<%BANNER%>

Multicultural Competence and Decision-Making Bias in Child Protective Services


PAGE 1

1 MULTICULTURAL COMPETENCE AND DE CISION-MAKING BIAS IN CHILD PROTECTIVE SERVICES By MICHAEL N. GHALI A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2006

PAGE 2

2 Copyright 2006 by Michael N. Ghali

PAGE 3

3 TABLE OF CONTENTS page LIST OF TABLES................................................................................................................. ..........4 ABSTRACT....................................................................................................................... ..............5 CHAPTER 1 INTRODUCTION................................................................................................................... .6 Overrepresentation of African-American Child ren in Child Welfare: A Closer Look............7 A Need for Cultural Competence...........................................................................................10 Multicultural Competence......................................................................................................11 Multicultural Competence and Child Welfare........................................................................14 2 METHODS........................................................................................................................ .....17 Participants................................................................................................................... ..........17 Instruments.................................................................................................................... .........18 Demographic Questionnaire............................................................................................18 Case Vignettes.................................................................................................................18 Appraisal Rating Questions.............................................................................................20 Multicultural Counseling Inventory................................................................................20 Marlowe-Crowne Social Desirability Scale....................................................................22 Procedure...................................................................................................................... ..........22 3 RESULTS........................................................................................................................ .......24 4 DISCUSSION..................................................................................................................... ....32 Current Limitations............................................................................................................ .....33 Recommendations for Future Research..................................................................................36 Implications................................................................................................................... .........38 APPENDIX..FORMS AND MEASURES................................................................................38 Informed Consent............................................................................................................... ....39 Vignettes: Form A............................................................................................................. ....41 Vignettes: Form B............................................................................................................. .....43 Appraisal Rating Questions....................................................................................................45 Multicultural Counselin g Inventory (MCI)............................................................................47 The Marlowe-Crowne Social Desirability Scale....................................................................50 Debriefing..................................................................................................................... ..........52 Feedback Form.................................................................................................................. .....54 REFERENCES..................................................................................................................... .........55 BIOGRAPHICAL SKETCH.........................................................................................................59

PAGE 4

4 LIST OF TABLES Table Page 3-1 Means, standard deviations, and correlations among va riables (N = 220)........................29 3-2 Results of MANOVA with combined rating indices as DVs and version (race of the child depicted) as IV..........................................................................................................30 3-3 Summary of hierarchical regression analysis of apprai sal rating indices on multicultural competence...................................................................................................30 3-4 Summary of hierarchical regression analyses of multicultural competence on african american and european american appraisal rating indices and Z-st atistics of race of the child as a moderator.....................................................................................................31

PAGE 5

5 Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science MULTICULTURAL COMPETENCE AND DE CISION-MAKING BIAS IN CHILD PROTECTIVE SERVICES By Michael N. Ghali December 2006 Chair: Mark Fondacaro Major Department: Psychology African-American children are overrepresented in the American Child Welfare system. Research has concluded that th is overrepresentation is not due to higher rates of abuse and neglect in African-American families, but more likely due to differential treatment of AfricanAmericans somewhere along the process of reporti ng, investigation, and service allocation. This study serves as an initial examination into th e relationships between multicultural competence, race, and decision-making bias in the investigation phase of the child protect ive service process. Participants were asked to read a series of case vignettes and make deci sions regarding 1) the extent to which they believe the information is ev idence of actual child abus e or neglect, 2) level of risk to the child and 3) level of reco mmended intervention, based upon the information presented in the vignettes. Multivariate Analysis of Variance (MANOVA) and regression analyses were used to analyze the data. Results from this sample of data suggest that ratings of perceived risk and recommended intervention do va ry based on the race of the child and that the multicultural competence of the investigators does significantly predict ratings of substantiation agreement, level of perceived risk, and recomm ended intervention. Results and limitations are discussed and recommendations for fu ture research are presented.

