and Bierman stated, "Matching patients and therapists has largely failed, perhaps because of methodological weaknesses and reliance on simplistic matching variables" (p. 408). The same authors suggested, in turn, that future research would improve with more specific, detailed models and variables for matching. Support for the value of specific definitions of major constructs might be seen in the effectiveness of research on the social influence model. Since the beginning of study on the social influence model, the constructs of attractivemess, expertness, and trustworthiness were tied to concrete behaviors and the trend has been toward even more specificity (Corrigan et al., 1980). It is possible that Holland's (1973) typologies are too general to be linked with actual behaviors. Descriptions of the six typologies are more exact than either the interpersonal transactions descriptors (Carson, 1969) or the A-B model descriptors (Cox, 1978; Whitheorn dc Betz, 1960). Nevertheless, the Holland typologies are interests, values, and attitudes. Studies on Holland's theory often use general behavioral outcomes such as duration in a career or choice of college major. Those outcomes may be too crude to be useful in matching research. Future applications of his typologies in matching, however, might improve if behavioral cues were defined for identifying and displaying the different 64 typologies. Berzins (1977) stated that effective client-counselor matching models would be developed only after research demonstrated how particular counselor actions influenced particular client behaviors. Using a model with personality constructs, then, might be premature. It may be that a strong, specific data base must first be developed through the social influence model (Strong, 1968) and other similar lines of research. Holland's Model This investigation looked at client-counselor personality matching for similarity. A vital component within that broader framework was Holland's (1973) theory of vocational choices and his personality typologies. Thus, findings in Holland's theoretical system have ramifications for any use of his constructs in matching research. These ramifications are discussed in this section. The criteria established for functionally defining Holland's constructs of congruence, consistency, and differentiation proved to be highly restrictive with prospective subjects; recall that only 3296 of those who volunteered to participate met the criteria on all three constructs. It was judged essential to continue including all three of the constructs, since Bruch (1978) and Holland (1973) regarded them as essential factors in typology definition and in typology- environment matching. The functional definitions used in this study, however, limited the sample size. An important factor directly related to more powerful statistics is the number of experimental units involved. Having more participants in the subject pool increases the opportunities for detecting true differences between groups (Borg 6r Gall, 1983). Inasmuch as this study was able to garner just one subject per cell of the Latin Square design, it is conceivable that a greater sample size would have raised the chance of identifying statistically significant relationships. 65 Still, the sample of 36 that was obtained was sufficient to identify a significant effect. Another implication of the smaller sample was its delimiting possible generalizations of the results. The sample pool was already restricted for generalizations, since the subjects were more highly educated and likely more active in Protestant churches than the population as a whole. The extreme selectivity of the criteria for subjects to participate probably limited generalizations even more. Thus, the significant results can only be applied to highly educated Protestants who attain VPI profiles deemed ideal by Holland (1978). The concern here was not that of insufficient volunteers; a total of 147 people completed the Vocational Preference Inventory. Instead, the problem was that the criteria for qualifying subjects to rate the audiotapes eliminated 68% of the volunteers. If all 147 persons had qualified to rate the tapes, it is possible that the Latin Square could have had three or four replications in each cell. From methodological and statistical standpoints, then, it is worthwhile to reexamine the roles that congruence, consistency, and differentiation play in a Holland theory-based study. Congruence, consistency, and differentiation were defined in Chapter II. That chapter also summarized research on each dimension. Since Holland's last major theoretical statement in 1973, the data trend indicates that consistency is unnecesaary (Erwin, I982; Nafziger et al., 1975; Spencer, 1982) and that differentiation is unnecessary (Lowe, 1981; Nafziger et al., 1974) in research on Holland's theory. Durign the same period of time, the data trend favors congruence as a useful construct (Holland et al., 1981; Osipow, 1983; Solomone 6r Slaney, 1978; Walsh, 1974). It would be possible, therefore, to explore the 66 ramifications of Holland's theory for client-counselor matching by excluding consistency and differentiation in selecting subjects. It is interesting to note that the distance variable in this project's analysis of variance and planned comparisons operated in the same manner as Holland's (1973) definition of congruence. Attaining significant results with the first planned comparison contrasts with the occupational studies questioning the utility of the construct. Perhaps congruence functions differently in personality matching studies than in vocational research. Another possibility is that congruence overlaps with some other unknown factor(s) such that its influence depends upon that factor or factors also being present. These possibilities are speculation, but might be considerations in planning future studies. Statistical significance was found with 36 subjects. Future studies, though, should seek more subjects and thereby increase their chances of obtaining significant results. Since more people would be available for participation, using congruence as the sole criterion for choosing subjects would be advantageous. Eliminating consistency and differentiation would simplify selection criteria. Instead of using the highest three scores on the Vocational Preference Inventory to define subjects' typologies, the highest score alone would suffice. The secondary and tertiary scores were used in this project to identify consistency and differentiation. Using congruence as the primary selection criterion in personality matching research would have limitationns. It was discussed in Chapter II that congruence was a process of personality matching, not an aspect of personality itself. Investigators could use congruence, since it appears to predict subject choices of environments that will be satisfying (Holland, 1973; Morrow, 1971; Walsh, 1974). Even so, congruence is not a personality factor and is not suitable as the sole criterion for personality matching. 