• unlimited access with print and download
    $ 37 00
  • read full document, no print or download, expires after 72 hours
    $ 4 99
More info
Unlimited access including download and printing, plus availability for reading and annotating in your in your Udini library.
  • Access to this article in your Udini library for 72 hours from purchase.
  • The article will not be available for download or print.
  • Upgrade to the full version of this document at a reduced price.
  • Your trial access payment is credited when purchasing the full version.
Buy
Continue searching

Counselor cognitive complexity: Instrument development and validation

Dissertation
Author: Laura Everhart Welfare
Abstract:
Because issues in counseling are often complex, it is difficult for counselors to accurately understand their clients' needs. A counselor's understanding is limited by his or her ability to recognize relevant client variables and comprehend their interactional impact on the client's overall needs (Blocher, 1983). This process is complicated, and requires advanced cognitive complexity. Counselor cognitive complexity has been linked with multiple aspects of counselor effectiveness (e.g., Borders, 1989; Fong, Borders, Ethington, & Pitts, 1997; Holloway & Wolleat, 1980). In fact, researchers have found support for the assumption that cognitive complexity increases during supervised counseling practice (e.g., Duys & Hedstrom, 2000). To date, however, these results primarily have been based in general measures of a counselor's cognitive complexity, despite evidence that level of complexity is domain specific (Crockett, 1965). Crockett (1965) reported that an individual's level of cognitive complexity can vary from topic to topic. Because of the domain-specific nature of cognitive complexity, a counseling-specific measure of cognitions is needed before researchers can fully understand counselor cognitive development. As such, the Counselor Cognitions Questionnaire was developed to fill the void in available instruments. It measures the complexity of counselors' cognitions about their clients. Development of the instrument and initial validation results will be described. Preliminary findings about the impact of general cognitive complexity on client-specific cognitive complexity and factors in counselor cognitive complexity will be explained.

iv TABLE OF CONTENTS Page LIST OF TABLES...........................................................................................................vii

LIST OF FIGURES..........................................................................................................ix

CHAPTER I.INTRODUCTION......................................................................................................1 Purpose of the Study.................................................................................................3 Statement of the Problem...........................................................................................7 Research Questions...................................................................................................9 Need for the Study..................................................................................................10 Definition of Terms.................................................................................................11 Brief Overview........................................................................................................12 II.LITERATURE REVIEW........................................................................................13 Counselor Cognitive Complexity...........................................................................13 The Bases of Cognitive Complexity:Theory and Measurement....................14 Kelly,1955............................................................................................14 Bieri,1955............................................................................................17 Crockett,1965........................................................................................17 Zajonc,1960..........................................................................................21 Harvey,Hunt,and Schroeder,1961.......................................................22 The Paragraph Completion Method.......................................................23 Conceptual/Integrative Complexity Method..........................................24 The Learning Environment Preferences Scale.......................................25 Loevinger............................................................................................26 Summary of The Bases of Cognitive Complexity...........................................29 Professional Development fromNovice to Expert..........................................30 Counselor Cognitive Development..........................................................................33 Developmental Models of Supervision............................................................33 Blocher,1983.........................................................................................33 Loganbill,Hardy,and Delworth,1982..................................................34

v Stoltenberg,1981...................................................................................37 Summary of Developmental Models of Supervision.......................................39 Studies of Counselor Cognitive Development using General Measures.........40 Holloway and Wolleat,1980.................................................................40 Goldberg,1974......................................................................................41 Holloway and Wampold,1986..............................................................43 Borders,Fong,and Neimeyer,1986.....................................................45 Borders,1989.........................................................................................46 Borders and Fong,1989.........................................................................48 Fong,Borders,Ethington,and Pitts,1997.............................................50 Duys and Hedstrom,2000......................................................................51 Little,Packman,Smaby,and Maddux,2005.........................................52 Brendel,Kolbert,and Fostor,2002........................................................54 Granello,2002........................................................................................55 Lovell,1999...........................................................................................56 McAuliffe and Lovell,2006..................................................................57 Summary of Studies of Counselor Cognitive Development using General Measures................................................................................58 Studies of Counselor Cognitive Development using Specific Measures.........59 Kurpius,Moran,and Benjamin,1985....................................................59 Falvey,Bray,and Herbert,2005...........................................................62 Ladany,Marotta,and Muse-Burke,2001..............................................63 Summary of Studies of Counselor Cognitive Development using Specific Measures...............................................................................66 Summary................................................................................................................66 III.METHODOLOGY.................................................................................................67 Research Questions and Hypotheses.......................................................................67 Participants...............................................................................................................68 Instrumentation........................................................................................................69 Development of the Counselor Cognitions Questionnaire..............................69 The Counselor Cognitions Questionnaire........................................................73 The Washington University Sentence Completion Test..................................77 Information Questionnaire...............................................................................81 Procedures................................................................................................................81 Data Analysis...........................................................................................................82 IV.RESULTS...............................................................................................................85 Sample Characteristics.............................................................................................85 Instrument Psychometrics........................................................................................88