PAGE 6

6 CHAPTER 1 INTRODUCTION Minority children and families are, at times, poorly served by available social and human services. This includes counseling, social work, health care, and other services. For example, African-American children are overrepresented in the Child Welfare system in America. This means that African American, or Black, children are represented in the Child Welfare system in higher proportions than they are in the genera l population of children in the United States. Although Blacks account for only 15% of all children in the United States, they account for 25% of substantiated maltreatment victims and comprise 45% of the total number of children in foster care (Chibnall, Dutch, Jones-Ha rden, Brown, Gourdine, 2003). In addition, these minority children spend longer in the chil d welfare system (Roberts, 20 02). These disproportions are evidence of the differential treatment of African-A merican and Black children and families in the child welfare system. Current trends in the general population give mo re cause for alarm. By the year 2050, the U.S. Census Bureau (2004) expects that peopl e who report to be White alone, Not Hispanic will make up only 50.1% of the population. As th e proportion of minoriti es in the general population gets closer to, and will ultimately ove rtake, the majority population, these and other problems in social/human services may compound th emselves if adjustments are not made in the way human service workers are trained, the way soci al services are delivered, and in the way the systems themselves are organized and operated. One model that shows promise in promoting positive changes in training professionals who serve minority groups is a model of Multicu ltural Counseling Competence developed in the field of counseling psychology (Sue & Arredond o, 1992). Perhaps this model can help to explain the problem of African-American over-repre sentation in child welfare by culturally less-

PAGE 7

7 aware, less-knowledgeable, or less-sk illed decisions made by child welfare workers. This paper will outline the problem of African-American ove r-representation in Child Protective Services (CPS), offer rationale for applying multicultural competence concepts to the problem, and report and discuss results from a study which aims to explore how those concep ts are related to or might help explain racial disparities in child protective service delivery. Overrepresentation of African-American Child ren in Child Welfare: A Closer Look African-American families were, and are, receiving child welfare services, often uninvited, at a proportionally greater rate when compared to white families. Roberts (2002) reports that black children make up 17% of the childrens ge neral population yet 42% of the child welfare population. In addition, once black children enter the child welfare system, th ey often experience different paths through the system than majority children. For example, black children end up in foster care more often than white children and are reunified with their natural families at a slower and lower rate (Roberts, 2002). Although the intent of the child welfare system is to make decisions in the best interest of the child, entering the child welf are system, and foster care in particular, is correlated with other negative life expe riences, such as delinquency and entering the juvenile justice system (Taussig, 2002), and can set the child up for a difficult road through life. One typical entr y point into the child we lfare system is through Ch ild Protective Services (CPS). Generally speaking, suspected abuse or negl ect is reported to a child abuse and neglect hotline. When the report is made, the hotline operator gathers demographic information regarding the family in question and informati on about the abuse or neglect and how it was identified, as reported by the often anonymous ca ller. The hotline operator then decides whether or not to accept the repor t as valid. If the report is accepted, the reported information is sent to Protective Investigation.

PAGE 8

8 A Protective Investigator (PI) is sent to investigate the report The PI has only a few hours to gather any information on which to make two crucial decisions. First, the PI must decide whether or not the abuse/neglect actually happened. If it is determined that the abuse/neglect has indeed occurred, the report is said to be substantiated. If no evidence of abuse/neglect is found, the report is unsubstantiated. Although states have laws to defi ne what is abuse and neglect, these laws can be interpreted differently by diffe rent people and there may be gray areas that remain unclear despite the laws. If evidence of abuse/neglect is found by the PI and the report is substantiated, the PI must then make a deci sion to either, leave the child home and provide services in that context, or remove the child fr om the home due to the im minence of future harm and place the child in a foster care setting. An investigator often makes an assessment of the level of immediate and future risk to the ch ild to help make these decisions (National Association of Child Welf are Administrators, 1999). Once the initial investigation is complete a nd those initial decisions made, the case is transferred to another worker, considered to be Protective Supervision (PS). This worker conducts an ongoing assessment to determine the safe ty of the child and needs to the family. The PS worker also decides what services to o ffer to the family and the manner in which those services will be provided (National Associati on of Child Welfare Administrators, 1999). Clearly, there are many important decisions made during the entire CPS pr ocess and it may be reasonable to think that cultural difference s, and perhaps misconceptions based on those differences, may impact how t hose decisions are made. Before we examine reasons for the overrepres entation of black childre n in the system, we can rule out some other possible explanat ions. One possible explanation for this overrepresentation of black children in the child welfare system is that they actually experience