67 Social Influence Model This research was meant to focus on topics other than the Social Influence Model. The relevance and strength of the empirical findings from that line of study, however, necessitated its consideration. The social influence model might have some explanations related to this study's focus on client-counselor matching for personality similarity. Two findings are discussed here. Both of these findings pertain to the attractiveness dimension and its influence by other factors. It was reported that the expertness dimension influenced subjects more than attractiveness (Corrigan et a1, 1980; Strong 6: Dixon, 1971). Expertness was not manipulated as a variable in the current project because interpretations of Bruch's (1978) and Holland's (1973) writings suggested that attractiveness was the only issue involved in matching for similarity through Holland's (I973) typologies. Perhaps, though, such an approach was overly simplistic. It may be that expertness interacts with attractiveness to the point that the two cannot be studied separately. In the brief counselor presentaions in this study, the issue of counselor expertise was never raised by the researcher or by subjects. Some subjects, however, might have counselor expertness as an unspoken expectation. Future studies on client-counselor matching through Holland's typologies ought to address the expertness issue in one of two ways. First, should the methodology used here be duplicated in another study, the investigator(s) should ask subjects how important counselor expertise is in their ratings. This could be accomplsiehd either orally or in written form (e.g., a Likert scale such as the vocational counseling interest scale at the bottom of the counselor choice rating scale). That approach would have the same assumption as the present study, that expertness exerted insignificant influence during the informed consent phase of client-counselor contact. Asking subjects about the importance of counselor 68 expertness also provides an explicit check on the possible influence of the dimension. The second manner of addressing expertness is to build it into the study. Such an approach would assume that counselor expertness is a major consideration for subjects from the earliest moments of counseling contact. The dimension might be incorporated in a number of ways. For instance, the taped counselor could be introduced as a Ph.D. who has extensive training and publications in career counseling. The Option of including expertness is preferable. Research on the social influence model identifies a consistent pattern of perceived expertness affecting attractiveness in early counseling phases (Corrigan et al., 1980). It seems, therefore, that since expertness is a major consideration for most of the counseling process, it is probably an important factor at the outset as well. A second finding from social influence research is what reviewers have labeled counselor "legitimate status," wherein the counselor has authority and influence simply be being in that professional role (Corrigan et al., 1980; Heppner 6r Heesacker, 1982). Clients' perceptions and ratings of counselors change over the course of counseling as they interact with the counselor (Bodden 6r Winer, 1978; Heppner 6r Heesacker, 1982). Little or no actual experience with the counselor at the beginning of the relationship, though, places a great deal of emphasis on the counselor's legitimate status in ratings of the counselor. Thus, subjects might rely heavily upon their pre—counseling impressions of counselors in general instead of solely upon the actual counselors' initial verbalizations. If subjects had similar stereotypic pre-counseling impressions, they would probably discriminate little between differing counselor presentations. Statistical analyses of ratings based on these stereotypic impressions would, in turn, likely 69 be nonsignificant, due to the restricted amount of between-group variability such ratings generate (Meyers 6c Grossen, I974). The impact of legitimate status might help explain the mixed results on the effects of pre-counseling information as mentioned in Chapter II (Berzins, 1977; Goldstein, 1980; Manthei, 1983). In the present study, though, subjects' ratings of the six audiotapes on the six point counselor choice rating scale ranged from two points to six. That is, the least discriminating rater used two of the six Likert scale points to rate the six tapes; the most discriminating rater used all of the six Likert scale points in rating the tapes. The mean number of points used by subjects was 4.22 (SD = 0.95). If legitimate status had a great deal of influence in subjects' ratings and they discriminated little among the six different counselor presentations, then they probably would have used a restricted range of points on the Likert scale. Apparently, legitimate status was not an issue in these ratings. The social influence model is a strong force in contemporary counseling research. This study directly incorporated some of its findings insofar as they clearly pertained to client-counselor matching. As the discussions in this section exemplify, however, the social influence model has many implications for the topic of this study. It is important that future matching studies based on the current one consider what is known about client-counselor matching studies for personality similarity, Holland's theory, and the social influence model. Otherwise, research results are likely to be nonsignificant or inadequate. Methodology A possible explanation for poor results from matching studies is weak research design and/or insufficient implementation. Deficiencies in design and analysis appeared in the one published study that was comparable in focus to this 70 project (Cox 6: Thoreson, 1977); Chapter II discussed these deficiencies. Efforts were made here to avoid problems of the Cox and Thoreson study. Efforts were also made to strengthen the study according to recommendations that matching research should use factorial designs (Berzins, 1977; Parloff et al., 1978). These efforts provided a strong test of the prOposed model of matching through Holland's (I973) typologies. The research design was valuable in this respect. The repeated measures analysis of variance checked main and interaction effects. The planned comparisons gave direct tests of the directional research hypothesis. Future matching studies should benefit from continuing to use more complex designs. A second source of concern in the implementation of the study as designed was the weakness of control possible for conditions under which subjects rated the audiotapes. Preferably, all of the subjects would have heard the tapes in the same physical setting, at more nearly the same time of the day, and more nearly within the same time span after completing the vocational preference inventory. Nevertheless, the repeated measures aspect of the study controlled for subject variations along these variables (Winer, 1962). A third methodological issue is the use of audiotapes for the counselor presentations. It was explained in Chapter III that audiotapes avoided various sources of potential confounding due especially to counselor physical appearance and age. The audiotapes also simplified the stimuli to which subjects reacted, as their only cues for evaluating the tapes were the counselor's words. Perhaps the audiotapes created some difficulties for this study, while they were averting some others. An issue in some studies was that subject attraction to counselors was based more on nonverbal aspects of the counselors (e.g., frequency of counselor smiling) than on counselor's verbal behaviors (Barak, Patkin, or Dell, 1982; Claiborn, 1979; Down or Pety, 1982). It was decided for this 71 study to consider client-counselor personality matching only in terms of verbal behavior patterns. Perhaps, however, the scope of the definition of personality should be expanded to include nonverbal behaviors. An expanded definition might benefit a study of this sort, where attempts are made to enhance counselor attractiveness through similarity. Videotaped or live counselor presentations would provide the nonverbal stimuli for subjects. A fourth potential difficulty with this study was its use of quasi-clients as subjects rather than people actually interested in receiving counseling. Borg and Gall (1983) and Goldman (1978) commented on how volunteer subjects in general would likely act differently than nonvolunteers. Goldstein (1980) stressed the same point for counseling research, since