|
Article Excerpt
This article describes the revision of the White Racial Consciousness Development Scale (D. Claney & W. M. Parker, 1989). A multistage approach including item generation, item refinement and selection, and evaluation of score validity and reliability was used to test construction and validation. Implications for theory, practice, and future research are discussed.
**********
Helms (1984) proposed a theory of White racial identity development based on the assumption that as White counselors gain a better understanding of their own racial identity attitudes, they will be more adept at understanding those of others. Although Helms's (1990) work has served as a catalyst for others to suggest alternative conceptualizations of White racial identity development (Ponterotto, 1988; Sabnani, Ponterotto, & Borodovsky, 1991; Sue & Sue, 1999), hers is considered to be the principal theory in the area, at least in part because of the availability of measures to assess it (Leach, Behrens, & LaFleur, 2002).
According to Helms's (1984, 1990, 1995) White racial identity theory, White people move from being naive about the differences between Blacks and Whites to learning to accept and appreciate the differences between races. Helms (1984) initially described five stages of White racial identity development: contact, disintegration, reintegration, pseudo-independence, and autonomy.
In Helms's (1995) revision, she changed the term stages to statuses. In each status, people are theorized to use different information-processing strategies or schemas to filter and interpret racial cues. Multiple statuses can coexist, and most people use several schemas to process information. The emergence of different statuses occurs in a particular sequence based on differentiations of the ego, with more mature statuses being more sophisticated than earlier statuses. The status that is used by a person in a given situation is determined by the dominance of the schema in the personality structure of the individual.
Although White racial identity models have been widely used and have intuitive appeal for educational, clinical, and research application, there are still problems in defining the underlying constructs of the model. Part of the challenge of refining the models of White racial identity involves creating reliable and valid scales to measure the constructs included in the theory.
Three major scales have been developed to assess White racial identity development: the White Racial Consciousness Development Scale (WRCDS; Claney & Parker, 1989); the White Racial Identity Attitude Scale (WRIAS; Helms & Carter, 1990); and the White Racial Identity Development Scale, developed by Corbett, Helms, and Reagan (1992), which measures the immersion/emersion status of White racial identity development. The White Racial Identity Development Scale has not been used in many published research studies (Leach et al., 2002). Another scale, the Oklahoma Racial Attitudes Scale (ORAS; Choney & Behrens, 1996), measures White racial consciousness according to Rowe, Bennett, and Atkinson's (1994) theory of White racial consciousness, developed in reaction to Helms's (1984, 1990) theories.
Rowe et al. (1994) defined White racial consciousness as one's awareness of being White and what that implies in relation to those who do not share White group membership. Although Rowe et al. maintained that their model is superior to Helms's (1990) model, Block and Carter (1996) questioned the validity of this claim. Block and Carter stated that "a rose by any other name is still a rose" (p. 327). Although the ORAS (a scale based on White racial consciousness theory) was designed to measure the White racial consciousness construct, the scale is not discussed in this article because it is not based on Helms's (1984, 1990, 1995) theory.
Claney and Parker (1989) were the first to develop a scale based on Helms's (1984) theory of White racial identity. The scale consists of 15 Likert-type items, with 3 items pertaining to each of Helms's (1984) proposed stages: contact, disintegration, reintegration, pseudo-independence, and autonomy. When the WRCDS was developed, the immersion/emersion status was not yet developed by Helms (1984) in her theory of White racial identity development. Since its development, the WRCDS has been used or suggested in a variety of studies (Parker & Moore, 1998; Sabnani & Ponterotto, 1992; Sabnani et al., 1991). The reliability of scores on the WRCDS using a Guttman scale for reliability was low, with the highest reliability reported being .56 on the Reintegration subscale and the lowest being .08 on the Autonomy subscale.
In 1990, Helms and Carter developed the WRIAS to assess White racial identity based on Helms's (1984) theory of White racial identity development. The WRIAS has more items than the WRCDS, with 10 items representing each status. As the WRIAS was increasingly used in the 1990s, questions regarding the reliability and validity of scores emerged. The scale has been criticized for lacking a strong empirical foundation, and it appears that the WRIAS does require item revision to improve the internal consistency and construct validity of its scores (Block, Roberson, & Neuger, 1995; Swanson, Tokar, & Davis, 1994). For example, Swanson et al. found that the Contact subscale lacked cohesiveness and that 32 of the total 50 items loaded on a general factor, leading to the conclusion that the data "do not support the psychometric adequacy of the WRIAS" (p. 215). In addition, others have raised questions concerning the reliability of the Contact subscale scores (Ottavi, Pope-Davis, & Dings, 1994). Studies by Carter (1990) and Pope-Davis and Ottavi (1992) upheld the scale's reliability of scores, yet other researchers have obtained low reliability coefficients, especially for the Contact, Pseudo-Independence, and Autonomy subscales (Block et al., 1995).
Choney and Rowe (1994) critically discussed both the WRCDS and the WRIAS. According to Choney and Rowe, the WRCDS's low internal consistency and lack of concurrent validity of scores with Helms and Carter's (1990) WRIAS raises serious questions as to the ability of the WRCDS to measure Helms's (1984, 1990) White racial identity theory constructs. Choney and Rowe suggested that adding more items to measure each subscale (i.e., Contact, Disintegration, etc.) could increase the reliability of scores on the WRCDS.
PURPOSE OF THE STUDY
Although three scales have been developed and used widely to measure Helms's (1984, 1990) theory of White racial identity, all have received criticism for either the lack of reliability of scores (Choney & Rowe, 1994) or for the lack of validity of scores (Behrens, 1997; Bennett, Behrens, & Rowe, 1993; Choney & Rowe, 1994; Tokar & Swanson, 1991). The purpose of this research study was to improve the validity and reliability of WRCDS scores by adding more items to each of the subscales of the WRCDS. We decided to update the psychometric properties of the WRCDS given our collegial relationship with one of the scale developers who encouraged our continuing research on the measure. Our hope was that a revised scale will be a psychometrically sound instrument that counselor educators and practitioners can use to help understand how they (if White), their White students, their White clients, and their White supervisees think about Black people/people of other races (Helms, 1995).
The project design had five stages. The purpose of the first stage was to develop new items for the scale based on the responses of White students recorded from both individual interviews and focus groups. The second stage of the project involved item refinement and selection. In the third stage of the project, we examined the factor structure underlying WRCDS items by administering the scale to 402 White undergraduates. Because exploratory factor analysis (EFA) is data driven (Van Prooijen & Van Der Kloot, 2001), the EFA method was used to make appropriate modifications as well as to reduce the number of items that were already developed based on Helms's (1984) racial identity theory. In the fourth stage, the revised scale's final version (WRCDS-R) was reevaluated using a confirmatory factor analysis (CFA) and reliability testing with new data for 168 counselors. A CFA was used to assess the revised model. Finally, to provide additional evidence of score validity, we examined whether the WRCDS-R could differentiate the counselor group from the undergraduate group.
STAGE 1: ITEM GENERATION
Consistent with the statuses proposed by Helms's (1984, 1990, 1995) model of White racial identity, a large pool of items was generated covering the five proposed dimensions of White individuals' attitudes toward Black individuals. Because of the multidimensional nature of the theory (Helms, 1984, 1990, 1995), we developed items using several qualitative data collection sources, such as a focus group discussion and individual interviews.
In qualitative inquiry, sampling occurs purposively: The researcher identifies participants who appear to be rich sources of data (Glesne, 1999; Kuzel, 1999). The participants for the focus groups (n = 19) and for the individual interviews (n = 6) were selected through criterion (all White), critical case (people...
|
