|
Article Excerpt
The authors describe the development of the California Brief Multicultural Competence Scale (CBMCS). The 21-item CBMCS was derived from principal component analysis, item content validated by a panel of experts, and confirmatory factor analyses. Several studies provided internal consistency, subscale intercorrelations, criterion-related validation, and assessment of possible social desirability contamination.
**********
How adequately have training programs prepared professional psychologists to work with culturally diverse populations? This question has been a focus of multicultural competence research for 3 decades. In 1973, the American Psychological Association (APA) endorsed multicultural training recommended by the National Conference on Levels and Patterns of Professional Training (Korman, 1976). The American Counseling Association compiled guidelines for standard test usage with multicultural populations (Prediger, 1994). Early and more recent comprehensive APA guidelines on multicultural education, training, research, practice, and organizational change addressed the centrality of cultural issues for all psychologists (APA, 1993, 2003). However, implementation in professional psychology has been meager except in Counseling Psychology programs (LaFromboise & Foster, 1992; Mio & Morris, 1990), and the characteristics of multicultural competence training have remained elusive (e.g., Murphy, Wright, & Bellamy, 1995; Quintana & Bernal, 1995).
A national examination of multicultural training suggested a model for quality multicultural training (Ponterotto, 1997, 1998), although sensitivity rather than proficiency has apparently resulted from additional training. Each area of professional psychology has reported survey data suggesting limited efficacy for available training (for a review, see Ponterotto & Alexander, 1996). Psychologists in most professional programs report inadequate formal training and supervised experience to prepare them for practice with multicultural clients (Allison, Crawford, Echemendia, Robinson, & Knepp, 1994). Adequacy of self-perceived awareness and skills as reported in a more recent survey of multicultural competencies (Holcomb-McCoy & Myers, 1999). In these two surveys, differences among participants in training, ethnicity, program affiliation, and graduation year may account for differences in findings. Other studies (e.g., Pope-Davis, Reynolds, Dings, & Nielson, 1995) also found that greater self-perceived multicultural competencies among students of color and ethnic minority status per se can provide in vivo cultural knowledge and personal experience with discrimination (Lee & Richardson, 1991; Ponterotto, Casas, Suzuki, & Alexander, 2001; Robinson & Ginter, 1999).
MULTICULTURAL COUNSELING COMPETENCY MODEL
The Cross-Cultural Counseling Competency Model (for a review, see Ponterotto, Fuertes, & Chen, 2000) contained in a Division 17 position paper (Sue et al., 1982) specified 11 competencies in three broad areas, including attitudes/beliefs, knowledge, and skills. Preserving these three superordinate areas, this model was subsequently expanded to 31 specific competencies (Sue, Arredondo, & McDavis, 1992). Operationalization of these competencies by 119 explanatory statements, within an American Counseling Association training context (Arredondo et al., 1996), led to an even more comprehensive formulation by Divisions 17 and 45 (Sue et al., 1998) with a further expansion to 34 competencies.
MULTICULTURAL COMPETENCY INSTRUMENTATION
A number of instruments have been developed to measure multicultural competencies as conceptualized in the Cross-Cultural Counseling Competency Model including the Cross-Cultural Counseling Inventory-Revised (CCCI-R; LaFromboise et al., 1991), the Multicultural Awareness, Knowledge, Skills Survey (MAKSS; D'Andrea et al., 1991), the Multicultural Counseling Awareness Scale-Form B (MCAS-B; Ponterotto & Alexander, 1996) the Multicultural Counseling Knowledge and Awareness Scale (MCKAS; Ponterotto, Gretchen, Utsey, Rieger, & Austin, 2002), the Multicultural Counseling Inventory (MCI; Sodowsky et al., 1998; Sodowsky et al., 1994), and the Multicultural Competency and Training Survey (MCCTS; Holcomb-McCoy, 2000). (The MCAS-B was used in the present study because the new version [MCKAS] was not available during the initial phases of our research.)
Several aspects of the development of these instruments and validation of scores from the instruments are relevant for purposes of comparison and for evaluating the adequacy of construct delineation. Evidence for psychometric properties (i.e., origin of items, scale development, scoring procedures, reliability, validation evidence, subscale development and clarity of subscale interpretation, and factor labels) have been examined for scores from each of the aforementioned instruments (Ponterotto & Alexander, 1996; Pope-Davis & Dings, 1995). Although the items in these instruments had a common source in the multicultural competency model, each instrument implemented somewhat different methods, and several versions of the model were used to create an item pool. Scale development procedures differed, numbers of items varied across instruments, and scoring procedures either added or averaged items. Internal consistency reliabilities were, in general, adequate. There was low to moderate subscale score consistency within instruments, but less than desirable subscale score consistency across instruments, while validation enterprises differed in kind and adequacy across instruments.
All instruments used exploratory factor analyses, although the MCI, MCAS-B, and MCKAS used confirmatory factor analyses as well. These factor analyses yielded one, three, or four factors with a variety of labels reflecting the structure coefficients obtained. The triad of original constructs was augmented by factors for Relationship in the MCI and Racial Identity Development and Multicultural Terminology in the MCCTS. The CCCI-R is uniquely a unidimensional measure, based on supervisor ratings, rather than a performance self-assessment by clinicians. Recent studies (i.e., Constantine et al., 2002; Ponterotto et al., 2002) delineated two consistent factors corresponding to self-perceived multicultural counseling skills and multicultural attitudes/beliefs.