it would be expected that actual counselees experience more distress than non-clients and behave more like clients as a whole. Manthei (1983) as well supported the idea of having actual clients as research subjects whenever possible. The low Pearson correlation between subject interest in actual vocational counseling and counselor evaluations plus to the low mean score of subjects' interests in vocational counseling in the present study suggest that lack of subject distress or urgency or interest in the research topic might be an issue here as well as in other matching studies. It seems wise, therefore, to conduct matching studies for personality similarity on Holland‘s (1973) typologies with actual counseling clients, whenever feasible. Numerous ethical and practical considerations, however, make this difficult (Borg 6: Gail, 1983). Final Remarks It was stated earlier in this chapter that the basic premise of client- counselor personality matching for similarity still seemed plausible. Given the myriad of difficulties associated with matching research over the past three 72 decades, however, matching studies per se might be replaced by more vigorous lines of research. The social influence model, for example, could assimilate the basic thesis of matching--seeking combinations of variables of client-counselor similarity or dissimilarity that improve counseling outcome or processes (Parloff et al., 1978). Assimilation of the matching thesis by one or more other lines of research (e.g., the social influence model) might be a positive step in the evolution of counseling psychology. Cumulative findings in the personality matching area that were summarized in Chapter 11 suggested that the matching thesis had some validity. The Chapters 11 and V discussions noted, though, that its development and utility also had serious deficiencies. Unless researchers overcome these deficiencies, the matching thesis will be useless in facilitating attainment of the possible benefits mentioned in Chapter Inimproving counseling outcomes and processes and improving counselor education. This study found support for the idea that clients and counselors could be matched for personality similarity. One implication of the results is that matches for similarity would encourage greater client attraction to a counselor which, in turn, would increase chances of the client's returning to the counselor. Continued client-counselor contact would hopefully result in greater client gains. Using Holland's (1973) typologies would help this matching by providing a strong theoretical base and simple screening instruments. APPENDICES APPENDIX A VPI RESULTS PROFILE SHEET ARTISTIC COUNSELOR SCRIPT Perhaps it would be helpful if I told you how I visualize counseling. My sense is that counseling is meant to help you explore things that concern or interest you. We will focus more on vocational issues, since that is what I look at here in my job. While you're getting a clearer picture of the working world, you may tune into other parts of your life that may surface. We can go into those, too, if you wish. Basically, I see myself as a creative person who will work with you in identifying some possible career directions. I'll try as hard as I can to tune into issues that you bring up. I'll gladly share my own intuitions and experiences with you, if you want. I recognize, though, my responsibility for helping you make your own decisions. Your job will be to help me pick up on what you honestly feel about whatever we talk about. The more I can know you as a unique individual, the more help I can give in moving toward the goals and objectives we will establish. You will need to express your feelings, give information related to the issues we discuss, exercise your individuality in making decisions, and creatively respond to tasks needed to reach each of your goals. I believe that much of this will happen naturally, as we both respond spontaneously to where we are in the whole process. Frequently, it is hard for people to get a clear sense of how to look at a concern. We usually try to keep the big picture in mind while looking at parts of it. Sometimes we'll focus on only a detail of the whole picture, but gradually we will paint in more and more parts until the whole picture is completed. For example, say you wanted to become more flexible in your approach to working so 73 74 you could relate better to coworkers or a possible employer in an interview. We would look at communication skills that would be helpful in these situations and work through any feelings you might have about using the skills. Since we will be looking at vocational issues, I think it would be important to share my feelings about the world of work. I don't want to force my values, feelings, and ideas on you; you can choose what you feel is important in looking at working. I see the world of work as a place where people grow and develop in their abilities to express themselves. In ideal working environments, people can maintain their individuality while interacting with other people in positive ways. There is freedom to be yourself, if you choose to exercise it in flexible ways. One other thing is important at this point: I want you to know that counseling is voluntary. You may quit whenever you wish. The interviews are private; that is, I won't discuss you or your concerns with anyone without your prior approval. Our interviews usually last about 50 minutes. It often takes about five sessions to resolve most concerns; sometimes less, sometimes more. Now, if you have questions about counseling, I'll be glad to answer them. 75 CONVENTIONAL COUNSELOR SCRIPT Perhaps it would be helpful if I shared with you my basic beliefs about counseling. I think that counseling is meant to help you work on things that concern or interest you. We will focus more on vocational issues, since this is the area in which I have been hired. If, while looking at career issues, we find that other areas of your life need attention, we can go into those, too, if you wish. Ultimately, it is up to you to decide the matters we will discuss. Basically, I see my job to be one of helping you. I'll conscientiously work with you to assist in identifying some possible career direction for you. I think our chances are best for doing that when we take a practical approach. I'll try as hard as I can to hear your point of view and to understand the issues that you bring up. In order to be efficient at this, I think it is important that we proceed in an orderly fashion. So part of my job is to use my training and experience to keep us on task. This can lead to practical payoffs, if we're both persistent and exercise some self-control. Your job will be to help me understand what you honestly think about whatever we discuss. The more data you can give me about yourself, the better assistance I can provide in moving towards the goals and objectives we will establish. So you will need to provide me with your opinions, with information related to the issues we raise, and with your cooperation in making decisions and in performing tasks that we both see as necessary. I believe that this will happen, if we keep your long—range goals in mind. Frequently, it seems hard for people to get a handle on how to focus on a concern. We usually try to zero in on a specific target that you would like to 76 accomplish. Sometimes this target is only a small part of your overall concern, yet gradually we will work on more and more parts until we have rsolved the larger issue. Another way of saying that is that we'll just take one step at a time. For example, say you wanted to develop a strategy for getting information about companies you could work for. We could identify and talk about the resources you could use to gather data, then have you get and