vi Research Hypotheses...............................................................................................92 Research Question 1........................................................................................93 Research Question 2........................................................................................98 Research Question 3......................................................................................100 Research Question 4......................................................................................102 V.CONCLUSIONS.....................................................................................................110 Overview................................................................................................................110 Discussion..............................................................................................................110 Counselor Cognitions Questionnaire............................................................110 Counselor Cognitive Complexity..................................................................114 Research Question 1.............................................................................115 Research Question 2.............................................................................116 Research Question 4.............................................................................118 Limitations of the Study.......................................................................................123 Implications for Counseling.................................................................................124 Suggestions for Future Research.........................................................................127 REFERENCES................................................................................................................129

vii LISTS OF TABLES Page Table 1 Sample Grid..........................................................................................................16 Table 2 Stages of Ego Development..................................................................................27 Table 3 Research Questions...............................................................................................83 Table 4 Sample Characteristics..........................................................................................86 Table 5 Race/Ethnicity of All Participants by Highest Counseling Degree Completed....86 Table 6 Counseling Specialization of Participants............................................................87 Table 7 Highest Counseling Degree Completed by Sex....................................................87 Table 8 Race/Ethnicity by Sex...........................................................................................87 Table 9 CCQ Scores by Highest Counseling Degree Completed......................................88 Table 10 CCQ Inter-Rater Reliabilities.............................................................................89 Table 11 ANOVA CCQ Differentiation Score Across Counseling Specialization...........90 Table 12 ANOVA CCQ Integration Score Across Counseling Specialization.................91 Table 13 General Cognitive Complexity Scores...............................................................92 Table 14 SCT Score by Highest Counseling Degree Completed......................................92 Table 15 CCQ Differentiation Score by SCT Level..........................................................93 Table 16 CCQ Integration Score by SCT Level................................................................93 Table 17 ANOVA:CCQ Scores Across SCT Levels........................................................96 Table 18 Post-Hoc Tukey’s Analyses................................................................................97 Table 19 Variance Explained:Counseling Experience and CCQ Differentiation.............98 Table 20 ANOVA:Counseling Experience and CCQ Differentiation..............................99

viii Table 21 Variance Explained:Counseling Experience and CCQ Integration...................99 Table 22 ANOVA:Counseling Experience and CCQ Integration....................................99 Table 23 Variance Explained:Counseling Experience and SCT Score..........................100 Table 24 ANOVA:Counseling Experience and SCT Score...........................................100 Table 25 Paired Sample Statistics....................................................................................101 Table 26 Paired Sample Correlations..............................................................................101 Table 27 Paired Sample t-test..........................................................................................102 Table 28 Pearson’s r Correlation Matrix.........................................................................103 Table 29 Variance Explained CCQ Differentiation.........................................................104 Table 30 Model Significance:CCQ Differentiation........................................................104 Table 31 Coefficients:CCQ Differentiation...................................................................105 Table 32 Variance Explained CCQ Integration...............................................................105 Table 33 Model Significance:CCQ Integration..............................................................105 Table 34 Coefficients:CCQ Integration.........................................................................106 Table 35 CCQ Score Across Years of Supervisory Experience......................................107 Table 36 Parsimonious Model:CCQ Differentiation......................................................108 Table 37 ANOVA Parsimonious Model:CCQ Differentiation.......................................108 Table 38 Parsimonious Model:CCQ Integration............................................................109 Table 39 ANOVA Parsimonious Model:CCQ Integration.............................................109

ix LIST OF FIGURES Page Figure 1 CCQ Differentiation Scores Within Levels of SCT............................................95 Figure 2 CCQ Integration Scores Within Levels of SCT..................................................96