A number of critical reviews from 1985 to 2001 provided evaluations of the multicultural competence construct (e.g., Constantine & Ladany, 2001; Ponterotto & Alexander, 1996; Ponterotto, Fuertes, et al., 2000; Ponterotto & Furlong, 1985; Ponterotto, Rieger, Barrett, & Sparks, 1994; Pope-Davis & Dings, 1995; Pope-Davis et al., 2001; Pope-Davis & Nielson, 1996). These reviews convey only limited support for the three-factor model. These instruments measure beliefs about services rather than attitudes, and caution is advised in using these instruments for the measurement of multicultural competence per se (Constantine et al., 2002).
Pope-Davis and colleagues (2001) examined consequences of omitting consumer perspectives on multicultural competency skills that included expectations and experiences. Research has been affected by inadequate empirical data and sociopolitical attitudes. Legitimate questions have been raised concerning multicultural competencies with systems larger than one individual client or with noncounseling interventions (Constantine & Ladany, 2001; Ladany, Inman, Constantine, & Hofheinz, 1997).
Research using these multicultural competency instruments provided promising correlational and regression analysis relationships to demographic and training variables; case conceptualization skills; and hypothesized, linked constructs (Ponterotto, Fuertes, et al., 2000). In summary, these authors suggested that the findings contributed to an enlarged intercultural perspective with a nonracist personal stance, enhanced worldview, awareness of oppression, and racial identity development. Statistical examination of social desirability influence has been recommended (Worthington, Mobley, Franks, & Tan, 2000). Sodowsky et al. (1998) reiterated this caveat using a new multicultural-specific social desirability scale with the MCI.
The sheer numbers and quality of thoughtful criticisms focused on these instruments suggest their perceived importance for professional training in counseling and psychology. However, there are substantive limitations to all instruments when they are used as diagnostic precursors to specific multicultural competence training objectives or as optimal pre-post measures of the effectiveness of such training. Discrepancies in historical findings may be attributed to the use of several independent instruments developed to measure the same constructs. It has been concluded that the "collective group of instruments is still only in their infancy" (Ponterotto, Fuertes, et al., 2000, p. 646), and continued factor analytic research on a larger scale with all instruments has been recommended (Ponterotto & Alexander, 1996). A single instrument, developed from available instruments, could potentially resolve some of the reported critical issues and shortcomings.
A RATIONALE FOR RESEARCH ON MULTICULTURAL COMPETENCY INSTRUMENTS
Multicultural competency instruments were constructed and evaluated primarily in the context of counseling training programs. These programs endorse the centrality of cultural issues, the coextensive nature of multicultural and clinical competencies, and the primacy of culture as a proximal variable in research. Because an acknowledged limitation of multicultural competency instruments is measurement of self-efficacy beliefs rather than skills, objective relationships between these beliefs and demonstrated ability to provide effective mental health services to culturally diverse clients are ultimately necessary (Lent, Brown, & Larkin, 1986). The research history of these instruments endorses their continued use; nonetheless, the lack of uniformity in what they measure suggests that their use to evaluate provider competencies is premature. There has been a call for more valid multicultural counseling instruments to replace the original measures (Atkinson & Israel, 2003), because their item contents were derived from committee consensus rather than empirical identification. It is also necessary to make the most responsible and efficacious use of these existing measures to ensure their continued presence as progenitors of more adequate training that can be evaluated by subsequent clinical outcomes. The potential utility of these existing measures may be simply diagnostic in the sense of legitimizing in-service training as an intervening step prior to evaluation of clinician skills with clients.
The primary objective of this research is to improve multicultural competence instrumentation as a precursor to consistent and replicable multicultural competence training for clinicians within a comprehensive system of mental health care. A secondary objective is to embed competence assessment and training within a quality-of-care model for ethnic minority populations. Ponterotto, Gretchen, and Chauhan (2000) described a number of relevant models, including the Multicultural Assessment-Intervention Process model (MAIP; see Dana, 1993, 1998, 2000) that provides a system of care context in which the present research is a major component. MAIP treats culture as a proximal variable with successive steps for adequate mental health care that include service delivery style, client acculturation or racial identity status, client-clinician matching, multicultural competence training, and use of cultural components in interventions within a format of pre-post client evaluation.
CONSTRUCTION AND DEVELOPMENT OF THE CALIFORNIA BRIEF MULTICULTURAL COMPETENCE SCALE (CBMCS)
This research was designed to create a single, brief instrument from items in several multicultural counseling competency measures predicated on responses from direct-service community mental health providers. The development of this new instrument responds to the call for continued scrutiny of the multicultural competency model itself and the operationalization of these beliefs concerning self-perceived competencies (Ponterotto et al., 2002).
This article presents the results of four studies. Study 1 examines items from subscales in the self-report instruments for social desirability. Subsequently, using these instruments, data from a relatively large sample of mental health practitioners were obtained. Thus, in Study 2 (which is subsequently trichotomized into Studies 2a, 2b, and 2c), a short form version of the CBMCS was developed from four instruments (CCCI-R, MAKSS, MCAS-B, MCCTS), factor analyzed, scrutinized by a panel of experts for content validation, and further verified by means of a confirmatory factor analysis. Study 3 provided an additional check for social desirability bias within the CBMCS subscales. Study 4 provided further support of the CBMCS with an additional confirmatory factor analysis on an independent sample. This brief instrument (the CBMCS) can be used as an empirical basis for the development of training programs that can be evaluated for efficacy in treatment settings.
STUDY 1: SOCIAL DESIRABILITY CHECK
This study related the Marlowe-Crowne Social Desirability Scale (SDS; Crowne & Marlowe, 1960) to each of the individual items in a comprehensive item pool representing four measures in order to eliminate those items correlating with the SDS from subsequent analyses.
Method
Participants. The Study 1 participants constituted a convenience sample of 54 mental health service providers at a Southern California community mental health agency. The...
|