organize the information you would need. Since we'll be looking at vocational issues, I think it would be important to share a little bit about how I view the world of work. In this way, I hope to let you know of my values about working, values which you may or may not share. I don't want to force my ideas on you; you're free to choose what you think is important in looking at work. I see the work world as a place to earn money in order to exist, but it is more than that, too. It can be a place where you can show yourself to be competent and dependable. By being conscientious and efficient, you can feel in control of yourself and your situation, which is important for everyone to feel. The more in control of yourself you can feel, the more satisfied you will be with your occupation. One other thing I want you to know is that counseling is voluntary. You may quit whenever you wish. The interviews are private; I won't discuss you or your concerns with anyone without your prior approval. Our interviews usually last about 50 minutes. It often takes about five sessions to resolve most concerns; sometimes less, sometimes more. Now, if you have questions about counseling, I'll be glad to answer them. 77 ENTERPRISING COUNSELOR SCRIPT Perhaps it would be helpful if I shared with you my basic beliefs about counseling. I think that counseling is meant to help you become successful at doing things that interest you. We will focus more on your building a successful career since vocational counseling is what my job is about. If, while looking at career issues, we find that other areas of your life need attention, we can go into those, too, if you wish. Ultimately, it is up to you to decide the matters we will discuss. Basically, I see my job as helping you get what you want. I am prepared to focus my energies on helping you identify some possible career direction. I think our chances are best for doing that when we talk openly and listen carefully to one another. I'll try as hard as I can to listen to and understand the issues that you bring up. I will gladly share my own insights and experiences about what it takes to get where you want to go. I also recognize, though, my responsiblity for helping you make your own decisons. Really, this whole process concerns your finding out what you have to offer and then finding ways to let people know it. Your job will be to help me understand what you honestly think and feel about whatever we discuss. The more I can know about what your strengths are, the better I can assist you in moving towards the goals and objectives we will establish. So you will need to provide me with your opinions, with information related to the issues we raise; and you'll have to make decisions and perform tasks that we both see as necessary to reach your goals. I believe that much of this will happen naturally and will be a pleasurable experience for both of us. 78 Frequently, it seems hard for poeple to focus in on a strategy. We usually try to zero in on a specific target that you would like to accomplish. Sometimes this target is only a small part of your overall strategy, yet gradually we will work on more and more parts until we have put the whole program together. Another way of saying that is that we'll just take one step at a time. For example, say you wanted to develop a more effective way to sell yourself in an interview with a prospective employer. We could identify and talk about the skills that would be helpful in this situation, set up ways to practice them, and also work on taking care of any concerns you might have about applying the skills. Since we'll be looking at vocational issues, I think it would be important to share a little bit about how I view the world of work. In this way, I hope to let you know of my values about working, values which you may or may not share. I don't want to force my ideas on you; you're free to choose what you think and feel is important in looking at work. I see the work world as a place to earn rewards. Money is certainly one of those rewards, and we'll look at ways for you to enhance your earnings. There are other rewards aside from money, though. For example, there is the pride you can have in yourself as you take on increasing responsibility where you work nd are increasingly identified as a leader. There is the respect that you will gain from your peers and your subordinates as you demonstrate your ability to achieve goals to lead others toward achieving theirs. One other thing is important at this point. I want you to know that counseling is voluntary. You may quit whenever you wish. The interviews are private; that is, I won't discuss you or your concerns with anyone without your prior approval. Our interviews usually last about 50 minutes. It often takes 79 about five sessions to resolve most concerns; sometimes less, sometimes more. Now, if you have questions about counseling, I'll be glad to answer them. _ 80 INVESTIGATIVE COUNSELOR SCRIPT Perhaps it would be helpful if I shared with you my basic beliefs about counseling. I think counseling is meant to help you investigate things that concern or interst you. We will focus more on vocational issues, since that is the area in which I have been hired. If, while analyzing career issues, we find that other areas of your life stimulate inquiry, we can look into those, too. Ultimately, it is up to you to decide the matters we will discuss. Basically, I see my role as helping you examine your situation. I am prepared to assist in identifying some possible career directions for you. I think our chances are best for doing that when we use a rational approach. I'll listen to and attempt to understand the issues that you bring up. I will gladly share my own insights and experiences with you. I also recognize, though, my responsibility for helping you make your own decisions. Really, this whole process involves both of us using a scientific approach and communicating with each other as precisely as possible. Your job will be to help me understand what you honestly think about whatever we discuss. The more I can know how you think, the better assistance I can provide in moving towards the goals and objectives we will establish together. So you will need to provide me with your opinions, with information related to the issues we raise, and with your cooperation in making decisions and in performing tasks that we both see as necessary to reach our goals. I believe that much of this will happen naturally, as we remain curious and systematically work together. 81 Frequently, it seems hard for people to get a handle on how to focus on a concern. We usually try to isolate a specific goal that you would like to accomplish. Sometimes this target is only a small part of your overall concern, yet gradually we will work on more and more components until we have resolved the larger issue. Another way of saying that is that we'll proceed one step at a time. For example, say you wanted to develop research skills so you could analyze possible employers in this area. We could identify and talk about the skill that would be helpful in this situation, set up ways to practice them, and also work on taking care of any concerns you might have about applying the skills. Since we'll be looking at vocational issues, I think it would be important to share a little bit about how I view the world of work. In this way, I hope to let you know of my values about working, values which you may or may not share. I don't want to force my ideas on you; you're free to choose what you think and feel is important in looking at working. I see the work world as more than a place to earn money in order to exist. It can also be someplace that provides an opportunity to learn more about life. Becuase I believe that people basically need and want to