1 CHAPTER I INTRODUCTION Because issues in counseling often are complex,it is difficult for counselors to accurately understand their clients’ needs.Effective counselors must be able to identify the multiple factors that influence a client’s presenting problemand integrate these factors into a meaningful framework.This process is complicated,and requires advanced cognitive complexity.In addition,clients’ issues vary in severity and chronicity,and their preferences and receptiveness to counseling vary,even during treatment.A client’s unique combination of cognitive,emotional,behavioral,and spiritual characteristics determines her or his needs in counseling.There are endless possible combinations of these characteristics;how can counselors know all of them? Counselors cannot know themall.Instead,a counselor’s understanding of a client is limited by her or his ability to recognize relevant client variables and comprehend their combined impact on the client’s overall needs.For example,if the counselor does not recognize the client’s relationship insecurity,then the counselor cannot consider the impact of this characteristic on the client’s generalized anxiety.In this example,the counselor’s understanding of the client is incomplete and may lead to ineffective or inefficient treatment.The more complex their level of thinking is,the more client characteristics counselors can recognize and process,and thus the more accurate their understanding of the client can be (Blocher,1983).

2 Such cognitive complexity is composed of two processes (Crockett,1965).The number of client characteristics a counselor can recognize is the counselor’s level of differentiation .Understanding how those characteristics fit together is the process of integration .Counselors who have high levels of cognitive complexity are able to recognize many client variables,allowing for a sophisticated understanding of client needs.Even intricate cases of inconsistencies,mutual influences,and paradoxes can be understood and addressed.Counselors who have low levels of cognitive complexity are more likely to see clients simplistically,focusing on concrete characteristics and using black/white decision-making.Thus,increasing the complexity of counseling students’ cognitions about clients and the counseling process is a stated (e.g.,developmental models) and implied goal of counselor education programs. Indeed,counselor level of cognitive development has been linked with multiple aspects of counselor effectiveness (e.g.,Borders,1989;Fong,Borders,Ethington,& Pitts,1997;Holloway &Wolleat,1980;Ladany,Marotta,&Muse-Burke,2001). Borders (1989) reported that counseling students with higher levels of cognitive complexity were better able to remain objective in the counseling session,and Fong et al. (1997) found that students with higher cognitive development used more complex and effective verbal skills and had more confidence in their work.Borders,Fong,and Neimeyer (1986) found that only students at higher levels of ego development described clients in interactional terms.Holloway and Wolleat (1980) reported that counselors with low levels of cognitive development formulated simplistic clinical hypotheses.Cognitive complexity also has been linked with case conceptualization skills (Ladany et al.,2001).

3 In sum,counselors at higher levels of cognitive development are better able to formulate a thorough,objective understanding of the client and communicate effectively and confidently in the counseling session. These findings emphasize the vital need for facilitating cognitive development in counselor training and supervision.In fact,researchers have found support for the assumption that cognitive complexity increases during supervised counseling practice (e.g.,Duys &Hedstrom,2000;Granello,2002).To date,however,these results primarily have been based in general measures of a counselor’s cognitive complexity, despite evidence that level of complexity is content specific (Crockett,1965).That is,an individual’s level of cognitive complexity varies fromtopic to topic.For example,an individual may be able to think very complexly about flower gardening but only simplistically about adolescent mental health issues.A counseling-specific measure of cognitions is needed before researchers can fully understand counselor cognitive development.Such a measure would allow researchers to assess students’ cognitions, devise interventions that increase their students’ complex thinking,and investigate the effectiveness of these interventions.Counselor educators and supervisors then could use empirically supported interventions to maximize student opportunities for cognitive growth (Evidence-based practice in counselor education).Additional research into counselor cognitive development is necessary to achieve these goals. Purpose of the Study Although the importance of cognitive complexity seems clear,theoretically and empirically,no single,content-specific,psychometrically sound instrument exists to