PAGE 9

9 higher rates of abuse/neglect. One study has inve stigated this claim, but could find no support for it. The Third National Incidence Study of Child Abuse and Neglect (NIS-3) conducted by the Department of Health and Human Services in 1993 found that bl ack children were not abused or neglected at a rate th at differed from that of other r acial groups. The report concluded that differential representati on of minorities in the child we lfare population does not derive from inherent differences in the rates at which they are abused or neglected. If rate of maltreatment is not the cause, perhaps severity of maltreatment is the reason more African American children are re ported to be abused, subsequently investigated, and entered into the system than majority children. Various researchers have examined this factor, but unfortunately have been unable to arrive at a consensus. For example, Hampton (1987) found that, in regards to physical abuse, severity of injury adds statistica lly significant power in discriminating white and black fami lies at both low and high SES. In contrast, in the context of a study controlling for more variables, Zuravin and Orme (1994) found that the childs race did not have a significant main effect on severity of abuse. Another possible explanation is that because the child welfare system is designed to address issues relating to poor families, and b ecause black families are disproportionately poor, this is the reason more black children are entered in to the system. However, Roberts (2002) points out that there is good reason to suspect that poverty cannot co mpletely explain the systems racial disparity. Roberts (2002) is refe rring to findings that approximately the same percentage of Latino and Black households earn less than $15,000, the income level most highly associated with child maltreatmen t (NIS-3, 1993), yet Latino children ar e placed in foster care at a rate identical to their proportion of the population.

PAGE 10

10 After setting aside these possi ble explanations, it seems likely that the childs race influences the overrepresentation of African American children in the child welfare system. In fact, prior research has revealed numerous referen ces to race as a contributing factor in disparate decisions in child welfare processes. For ex ample, Lamb (1979) found that of 17 attribute variables entered into an anal ysis of variance with case outcome as the DV, only 4 of those attributes significantly increased the explained vari ance. Race of the child was one of those four variables and contributed signifi cantly to the total explained variance. Eckenrode, Powers, Doris, Munsch, and Bolger (1988) conducted a study in which th ey examined the child welfare records in the state of New York from the year 1985. Their data suggest ed that for physical abuse reports, no background charac teristics other than ethnicity had an effect on substantiation, but this effect was quite strong, accounting alone for 8% of the variance in substantiation (p. 15). More recently, the NIS-3 (1993) report reveals findings that suggest that the different races receive differential attention somewhere during the process of referral, in vestigation, and service allocation. A Need for Cultural Competence Some within the child welfare system are rea lizing the need to diagnose and make needed adjustments in regards to the problem of over-re presentation. In a study authored by Chibnall, Dutch, Jones-Harden, Brown, and Gourdine (2003) the authors conclude d that the need for continuing research in the area of overrepresentation is evident. Researchers, policy-makers, and practit ioners have divergent views on the causes of minority over-representa tion. This phenomenon may be the result of a disproportionate need for services or of systematic racial influences on decision-making at any number of points along the continuum of child welfare services including reporting, investigation, substantiation, and placement. Resear chers have attempted to explore levels of need and to examine how race affects childrens experiences at each of these points, but findings have been inconsistent. Where racial