know as much about their worlds as they can, I encourage individuals with whom I'm working in counseling to remain curious and take note of the intellectual stimulation in their work environment. I think that a commitment to continued learning helps one feel stimulated at work and more satisfied. One other thing is important at this point. I want you to know that counseling is voluntary. You may quit whenever you wish. The interviews are private; that is, I won't discuss you or your concerns with anyone else without your prior approval. Our interviews usually last about 50 minutes. If often takes about five sessions to resolve most concerns; sometimes less, sometimes more. Now, if you have questions about counseling, I'll be glad to answer them. 82 REALISTIC COUNSELING SCRIPT I want to tell you about how I go about counseling. I think counseling is meant to help you work on things that are important to you. We will focus more on vocational issues, since that is what my job is. If while we're working on career issues we find that other parts of your life need work, we can go into those, too, if you want. It is up to you to decide what we will talk about. Basically, I see myself as working for you. I'll work with you in identifying some possible career directions. I think our chances are best for doing that when we get hands on experience with issues you bring up. I'll talk about what I think about your situation and direct experiences I've had, if you want. But I won't make any decisions for you. That is your job. Really, this whole thing depends on you figuring out what you want by getting direct experience with the working world. Your job will be to let me know what you honestly think about whatever we discuss. The more I can know about you, the better job I can do in moving towards what you want to accomplish. So you will need to provide me with your opinions and with information. You will have to make decisions and do the things that we are necessary to reach your goals. A lot of times it is tough for people to get their hands on what they want to do. We usually try to tackle one piece at a time. This piece may be only one player in the game. Yet gradually we will work on more and more pieces until the field is clear. For example, say you wanted to get a clearer idea of machines or tools that you would like to work with and that you would be good at using. We could do some aptitude testing to see what you would be good at, then 83 arrange for. you to go to work sites to get a hands-on feel of what it would be like to work with those aptitudes. Then we could tackle another part and eventually take care of it, too. Since we will be looking at vocational issues, I think it is important to tell you a little bit about how I view theworld of work. I don't want to force my ideas on you; you are free to choose what you think is important in looking at working. I see work as a place to earn money in order to get the things you want. It can also be a place where you can get other tangible rewards. If you work it right, you can get paid for doing things that you would have done anyway. For example, say you like to be outside and hike a lot. There are jobs that have outdoor activities built in. Or say you like to work with your hands. That can surely be built in, too. Maybe animals—same thing. The list can go on and on. One other thing is important at this point. I want you to know that counseling is voluntary. You may quit whenever you wish. The interviews are private; that is, I won't discuss you or your concerns with anyone else without your prior approval. Our interviews usually last about 50 minutes. If often takes about five sessions to resolve most concerns; sometimes less, sometimes more. Now, if you have questions about counseling, I'll be glad to answer them. 84 SOCIAL COUNSELOR SCRIPT Perhaps it would be helpful if I shared with you my basic beleifs about counseling. I think that counseling is meant to help you work on things that concern or interest you. We will focus more on vocational issues, since that is the area in which I have been hired. If while looking at career issues we find that other areas of your life need attention, we can go into those, too, if you wish. Ultimately, it is up to you to decide the matters we will discuss. Basically, I see myself as a helper. I will cooperate with you to assist in identifying some possible career direction for you. I think our chances are best for doing that when we develop a friendly, sharing relationship. I'll try as hard as I can to listen to and understand the issues that you bring up. I will gladly share my own insights and experiences with you, if you want them. I also recognize, though, my responsibility for helping you make your own decisions. Really, this whole process concerns your growth and development, and I enjoy seeing that happen. Your job will be to help me understand what you honestly think and feel about whatever we discuss. The more I can know you as a unique person, the better assistance I can provide in moving towards the goals and objectives we will establish together. You will need to provide me with your opinions, with information related to the issues we raise, and with your cooperation in making decisions and in performing tasks that we both see as necessary to reach our goals. I believe that much of this will happen naturally as we get to know and trust each other more. 85 Frequently, it seems hard for peOple to get a handle on how to focus on a concern. We usually try to isolate a specific goal that you would like to accomplish. Sometimes this target is only a small part of your overall concern, yet gradually we will work on more and more components until we have resolved the larger issue. For example, say you wanted to develop more communication skills so that you could relate better to coworkers or to a possible employer in an interview situation. We could identify and talk about the communication skills that would be helpful in these situations, set up ways to practice them, and also work on taking care of any concerns you might have about applying the skills. Since we will be looking at vocational issues, I think it is important to tell you a little bit about how I veiw the world of work. I have my own values about working, values which you may or may not share. I don't want to force my ideas on you; you are free to choose what you think is important in looking at working. I see the work world as more than a place to earn money in order to exist. It can also be someplace in which peOple grow and develop in their personal lives. Because I believe that people basically need and want to positively interact with other peOple, I usualy encourage individuals with whom I'm working in counseling to learn ways to be socially skilled, comparative, understanding of others' thoughts and feelings, flexible, and knowledgeable of ways to help people. I think that these capabilities create a happier work environment. And the more harmony that exists in that environment, the more satisfied many peOpIe are with their occupations. One other thing is important at this point. I want you to know that counseling is voluntary. You may quit whenever you wish. The interviews are private; that is, I won't discuss you or your concerns with anyone else without your prior approval. Our interviews usually last about 50 minutes. If often takes 86 about five sessions to resolve most concerns; sometimes less, sometimes more. Now, if you have questions about counseling, I'll be glad to answer them. APPENDIX B SPU LETTER 87 do: .3836»... 9:328 k. .22 ..§:8u0 82a 35650 .22 2... .355 08:8 E 62.... 9:690:22 ozEaon 33oz «ED .050 .672 cozaoo> vfiou E) Standard Scores 8.5 20555.. 