4 measure counseling-specific cognitive complexity.There are well-established measures of overall cognitive developmental level,such as The Sentence Completion Test of Ego Development (SCT;Hy &Loevinger,1998;Loevinger &Wessler,1970),which captures counseling-relevant information about the respondent’s system for perceiving self,others, and relationships.It does not,however,capture cognitions specific to clients,the counseling relationship,or the counseling process.This is a significant limitation,as cognitive complexity is context-specific and can vary widely by topic within the individual (Crockett,1965).Using a general measure of cognitive complexity,then,may lead to faulty conclusions or underestimated effects. A second well-established measure of cognitive complexity is the Role Category Questionnaire (RCQ;Crockett,1965),as employed by Duys and Hedstrom(2000). Scorers of the RCQ tally the number of interpersonal characteristics a respondent uses to describe a liked peer and a disliked peer.The RCQ appears to be a good assessment of interpersonal construct differentiation ,but the respondent’s ability to integrate these constructs is not measured,despite the critical nature of both processes.In addition,it is unclear if one’s complexity of cognitions about a peer is representative of one’s complexity of cognitions about a client. Certainly,cognitive-oriented instruments designed specifically for counseling research are available.These clinical assessments have limited psychometric evidence, however,or are focused on only one aspect of the content of counselor cognitions (e.g., treatment planning).For example,Falvey (Falvey,1994;Falvey et al.,2005) created the Clinical Treatment Planning Simulation to gather information about case

5 conceptualization,diagnosis,and treatment planning.Respondents’ written case conceptualizations are scored using an unwieldy scoring rubric developed by expert panelists.The respondent receives points for including pieces of information about the case that the experts deemed important (e.g.,symptoms such as impulsivity or forgetfulness,diagnostic criteria such as impairment at school and home) and loses points for omitting important information (e.g.,duration of the problem).Thus,the scores reflect counselor content knowledge of the symptomology and treatment practices for the single client issue presented and have no apparent relationship with the counselor’s level of cognitive complexity. Kurpius,Benjamin,and Morran (1985) used a thought-listing technique in their study of counseling cognitions.Respondents were given a total of 12 minutes to list thoughts stimulated by video clips of a client.Raters then awarded one point for each thought that represented one of four clinical hypothesis categories:1) reflecting the client issues,2) stating a factor of the client’s issue,3) relating a factor to the issue,and 4) differentiating or integrating problems.In addition,one point was given for thoughts that asked or answered questions about the client’s issue,indicated future plans for counseling,revealed counselor coping skills,or supported the counselor’s self-efficacy. A sumscore was used as an indication of the respondent’s ability to performthe conceptual tasks required of counselors.Herein lies the weakness of this instrument.For example,listing five client issues (e.g.,drug abuser,unemployed,poor,homeless,and suicidal) would earn the participant a score of five.According to this scoring system, listing five much more complex thoughts also would earn the a score of five (e.g.,

6 dejected,lack of purpose in life,few meaningful relationships,low self-efficacy about work,lack of positive role models).In effect,the sumscore does not truly reflect the differences in the cognitive abilities of the participants.Though this type of measure seems relevant to counseling performance,it does not appear to be a precise measure of the complexity of cognitions about clients or counseling. Ladany,Marotta,and Muse-Burke (2001) studied the relationship between counselor complexity of case conceptualization and supervisory style preference. Participants were given a written intake of a client and asked to write at least three sentences describing what they believed to be the origins of the client’s issue and an effective treatment plan for addressing the issue.Case conceptualization integrative complexity was measured using a social perception procedure developed by Suedfield, Tetlock,and Streufert (1992).Coders rated participant responses based on four levels of integrative complexity:1 = low differentiation and integration ,2 = moderate/high differentiation and low integration ,3 = moderate/high differentiation and moderate integration ,4 = high differentiation and integration .The differentiation score reflected a sumof the number of reported factors in the origin of the client’s issue (similar to Kurpius et al.,1985).Coders assigned an integration score based on participants’ articulation of interrelated factors in the client issue or multiple options for the treatment plan.The wording chosen for the directions is a major threat to the validity of the measure.“Write at least three sentences” does not tell the participant that the goal is to provide a thorough case conceptualization and treatment plan,which is what the researchers contend that the participant response represents.In addition,it is unclear if

7 this method captures information about the counselors’ level of cognitive complexity or merely their knowledge of causes and symptoms of the client’s presenting problem, similar to Falvey’s (1994) measure. These measures capture information about the amount of knowledge participants have about counseling relevant topics and tasks (e.g.,recognizing symptomology as outlined in the Diagnostic and Statistical Manual of Mental Disorders [DSM;American Psychiatric Association,2005] and developing a treatment plan based in the client’s presenting issues).Although a counselor’s level of cognitive complexity underlies these tasks and limits the sophistication of the counselor’s performance,measuring performance on the task itself does not necessarily represent cognitive complexity. Statement of the Problem The existing instruments,briefly described above,explicate some of the challenges in measuring counselor cognitive complexity.The instruments provide methods for collecting cognitions (e.g.,sentence completion,thought listing,paragraph completion) or assessing counseling performance (e.g.,diagnosis and treatment planning),but no single instrument has both a counseling-specific complexity measure and psychometric stability.The goal of this project,then,is to investigate the psychometrics of a new instrument,the Counselor Cognitions Questionnaire (CCQ; Welfare &Borders,2006),as a reliable measure of counselor cognitive complexity specifically about clients that can easily be used by counselor educators,supervisors,and researchers.Each research question will gather evidence of validity of the CCQ. Additionally,the results will clarify aspects of counselor cognitive complexity.