PAGE 11

11 differences have been found, the reas ons for these differences remain unclear. (pg. 15) The Chibnall et al. (2003) study involved a focus group of agency administrators, field workers, and policy makers in the field of child welf are from across the country. The participants identified racial bias as a common problem that frequently interfered with good decision making (Executive Summary, p. iii). Participants also mentioned the need for culturally competent and experienced staff as a strategy to improve services to families of color and families in general (p. 34), noting that staff of ten lack exposure to different cultures and had no context for understanding the norms and practices of other cultures. Others have also reported the need for cult ural competence in the child welfare system. Pierce and Pierce (1996) not ed that workers must recognize that their culture determines how they define family, what values are emphasized and even how to greet a person properly (p. 719). Cohen (2003) referenced a specific dimensi on of cultural competence when she referred to attitudes (i.e., awareness) of the child welfare staff as an importan t variable in decision-making. Finally, one recommendation of th e Chibnall et al. (2003) study wa s that future research should examine the effects of cultural competency and cultural sensitivity training on actual child welfare practice. Multicultural Competence Researchers in the field of social work and child welfare have begun to develop models of cultural competence. For example, McPhat ter (1997) presented a Cultural Competence Attainment model. This model consists of three interrelated co mponents: Enlightened Consciousness, Grounded Knowledge Base, and Cumu lative Skill Proficiency. Despite available models, such as the Cultural Competence Attainme nt model, research regarding the assessment of practitioner and student levels of cultural competence and how these levels affect practice,

PAGE 12

12 interventions, and outcomes in ch ild welfare has been lacking. Some believe that the reason for this lack of documented research is due to i nsufficient development of multicultural competence practice measures (Walters & Wheeler, 1998). It seems that until valid measures of cultural competence have been developed in the field of social work, cultural competence research based on existing models in the fi eld will be stalled. Perhaps the most well-developed and empi rically investigated model of cultural competence exists in the field of counseling ps ychology. In 1980, Division 17 (Counseling) of the APA put their Educational and Training Committee to work on developing minimal crosscultural counseling competencies for training pr ograms to incorporate into their training and curricula (Sue et al., 1982). The result was the first formally documented characteristics of the culturally skilled counselor posited in Sue et al.s position paper. This model has been developed over the past two decades and has come to be known as a model of Multicultural Counseling Competencies. The model is built on three main dimensions. The first dimension, Awareness, refers to the process of becoming aware of ones own assumptions about human behavior, values, biases, preconceived notions, personal limitations, and so forth. The second dimension, Knowledge, refers to attempts to und erstand the worldview of culturally different clients without negative judgment. The third di mension, Skills, refers to actively developing and practicing appropriate, relevant, and sensitive inte rvention strategies and skills in working with culturally different clients. These three dimensi ons are conceptually very similar to the three dimensions contained in the Cultural Compet ence Attainment model described above. Considering the findings, assertions, and recommenda tions discussed in the literature, it seems that the dimensions of awareness, knowledge, a nd skills might play a role in how decisions are made in the provision of child welfar e and child protective services.

PAGE 13

13 In fact, some research has already found a connection between these dimensions and differential decisions when race is involved. A study of Chinese, Hispanic, and White parents found that the three cultural groups demonstrated differences in determinations of both the presence and severity of abuse and neglect (Hong and Hong, 1991). Specifically, the Chinese raters in their study tended to judge parental conduct less ha rshly and recommended agency intervention less frequently than Hispanics and wh ites (pg. 5). The authors discuss the findings in terms of cultural explanations and note that cu ltural sensitivity is called for when subjects reactions and response patterns can be attributed to characteristics of thei r culture. Cohen (2003) notes that child welfare worker s ultimately make decisions on family functioning, parent-child relationships, risks, and child safety based more on their perceptions, attitudes, and judgments than on cultural factors that influence child rear ing (p. 149). A possible implication of this statement is that child welfare workers do not possess enough cultural awareness, knowledge, and skill to make decisions based on cultural fact ors. As a consequence, they rely on their perceptions, attitudes, and judgments rather than cultural factors to make decisions. One advantage of the Sue et al. model of multic ultural competence is that various measures of multicultural competence have been developed from this model and used in related research. The most intensive effort to develop such tool s occurred in the early 1990 s (DAndrea, Daniels, & Heck, 1991; LaFromboise, Coleman, & Herna ndez, 1991; Ponterotto & Casas, 1991; Sodowsky, Taffe, Gutkin, & Wise, 1994). The resulti ng measures were all initially created from the three dimensions in the Sue et al. model: Awareness, Knowledge, and Skills, and are organized into two (Awareness & Knowledge/Ski lls), three (Awareness, Knowledge & Skills), or four (Awareness, Knowledge, Skills, & Relations hip) factors based on the items contained in the measure. Skill s