2.: so“. war—OK; uq<2 cm mm On. 3. On mm om we 2. ms om ou< _ q 11... ~ c an n n ... _ o 1 711.. a 1 n a o o a n . o o I O? a a e . _ h s a N . o . _ Q h n p N N n' o u a v 2 . a n a 1 m . a II n 2 ~ jjml : a n v n . cm 2 o a m. N— 0 0 v s v Q n W a a c 2 2 : a i . . . . . . an. 9- Op 0 0 h . 0 fl 2 2 h 2 a a 2. 2 m u : a : a a c 2 : 2 a 2 2 2 a 2 o 2 2 : 2 .2 : o. a. n. a. 2 2 no on 2 2 2 a 2 2 2 2 12 1| 2 2 2 2 2. z 2 2 VN v— : 2 2. 1 an 2 I z 2 on on 2 3 am a: co .2 Eu 230 8m 3 .8: u. On a m b O m v n N u 2.32 9:56: a £2 3 >mohzw>z_ wozmmmumcn— 4 APPENDIX C CHURCH LETTER 88 Dear Colleague: .As you may know, I have been working on my dissertation since my return to SPU in January. 1980. Besides the demands of regular academic duties, with which you are all familiar. various unplanned circumstances have delayed the completion of the doctoral research. The time now seems to be right to. press on with the project and I am seeking your assistance. I will explain the dissertation proposal briefly. minus a small amount of information that could bias possible outcomes. I am interested in learning how people react to different types of vocational counselors. These people would listen to six 4 minute audiotape recordings of six different coun- selor self-introductions. Each person would then rate each tape according to how much the person might prefer the coun- selor in an actual occupational counseling situation. In order to provide a strong test of the concepts underlying this study, it is necessary for me to establish that in fact each participant fits into the vocational theory being investi- gated. This explains the Vocational Preference Inventory enclosed here. The VPI is scored to identify how each test taker fits into the theoretical basis for the study. Those people who can be alligned with the theory will be asked to rate the 6 audiotapes. Individuals fitting into the theory less clearly will not be asked to score the tapes. If you are willing to volunteer. I ask that you take the follow- ing steps: 1. Complete the VPI as soon as possible; it requires 15-20 minutes. 2. Read and sign the informed consent form. 3. Return both documents to me on Wednesday, September 9 or Thursday, September 10. 4. Be prepared to use 40 minutes of your Spare time on Friday. September 11 to rate the audiotapes. If you are one of the chosen few for the project. I will contact you at Faculty Retreat as to time and place for the 89 tape reviews. VPI's for persons not qualify- ing will be destroyed immediately, unless you request the return of your own form. Should you be unwilling to take part. please return the blank .VPI and informed consent forms so I can use them in future studies. If you wish to learn about details of the research. I will be glad to meet with you after the retreat to discuss it further. I realize that the timing of this undertaking is an imposition on the precious little free time we have at the retreat. That occasion, however. is the only block of time available before we are immersed in our tasks for the upcoming quarter. I ask your help too because the community college group I tested this summer failed to qualify for hearing the tapes; the im- plication was that I required a more vocationally mature group of people and you fill the bill. ' I might mention that my request to you has been cleared by Faculty Council. I am also willing to clarify aspects of the study, if you desire. You can reach me at my office (X2210) or at home (285-3506). Thank you. Respectfully. Don MacDonald Counseling Education APPENDIX D COUNSELOR SCRIPTS 90 As you may know, I have been working for quite a while now to complete a dissertation for the Ph.D. degree at Michigan State University. I am close to having that accomplished except for the major step of having enough qualified people volunteer to help me gather some information on the topic. That is the purpose of this note to you; I am seeking your assistance. First, let me quickly explain the project to you. The study idea canes from practices in helping peOple identify areas in which they might like to‘have a job. An important part of the process is getting a job seeker with a vocational counselor and the two of them working together well. What I am looking at in this working relationship is how much a person likes different types of career counselor's on a first impression basis. Even if you are not now interested in occupational counseling, your opinions are valuable to me and can give useful ideas on these impressions. If you agree to volunteer, I will be asking you to take part in one and perhaps two tasks. As a volunteer, you will complete a brief (lo-15 minutes) vocational survey, where you will simply mark any occupations that seem interesting to you. I will score the survey, which will tell me those people who "qualify" for the next step in the process. By “qualify" I mean that it is necessary that your survey scores work with the theory that I am testing. If you don't qualify, it says nothing is wrong with you or your chosen occupation. It simply says that the survey or the theory behind it is not sensitive enough to clearly identify your job interests. Such clear identification is essential, however, for me to do the study, so I would only ask those with certain types of results to do the next part. The next phase consists of simply listening to six different audiotapes. Each tape runs about 4 minutes, so your total time commitment here would be approx- imately 25 minutes. After hearing all six tapes, you would just write down how much you liked each presentation. ' My hope is that all of these matters could be taken care of at church on two different Sundays. Where it did not work out at church, because of logistics or your personal preference, I would be asking to do them in one evening at your home. I wish I could pranise sane great outccme that would immediately benefit you. I can provide a free analysis of the vocational survey results plus discuss questions you may have about career development in general. Primarily, though, you will be doing me a big favor by participating. I have received help .from faculty and staff at Seattle Pacific University, where I work. But your contributions are greatly needed as well. Should you decide to volunteer, please inform me as soon as possible. Telling me before or after church would be'fine. This will allow me to schedule the two phases to the project within the next 4 weeks. If I don't hear from you withing two weeks, I will telephone to confirm your decision. You are, of course, free to decline participation at any'time. Information you share in the course of the study will also be held in confidence. Sincerely, APPENDIX E DATA COLLECTION RELEASE FORM 91 SEATTLE PACIFIC UNIVERSITY Department of Counseling Education DATA COLLECTION RELEASE FORM* I consent to have information on my behavior collected, recorded, and stored for the purposes of research in counseling psychology, as long as the following conditions are observed: 1. 