8 Instrument development is a multifaceted,multiphase task.This study will contribute to the validation of the CCQ by exploring its empirical relationship with Loevinger’s general measure of cognitive complexity,The Sentence Completion Test (Hy &Loevinger,1998;Loevinger,1976;Loevinger &Wessler,1970).The SCT measures conceptual developmental level across 10 ego stages.It is the most counseling- relevant measure of general complexity because it captures relevant information about how respondents perceive themselves,others,and relationships.Since cognitive complexity level may vary across different topics,however,the general measure may not accurately represent cognitive complexity completely as it relates to counseling.Some individuals who have high overall levels of cognitive development may have low levels of cognitive complexity about clients.Two individuals with identical levels of general cognitive development may differ in counselor cognitive complexity,perhaps due to their training or experience level.On a group level,however,a direct relationship between the two is expected. Thus,a second focus of the study is to investigate the effect of counseling experience on general cognitive complexity and client-specific cognitive complexity.In past research on counselors,general cognitive complexity has been found to change little over time,perhaps because of the broad levels of cognitive complexity represented by the scores (Hy &Loevinger,1996).Nevertheless,complexity of counselor cognitions about clients is believed to increase during training and counseling experience. In addition,this study will explore the empirical relationship between level of counselor cognitive complexity and self-perceived counselor effectiveness.The

9 complexity of counselor cognitions about a client with whomthey have felt effective will be measured,as will complexity of cognitions about a client with whomthey have felt less effective.Past literature supports the assumption that highly complex counselors demonstrate more skills of effective counselors (e.g.,Borders,1989;Fong,Borders, Ethington,&Pitts,1997;Holloway &Wolleat,1980;Ladany,Marotta,&Muse-Burke, 2001).It is expected that even within the individual,cognitions about clients with whom the counselor feels effective will be more complex than cognitions about clients with whomthe counselor felt ineffective. Finally,identifying counselor characteristics that influence cognitive complexity is a major step toward understanding counselor development.Some evidence suggests that age (Loevinger &Wessler,1970) and experience (e.g.,Duys &Hedstrom,2000) impact cognitive complexity.Other researchers (e.g.,Borders,1989) have speculated that overall cognitive complexity has a ceiling effect on domain-specific cognitive complexity.A multiple regression analysis will yield empirical support for the impact of these factors on counselor cognitive complexity. Research Questions Accordingly,this study is designed to address the following research questions: 1.Are counselors’ scores on the Counselor Cognitions Questionnaire significantly different across Sentence Completion Test (SCT) ego development levels? 2.Does duration of counseling experience predict scores on the measures of cognitive complexity (Counselor Cognitions Questionnaire and Sentence Completion Test)?

10 3.Do scores on the Counselor Cognitions Questionnaire vary significantly between clients with whomthe counselor felt effective and less effective? 4.What factors (i.e.,SCT score,age,graduate training,paraprofessional work experience,and counseling experience) influence scores on the Counselor Cognitions Questionnaire? Need for the Study A user-friendly,valid,reliable measure of counselor cognitive complexity is needed to facilitate research on counselor cognitive development.With such an instrument,researchers could identify the cognitive factors most critical to counselor effectiveness.Such knowledge also would informcounselor education and supervision curriculum.Interventions to facilitate cognitive development could be created, empirically tested,and integrated into curricular experiences in order to maximize counselor effectiveness and therefore client outcomes.Measures currently available are arduous to administer or score,which limits their usability,or focus on something other than complexity of counselor cognitions about clients (e.g.,general cognitive development,content knowledge of diagnostic indicators,and complexity of cognitions about peers).With a quick and accurate assessment tool,counselor educators and supervisors could easily evaluate their students’ progress.To achieve these goals,the measure must be simple to administer and straightforward in scoring even though cognitive processes are complex.The goals of this study are to further validate a new instrument,the Counselor Cognitions Questionnaire,as a measure of counselor cognitive complexity and explore factors that impact counselor cognitive development.

11 Definition of Terms Cognitive complexity is the level of differentiation and integration in an individual’s cognitive system(Crockett,1965). Counselor cognitions are counselors’ thoughts and impressions of their clients and the counseling relationship. Differentiation is the number of available constructs in an individual’s cognitive system about a domain. Integration refers to the ability to recognize relationships among cognitive constructs about a particular domain. Construct refers to a cognitive template through which an individual understands or assigns meaning to her or his world.A construct also can be called a schema. Cognitive development refers to an increase in an individual’s level of cognitive complexity. Ego developmental levels are part of a ten-stage systemfor grouping individuals according to the complexity of their cognitive,interpersonal,and moral perceptions of self and others (Hy &Loevinger,1996). Counseling Experience is the duration of time a counselor has provided professional counseling services. Supervision is an individual or group educational process in which a supervisor attempts to facilitate the development of a supervisee by processing her or his experiences providing counseling services.

12 Supervisory Experience is the duration of time an individual has provided counseling supervision. Brief Overview This study is presented in five chapters.The first chapter has provided an introduction to cognitive complexity,cognitive development,and rationale of the CCQ. The purpose of the study,statement of the problem,and need for the study are outlined in this introduction.In addition,definitions of key terms are included.The second chapter contains a review of the literature as it relates to cognitive psychology,counselor cognitive development,and cognitive assessment.The third chapter includes the methodology to be used in the study,including participants,sampling method, instruments,and data analyses.A detailed scoring protocol for the CCQ is explained, and development and preliminary validation to date are presented.The fourth chapter presents the results of this research according to each research question.Finally,the fifth chapter summarizes the study and includes limitations and recommendations for future research in the area of counselor cognitive development.

13 CHAPTER II LITERATURE REVIEW Counselor Cognitive Complexity Client issues in counseling are often complex.Effective counselors are able to identify and integrate multiple factors to reach an accurate understanding of complex client needs.That process of identifying and integrating pieces of information occurs in the counselor’s cognitive system.Individuals with highly developed systems can understand varied and intricate experiences,while individuals with less developed cognitive systems understand experiences more simplistically.Thus,the complexity of a counselor’s cognitive systemis fundamental to effective practice. Counselor educators and supervisors strive to prepare students to be effective counselors.Counselor cognitive development is a recognized goal of counselor preparation programs.Research into the effectiveness of counselor education practices informs curricular changes that improve counselor preparation.To this end,this study explores the measurement of counselor cognitive complexity.The following review includes the literature on general cognitive development,assessment methods that measure cognitive complexity,the development of professional expertise,empirical research regarding cognitive development,and qualitative and quantitative research on counselor cognitive development.

14 Several definitional distinctions must be considered in reading this review. Unless otherwise specified,cognitive complexity refers to the level of differentiation and integration in an individual’s cognitive system(Crockett,1965).Differentiation is the number of available constructs in an individual’s cognitive systemabout a domain,such as a client.Integration refers to the ability to recognize relationships among cognitive constructs in a particular domain.A construct is a cognitive template through which an individual interprets or assigns meaning to her or his world.A construct also can be called a schema.Cognitive development refers to an increase in an individual’s level of cognitive complexity. The Bases of Cognitive Complexity:Theory and Measurement Kelly,1955 Personal Construct Theory (Kelly,1955) is based on the assumption that individuals create conceptual templates (or constructs) that allow themto understand the things that they experience.Constructs are activated by stimuli and then used by the individual to understand the situation and respond appropriately.Kelly contended that individuals can refine the meaning they ascribe to a stimulus by learning from experiences.Therefore,an individual’s systemof cognitive constructs can become more complex over time.Each person’s systemis unique and based on her or his own experiences.Kelly defined a personal construct as an awareness of how two things are alike in a way that differentiates themfroma third thing.In this definition,a construct is dichotomous and mutually exclusive.That is,something can be small or it can be large, but not both.In this example,the construct is small-large.When exposed to a stimulus