2. Participant's Signature: Date: My name does not appear on any of the data collection forms: my identity will be concealed through the use of a code system. The data storage will insure confidentiality: my identity will be known only by the principle researcher(s). Upon request, I can have data collected on myself destroyed upon completion of the study. Any written or oral reports on the study, including publica- tions, will conceal my identity. I have been informed that I am a participant in a research 3 tUdy e 8 I have been informed that my participation in this study is voluntary and that I can withdraw at any time. Any potentially harmful physical, mental, or emotional effects have been indicated to me. I understand that the project is meant to contribute to the advancement of scientific knowledge in the field of counsel- ing psychology. Study Title: Researcher(s): *Designed to satisfy ethical guidelines for research found in: American Psychological Association, Ethical Principles In The Conduct Of Research With Human Participants. Washington, D.C: 1973. BIBLIOGRAPHY BIBLIOGRAPHY Alston, H. L., Wakefield, J. A., Doughtie, E. B., 6: Bobele, M. (1976). Correspondence of constructs in Holland's theory for male and female college students. Journal of Vocational Behavior, 8_, 85-88. Bandura, A. (1978). The self system in reciprocal determinism. American Psychologist, 33, 344-358. Barak, A., Patkin, 3., 6: Dell, D. M. (1982). Effects of certain counselor behaviors on perceived expertness and attractiveness. Journal of Counseling Psychology, 22, 261-267. Barak, A., 6: Rabbi. (1982). Predicting persistence, stability, and achievement in college by major choice consistency: A test of Holland's consistency hypothesis. Journal of Vocational Behavior, 29, 235-243. Bent, J., Putnam, D. G., Kiesler, D. J., 6: Nowicki, 5., Jr. (1976). Correlates of successful and unsuccessful psychotherapy. Journal of Consulting and Clinical Psychology, 43, I49. Bergin, A. E., 6: Lambert, M. J. (1978). The evaluation of therapeutic outcomes. In S. L. Garfield 6: A. E. Bergi (Eds), Handbook of psychotherapy and be_havior changes: An empirical analysis (2nd edj. New York: John Wiley 6: Sons. Bergin, A. E., at Suinn, R. M. (1975). Individual psychotherapy and behavior therapy. Annual Reveiw of Psychology, 26, 509-556. Bodden, J. L. at Winer, J. L. (1978). A model of client impression formation. Counselor Education and Supervision, _18, 21-28. Borg, W. R., at Gall, M. D. (1983). Educational research: An introduction (4th ed.). New York: Longman. Brammer, L.M., 6c Shostrom, E. L. (1982). Therapeutic psycholgy (4th ed.). Englewood Cliffs, NJ: Prentice-Hall. Bruch, M. A. (1978). Holland's typology applied to client-counselor interaction: Implications for counseling men. Counseling Psychologist, 7(4), 26-32. Campbell, D.. P. (1977). Strong-Campbell interest inventory manual. Palo Alto, CA: Stanford University. Campbell, D. T., 6: Stanley, J. C. (1966). Ex rimental and quasi-experimental designs for research. Chicago: Rand McNalIy. 92 93 Carkhuff, R. R. (1972). The development of systematic human resource development models. Counseling Psychologist, 3(3), 4-11. Carson, R. C. (1969). Interaction concepts of personali_y. Chicago: Aldine. Carter, J. A. (1978). Impressions of counselors as a function of counselor physical attractiveness. Journal of Counseling chhology, 33, 28-34. Cash, T. F., Begley, P. J., McCown, D. A., a Weise, B. C. (1975). When counselors are heard but not seen: Initial impact of physical attractiveness. Journal of Counseling Psychology, _22, 373-279. Cash, T. F., or Salzbach, R. F. (1978). The beauty of counseling: Effects of counselor physical attractiveness and self-disclosures on perceptions of counselor behavior. Journal of Counseling Psychology, 35, 283-291. Cheney, T. (1975). Attitude similarity, topic importance, and psychotherapeutic attraction. Journal of Counseling Psychology, 2_2, 2-5. Claiborn, C. D. (1979). Counselor verbal intervention, nonverbal behavior, and social power. Journal of Counselirgficholggy, 2_6, 378-383. Corrigan, J. D., Dell, D. M., Lewis, K. D., At Schmidt, L. D. (1980). Counseling as a social influence process: A review. Journal of Counseling Psychology, ZZ, 395“““1. Corsini, R. J. (1977). Introduction. In R. J. Corsini (Ed.). Current personality theories. Itasca, IL: F. E. Peacock. Cox, J. G., or Thoreson, R. W. (1977). Client-counselor matching: A test of the Holland model. Journal of Counseling Psychology, 2_4, 156-161. Daher, D. M., 6: Banikiotes, P. G. (1976). Interpersonal attraction and rewarding aspects of disclosure content and level. Journal of Personality and Social Psychology, 33, 492-496. Dietzel, C. S., & Abeles, N. (1975). Client-therapist complementarity and the therapeutic outcome. Journal of Counseling Psychology, Q, 264-272. Dowd, E. T., or Pety, J. (1982). Effect of counselor predicate matching on perceived social influence and client satisfaction. Journal of Counseling Psychology, 2_9, 206-209. Erwin, T. D. (1982). The predictive validity of Holland's construct of consistency. Journal of Vocational Behavior, Q, 180-192. Frank, J. D., Hoehn-Saric, R., Imber, S. D., Liberman, B. L., or Stone, A. R. (1978). Effective ingredients of successful psychotherapy. New York: Brunner/Mazel. Frantz, T., 6: Walsh, E. (1972). Exploration of Holland's theory of vocational choice in a graduate school environment. Journal of Vocational Behavior, 3, 223-232. 94 Glass, G. V., 6: Stanley, J. C. (1970). Statistical methods in education and psychology. Englewood Cliffs, NJ: PrenticeHall. Goldman, L. (1978). Introduction and point of view. In L. Goldman (Ed.). Research methods for counselors. New York: John Wiley 6: Sons. Goldstein, A. P. (1971). Psychotherapeutic attraction. New York: Pergamon. Goldstein, A. P. (1980). Relationship-enhancement methods. In F. H. Kanfer or A. P. Goldstein (Eds.). Helping people change (2nd ed.). New York: Pergamon. Goldstein, A. P., Heller, K., at Sechrest, L. B. (1966). chhotherapy and the psychology of behavior change. New York: John Wiley 6: Sons. Gomes-Schwartz, B., Hadley, S. W., 6: Strupp, H. H. (1978). Individual psychotherapy and behavior therapy. Annual Review of Psychology, 23, 435-471. Gottfredson, L. S. (1979). Aspiration-job match: Age trends in a large representative sample of young white men. Journal of Counseling Psychology, 2_6, 319-328. Hackman, H. W., dc Claiborn, C. D. (1982). An attributional approach to counselor attractiveness. Journal of Counseling Psychology, 23, 224-231. Hare-Mustin, R. T., Marecek, J., Kaplan, A. G., 6: Liss-Levinson, N. (1979). Rights of clients, responsibilities of therapists. American Psychologist, fl, 3-16. Heck, E., 6: Davis, C. (1977). The differential treatment model and drug counseling: An epilogue. In C. Davis 6: M. Schmidt (Eds.). Differential treatment of drug and alcohol abusers. Palm Springs, CA: ETC. Helms, J. E., 6: Simons, J. A. (1977). The influence of counselor's age and sex on college students' counselor preferences. Journal of College Student Personnel, _12, 115-123. Heppner, P. P., 6: Dixon, D. N. (1981). A review of the interpersonal influence process in counseling. Personnel and Guidance Journal, 33, 542-550. Heppner, P. P., 6: Heesacker, M. (1982). Interpersonal influence process in real- life counseling: Investigating client perceptions, counselor experience level, and counselor power over time. Journal of Counseling Psychology, 23, 215-223. Hersen, M., or Barlow, D. H. (1976). Single case gxperimental designs: Strategies for studying behavior change. New York: Pergamon. Higgins, E. B., 6: Warner, R. W., Jr. (1975). Counseling blacks. Personnel and Guidance Journal, 5_3, 382-386. 95 Hoffman-Graff, M. A. (1977). Interviewer use of positive and negative self- disclosure and interviewer-subject sex pairing. Journal of Counseling Psychology, 25, 184-190. Holland, J. L. (1972). Professional manual--the self-directed search. Palo Alto, CA: Consulting Psychologists. Holland, J. L. (1973). Making vocational choices: A theory of careers. Englewood Cliffs, NJ: Prentice-Hall. Holland, J. L. (1976). A new synthesis for an old method and a new analysis of some old phenomena. Counseling Psychologist, 3(3), 12-15. Holland, J. L. (1978). Manual for the vocational preference inventcm. Palo Alto, CA: Consulting Psychologists. Holland, J. L., dc Gottfredson, G. D. (1976). Using a typology of persons and environments to explain careers: Some extensions and clarifications. Counseling Psychologist, 3(3), 20-29. Holland, J. L., Magoon, T. M., or Spokane, A. R. (1981). Counseling psychology: Career interventions, research, and theory. Annual Review of Psychology, 32, 279-305. Hosford, R. E. (1969). Behavioral counseling: A contemporary overview. Counseling Psychologist, 3(4), 1-32. Howell, D. C. (1982). Statistical methods for psychology. Boston: Duxbury. Hughes, H. M. (1972). Vocational choice, level, and consistency: An investigation of Holland's theory on an employed sample. Journal of Vocational Behavior, 2, 377-388. Huston, T. L., 6: Levinger, G. (1978). Interpersonal attraction and relationships. Annual Review of Psychology,. 