15 “paperclip,” one’s construct of smallness or largeness is activated (i.e.,Is this itemsmall or large?).The paperclip is classified as small.That classification is part of the individual’s understanding of the stimulus “paperclip.” Multiple constructs are activated for each stimulus,depending on the context of the stimulus. The Construct Grid Method (CGM;Kelly,1955) is the method Kelly developed to assess the complexity of an individual’s construct systemabout a particular domain. The original method,called The Role Construct Repertory Test,has been modified to increase its versatility.Persons taking a traditional CGMare asked to consider in what important way two elements are alike and how those two elements are different froma third element.For example,Person A and Person B are both timid,while Person C is assertive.This assessment technique is very flexible in that it can be applied to any domain (e.g.,vegetables,people,countries).The elements (i.e.,Persons A,B,and C in the above example) can be designated by the assessor or the respondent (e.g.,How are your two brothers Andy and David alike and different from your brother Ben?). However,Kelly recommended asking the respondent to generate the construct (i.e.,timid - assertive in the above example).In a single construct grid the respondent considers six or more elements and many constructs,depending on the situation.The “grid” is developed with the elements and constructs.There are many methods for creating and scoring the grid.A common method is using a rating scale grid (Beail,1985).In this method each element (e.g.,client A,client B,and client C,etc.) is given a rating on each construct (e.g.,1 = very timid,2 = somewhat timid,3 = somewhat assertive,4 = very assertive).

16 Table 1 Sample Grid Client A Client B Client C Client D Client E Client F Client G Client H Timid 1 1 4 2 4 3 1 2 Assertive Reliable 1 2 2 4 2 2 3 1 Unreliable Analysis of the ratings is complex.Graphs,correlations,factor analyses,and cluster analyses have been used to analyze grid scores (Landfield &Epting,1987).All analyses measure the similarities and differences among the elements and constructs. Because of the adaptability and its relevance to understanding how people understand other people,the grid technique has been used in counseling research. Borders,Fong,and Neimeyer (1986) used the CGMto assess counselor cognitions about clients.Respondents listed eight clients who were then grouped into eight combinations of three clients.The respondent was presented with each grouping of three and asked to describe how two of the clients were similar and how they were different fromthe third client (e.g.,these two clients are “straightforward” while the third is “manipulative”). However,due to the complexity and variety of the scoring procedures,the CGMhas not been used often.In addition,the measure does not capture the respondent’s complete cognitions about an individual or client.The constructs chosen may not represent the full breadth or depth of the respondent’s understanding.

17 Bieri,1955 Bieri (1955),a student of Kelly,defined cognitive complexity as the capacity to interpret social behavior in a multidimensional way.He believed that individuals use their perceptions to predict others’ behaviors.Cognitive systems that are more complex allow more highly differentiated impressions of people,while simple cognitive systems provide poorer differentiation .Simple cognitive systems can develop into more complex cognitive systems with experience.Bieri measured cognitive complexity using Kelly’s (1955) construct grid,but,rather than analyzing the response patterns or differences among the constructs on the grid,he simply counted the number of constructs used by the respondent.Bieri contended that the number of constructs in a sample is representative of the total number of constructs in the respondent’s cognitive system.Using the CGM to obtain a count of constructs clarified the theoretical definition of cognitive complexity. Operationally defined by this assessment technique,cognitive complexity is the number of constructs in the cognitive system.One limitation of this method is that the sample of constructs on the grid is dependent on the elements in question.That is,if the grid is used to explore family relationships,then the elements are family members and the constructs are about the family members.Such a narrow sample may not be representative of overall cognitive complexity,nor complexity of elements in other domains. Crockett,1965 Walter Crockett (1965),also a student of Kelly,expanded on his mentor’s theories in his chapter,“Cognitive Complexity and Impression Formation.” Crockett

18 contended that an individual’s impression of another person is a function of the behavior and appearance of the perceived person;the relationship between the perceiver and the perceived;and the cognitions,beliefs,motives,intentions,personality,and psychological state of the perceiver.The perceiver only directly observes a few characteristics of the perceived,but makes extended inferences about many other characteristics.The inferences are made in the perceiver’s cognitive system.Crockett focused on the complexity of the cognitive system as it relates to the process of forming impressions of others.Like Kelly and Bieri,he too believed that a person’s cognitive systembecomes more complex over time as he or she encounters new experiences.Crockett clarified, however,that if an individual does not experience new stimuli in a particular domain,her or his cognitive systemin that domain will not become more complex.Therefore,the complexity in one domain does not represent the complexity in another domain or overall complexity.The domain-specific nature of cognitive systems is quite relevant in understanding and conducting research on cognitive complexity.For example,the current study is targeting the complexity of cognitions that counselors have about their clients.Per Crockett’s dictum,measuring the complexity of cognitions about anything other than clients would not be a valid measure of a counselor’s complexity of thinking about a client. Crockett (1965) also provided a specific definition of cognitive complexity:A complex cognitive systemcontains a large number of constructs that are integrated hierarchically.He defined two aspects of cognitive complexity:the number of constructs in the systemis the degree of cognitive differentiation ,while the complexity of the