23, 115-156. Johnson, D. H. (1978). Students' sex preferences and sex role expectations for counselors. Journal of Counseling Psychology, Q, 557-562. Kaul, T. J., 6: Schmidt, L. D. (1971). Dimensions of interviewer trustworthiness. Journal of Counseling Psychology, 3, 542-548. Kirk, R. E. (1972). Classification of ANOVA designs. In R. E. Kirk (Ed). §tatistic_el issues: A reader for the behavioral sciences. Monterey, CA: BrooksICole. Kivlighan, D. M., Hageseth, J. A., Tipton, R. M., 6: McGovern, T. V. (1981). Effects of matching treatment approaches and personality types in group vocational counseling. Journal of Counselim Psychology, g, 315-320. Krauskopf, C. J., Baumgardner, A., at Mandracchia, S. (1981). Return rate following intake revisited. Journal of Counseling Psychology, g, 519-521. 96 Krumboltz, J. D., Mitchell, A. M., 6: Jones, G. B. (1976). A social learning theory of career selection. Counseling Psychologist, §(l), 71-81. Leonard, R. L., Jr., Walsh, W. B., a: Osipow, S. H. (1973). Self-esteem, self- consistency, and second vocational choice. Journal of Counseling Psychology, 20, 91-93. Lewis, K. N., 6: Walsh, W. B. (1980). Effects of value-communication style and similarity of values on counselor evaluation. Journal of Counseling Psychology, g, 305-314. , Lowe, B. (1981). The relationship between vocational interest differentiation and career decisions. Journal of Vocational Behavior, 2, 346-349. Manthei, R. J. (1983). Client choice of therapist or therapy. Personnel and Guidance Journal, 31, 334-340. Menninger, K. (1958). Theory of psychoanalytic technigue. New York: Harper. Meyers, L. S., 6: Grossen, N. E. (1974). Behavioral research: Theory, procedure, and design. San Francisco: W. H. Freeman. Morrow, J. M. (1971). A test of Holland's theory of vocational choices. Journal of Counseling Psychology, 3, 422-425. Nafziger, D.*H., Holland, J. L., 6: Gottfredson, G. D. (1975). Student-college congruency as a predictor of satisfaction. Journal of Counseling Psychology, 22, 132-139. Nafziger, D. H., Holland, J. L., Helms, S. T., 6: McPartland, J. M. (1974). Applying an occupational classification to the work histories of young men and women. Journal of Vocational Behavior, _5_, 331-345. Nilsson, D. E., Strassberg, D. S., & Bannon, J. (1979). Perceptions of counselor self-disclosure: An analogue study. Journal of Counseling Psychology, 2_6, 339-40“. O'Neil, J. M., Magoon, T. M., a Tracey, T. J. (1978). Status of Holland's investigative personality types and their consistency levels seven years later. Journal of Counseling Psycholgy, g, 530-535. Osipow, S. H. (1973). Theories of career development (2nd ed.). Englewood Cliffs, NJ: Prentice-Hall. Osipow, S. H. (1983). Theories of career development (3nd ed.). Englewood Cliffs, NJ: Prentice-Hall. Parloff, M. B., Waskow, I. E., & Wolfe, B. E. (1978). Research on therapist variables in relation to process and outcome. In S. L. Garfield 6: A. E Bergin (Eds.). Handbook of psychotherapy and behavior change: An empirical analysis (2nd ed.). New York: John Wiley 6t Sons. 97 Phillips, J. S., 6: Bierman, K. L. (1981). Clinical psychology: Individual methods. Annual Review of Psychology, 32, 405-438. Razin, A. M. (1977). The A-B variable: Still promising after twenty years? In A. S. Gurman 61 A. M. Razin (Eds.). Effective psychotherapy: A handbook of research. New York: Pergamon. Rogers, C. R. (1951). Client-centered therapy. Boston: Houghton Mifflin. Salomone, P. R., 8: Slaney, R. B. (1978). The applicability of Holland's theory to nonprofessional workers. Journal of Vocational Behavior, l_3, 63-74. Scheid, A. B. (1976). Clients' perceptions of the counselor: The influence of counselor introduction and behavior. Journal of Counseling Psychology, 2_3, 503-508. Schmidt, L. D., 6: Strong, S. R. (1971). Attractiveness and influence in counseling. Journal of Counseling Psychology, _13, 348-351. Scott, W. A., dc Wertheimer, M. (1962). Introduction to psychological research. New York: Wiley. Shertzer, B., 6: Linden, J. D. (1979). Fundamentals of individual appraisal: Assessment techniques for counselors. Boston: Houghton Mifflin. Shertzer, B., 6: Stone, 5. C. (1980). Fundamentals of counseling (3rd ed.). Boston: Houghton Mifflin. Spencer, M. (1982). A test of Holland's theory with a sample of adults employed in professional roles in an academic setting. Unpublished masters thesis, Seattle Pacific University. Spokane, A. R., 6: Derby, D. P. (1979). Congruence, personality pattern, and satisfaction in college women. Journal of Vocational Behavior, _13, 36-42. Stewart, N. R., Winborn, B. B., Johnson, R. G., Burks, H. M., 6: Engelkes, J. R. (1978). Systematic counseling. Englewood Cliffs, NJ: Prentice-Hall. Strong, S. R. (1968). Counseling: An interpersonal influence process. Journal of CounselimL Psychology, Q, 81-87. Strong, 5. R., 6: Dixon, D. N. (1971). Expertness, attractiveness, and influence in counseling. Journal of Counseling Psychology, 13, 562-570. Strong, S. R., Taylor, R. G., Bratton, J. C., at Loper, R. G. (1971). Nonverbal behavior and perceived counselor charactersitics. Journal of Counseling Strupp, H. H., Fox, R. E., Jr Lessier, K. (1969). Patients view their psychotherm. Baltimore: Johns Hopkins. Strupp, H. H., Hadley, S. W., & Comes-Schwartz, B. (1977). PSYChotherapy for better or for worse. New York: Jason Aronson. 98 Thoreson, C. E., at Coates, T. J. (1978). What does it mean to be a behavior therapist? Counseling Psychologist, 2(3), 3-21. Tolbert, E. L. (1980). Counseling for career development (2nd ed.). Boston: Houghton Mifflin. Venzor, E., Gillis, J. 5., 6c Beal, D G. (1976). Preference for counselor repsonse style. Journal of Counseling Psychology, 23, 538-542. Wakefield, J. A., Jr., Yom, B. L., Doughtie, E. B., Chang, W. C., 6: Alston, H. L. (1975). The geometric relationship between Holland's personality typology and the vocational preference inventory for blacks. Journal of Counseling Psychology, 2_2, 58-60. Walsh, W. B. (1974). Consistent occupational preferences and personality. Journal of Vocational Behavior, 1, 145-153. Walsh, W. B., dc Lewis, R. O. (1972). Consistent, inconsistent, and undecided career preferences and personality. Journal of Vocational Behavior, 2, 309-316. Walsh, W. B., 6: Russell, J. H. (1969). College major choice and personal adjustment. Personnel and Guidance Journal, 32, 685-689. Weinrach, S. g. (1979). The Holland model: What it is and how it works. In S. G. Weinrach (Ed.). Career counseling: Theoretical and practical perspectives. New York: McGraw-Hill. Whitehorn, J. C., d: Betz, B. J. (1960). Further studies of the doctor as a crucial variable in the outcome of treatment with schizophrenic patients. American Journal of Psychiatry, 27:, 215-223. Whiteley, J. M., Burkhart, M. Q., Harway-Herman, M., at Whiteley, R. M. (1975). Counseling and student development. Annual Review of Psychology, 26, 337-366. Winer, B. J. (1962). Statistical principles in experimental design. New York: McGraw-Hill. Wolberg, L. R. (1977). The technique of psychotherapy: Volume 2 (3rd ed.). New York: Grune 6r Stratton. Wright, J. A., a: Hutton, B. B. (1977). Influence of client socioeconomic status on selected behaviors, attitudes, and decisions of counselors. Journal of Counseling Psychology, 2_4, 527-530. MICHIGAN STATE UNIV. LIBRARIES 1|1111111111111”1111IllIll11111111111111111111 31293000650329