19 relationships and connections among the constructs is referred to as the degree of integration .Crockett explored ways to measure differentiation and integration .First,he explained that an exhaustive measure of differentiation and integration is not possible. Instead,researchers must use a targeted sample of the cognitive system and assume that the sample is representative of the whole.Crockett developed a measure that sampled a broader range of constructs than the Kelly (1955) or Bieri (1955) measures.The Role Category Questionnaire (RCQ;Crockett,1965) captures respondent descriptions of eight different individuals in varying roles.The roles include,for example,an older,liked man and a same age,disliked female.The total number of unique constructs used to describe the eight people is the respondent’s interpersonal construct differentiation score.This measure of construct differentiation reportedly is more valid than past measures since it includes impressions of different types of people (Crockett,1965). Scoring of the RCQ requires identification of the individual constructs listed in the response.Individual constructs are defined as any characteristic,quality,trait, motivation,belief,habit,mannerism,or behavior (Crockett,1965).A point is given for each unique construct listed in the description of each peer.If,within the response,two phrases seemto mention similar constructs,the respondent is given the benefit of the doubt and multiple points are awarded.This is the case except in the case of adjectives describing a noun,in which case the two words represent one construct and are awarded one point.For example,“clever and mischievous” would earn two points,while “insensitive jerk” would earn one.Finally,general statements about humanity are not

20 scored.That is,“we all have our faults” or “I wish people were more like him” would not earn points. Coders are trained using the manual created by the authors (Crockett et al.,1974). The recommended process includes reading the coding rules,practicing by coding sample responses,discussing the coding rules used to rate each response,and seeking guidance froman experienced coder.As a final stage,a novice and expert coder should rate the same responses and inter-rater reliability should be assessed.Inter-rater reliability in studies using this instrument frequently exceed 0.90,demonstrating sufficient reliability for use in research.Crockett reported a 4-month test-retest reliability of 0.95 for the eight role version of this instrument (Crockett,1965).A more recent version of the RCQ includes only two roles (liked and disliked peer).It earned 4-week test-retest reliability scores of 0.84 and 0.86 (O’Keefe,Sheppard,&Streeter,1982).These scores represent adequate test-retest reliability for group level data with adult participants.Use of the RCQ with children or adolescents may show lower test-retest correlations due to the natural development of respondents in that age group. Evidence of validity is also available for the RCQ.Researchers (e.g.,Beatty & Payne,1984) found RCQ scores to be positively correlated with adults’ performance on a measure of social perspective taking called the Social Perspectives Task (Hale &Delia, 1976).Angell (2000) found scores on the RCQ and the Paragraph Completion Task had a positive correlation (r =.37,p <.05).This finding also supports the hypothesis of this study that general cognitive complexity is related to,but not necessarily synonymous with,complexity of cognitions about a specific domain.Angell also found the RCQ and

Full document contains 149 pages
Abstract: Because issues in counseling are often complex, it is difficult for counselors to accurately understand their clients' needs. A counselor's understanding is limited by his or her ability to recognize relevant client variables and comprehend their interactional impact on the client's overall needs (Blocher, 1983). This process is complicated, and requires advanced cognitive complexity. Counselor cognitive complexity has been linked with multiple aspects of counselor effectiveness (e.g., Borders, 1989; Fong, Borders, Ethington, & Pitts, 1997; Holloway & Wolleat, 1980). In fact, researchers have found support for the assumption that cognitive complexity increases during supervised counseling practice (e.g., Duys & Hedstrom, 2000). To date, however, these results primarily have been based in general measures of a counselor's cognitive complexity, despite evidence that level of complexity is domain specific (Crockett, 1965). Crockett (1965) reported that an individual's level of cognitive complexity can vary from topic to topic. Because of the domain-specific nature of cognitive complexity, a counseling-specific measure of cognitions is needed before researchers can fully understand counselor cognitive development. As such, the Counselor Cognitions Questionnaire was developed to fill the void in available instruments. It measures the complexity of counselors' cognitions about their clients. Development of the instrument and initial validation results will be described. Preliminary findings about the impact of general cognitive complexity on client-specific cognitive complexity and factors in counselor cognitive complexity will be explained.