PAGE 14

14 Multicultural Competence and Child Welfare Conceptual links can be made between decisions at different stages of the CPS process and specific items on these measures. For example, because PIs do not usually have more than essentially a few of hours to make such influe ntial decisions, initial impressions could be extremely influential in how situations are perc eived and, subsequently, ho w decisions are made. In this type of scenario, it may be important that the PI be aware of differential interpretations of nonverbal communication (e.g., personal space, eye contact, handshakes) within various racial/ethnic groups (Ponterotto, Gretchen, Utsey, Rieger, & Au stin, 2002, item from MCKAS). Again, as a worker it may be crucial to your de cisions to have an understanding of how your cultural background has influenced the way you th ink and act (DAndrea et al., 1991, from an item on the MAKSS). Davidson-Arad (2001) found empirical evidence that there is a sig nificant relationship between the decision to remove children at risk from home . and assess ments of their parents as uncooperative . . (pg. 127) Parental cooperation was measured by three questions regarding parents cooperation with the child protection work er, motivation for change, and acceptance of responsibility for the problem. (pg 132). This finding provides grounds for another conceptual link between workers deci sions and items on measures of multicultural competence in that it may be useful for the child protection worker to have a working understanding of certain cultures or be fam iliar with nonstandard English (Sodowsky et al., 1994, from items on the MCI) when working with minority families. Roberts (2002) noted a discrepancy between the types of services usually offered to white clients as compared to black clients. For exam ple, white clients are more often referred for housing assistance and black clients are more ofte n referred for mental health services and parenting classes. When a worker decides, fo r example, which services to provide for the

PAGE 15

15 family, hopefully that worker is confident in hi s/her ability to effec tively assess the mental health needs of a person from a cultural backgrou nd different from [their ] own (DAndrea et al., 1991, from an item on the MAKSS) or is knowledgeab le of the relative importance of family, community, and faith based on the clients culture and variations within the clients culture. Considering the links, both empirical and concep tual, between racial di sparities in the child welfare system and multicultural competence, re search directly examining the influence of multicultural competence in racially disparate decisi ons is essential. Previous research has not included multicultural competence and child welfar e decision making as variables in the same experimental paradigm; the propos ed research aims to do so. Specifically, this author is interested in exploring the follo wing research questions: Are deci sions made in the context of child protective services different dependi ng the race of the child ? Does multicultural competence relate to decisions made in the child protective service context? Does race of the child moderate the relationship between multicultural competence and decisions in the context of child protective services? Exploratory in nature, this study will examine the rela tionship between multicultural competence on the one hand and child-related in vestigative decisions (i.e., substantiation agreement, assessed level of risk, recommended level of intervention) on the other. Based on empirical evidence from field data linking race to differential decisions in the child protective service context (Eckenrode et al., 1988; Lamb, 1979; Pierce & Pierce, 1996) and anecdotal evidence from previous research citing the abilit y of case vignettes to e licit those differential decisions (Roberts, 2002; Shapira & Benbenishty, 1993) the aim of th e current study is to test the following hypotheses:

PAGE 16

16 HYPOTHESIS 1: Mean appraisal ratings of substa ntiation agreement, level of risk, and recommended intervention will be higher when case vignettes depict African American children compared to when vignett es depict Caucasian children. A between-subjects regression model will be the basis of testing the second and third hypotheses. Drawing on recent work by Bellini (2003) showin g a significant effect of multicultural competence on client outcomes, a second hypothesis was developed: HYPOTHESIS 2: After controlling for social desi rability, multicultural competence will account for unique variance in appr aisal ratings of substantiati on agreement, level of risk, and recommended intervention, regardless of the race of the child depicted in the vignette. Multicultural competence of the counselor has b een shown to relate to vocational outcomes when the client is African American, but the rela tionship is weaker when the client is Caucasian (Bellini, 2003). Similarly, we expect multicultu ral competence of the caseworker to relate to ratings of substantiation, level of risk, and r ecommended intervention when the child is African American, but we expect the relationship to be weaker when the child is Caucasian. HYPOTHESIS 3: Race of the child depicted in the vignette will moderate the relationship between multicultural competence and ratings of substa ntiation agreement, level of risk, and recommended interventi on, such that multicu ltural competence will account for a greater amount of unique variance in ratings (i.e., substantiation agreement, level of risk, recommended intervention) wh en examining ratings of vignettes depicting African American children a nd will account for less variance when examining ratings of vignettes depicting Ca ucasian children.

PAGE 17

17 CHAPTER 2 METHODS This study was conducted via the internet. Pa rticipation was voluntary and participants were not be reimbursed for their participation in the study. An online survey, including an informed consent process and debriefing was poste d on a secure server and two separate methods of participant recruitment were used; the undergraduate participant pool at a large Southeastern university was utilized, and graduate students of social work as well as caseworkers were accessed through emails requesting participation. Detailed procedures are described below. Following submission of their responses, particip ants were provided links to various resources related to multicultural competen ce in social work and were also given the opportunity to voluntarily complete a feedback form regarding their experience as a participant in the study. Criteria for inclusion in the study included age (18 & over), current en rollment in a human service-related program of study (e.g., sociol ogy, children and family services, psychology, social work) or current employment in a child welfare agency that conducts child protective investigations, and completion of the entire survey. Participants Combined methods of particip ant recruitment resulted in 266 total survey response submissions. Of these, 45 were not complete an d were dropped from all data analyses. The resulting final sample (N=221) was comprised of 157 (71%) undergraduate students, 25 (11.5%) graduate students, 23 (10.5%) caseworkers, a nd 15 (7%) who marked other, but whose reported occupations directly or indirectly invo lve child welfare educatio n or service provision. Females made up the majority of the sample; 164 (74%) compared to 57 (26%) males. The ethnic breakdown of the sample wa s also uneven, with Caucasians making up the majority with 162 (73.5%), compared to 15 (7%) African-Americans, 11 (5%) Asian-Americans, 18 (8%)

PAGE 18

18 Hispanic-Americans, and 11 (5%) participants who marked other. The distribution of degree/job categories is of par ticular consideration. The high percentage of undergraduate psychology students is of viable use because it is reported that caseworkers who are most likely to retain their positions are those with underg raduate degrees (Bernotavicz, n.d.) and less than 30% of child welfare workers possess degree s specific to social work (CWLA, 1999). Instruments The instruments used in this study included a demographic questionnaire a series of case vignettes describing a potential case of abuse or neglect, three apprai sal rating questions, the Multicultural Counseling Inventory (MCI) (S odowsky, Taffe, Gutkin, and Wise, 1994), and the Marlowe-Crown Social Desirability S cale MCSDS (Crowne & Marlowe, 1960). Demographic Questionnaire The demographic questionnaire included basic demographic items as well as questions related to child welfare work e xperience. Answers to these questions were used in follow-up post hoc analyses. Participants were not aske d for any identifying information (e.g., name, date of birth). Case Vignettes Case vignettes were used in this study to give respondents a basis on which to make decisions (i.e., ratings) regarding probability of substantiation, le vels of risk, and levels of recommended intervention they perceived based on the information presented in the vignette. Using case vignettes to have participants make decisions about the level of risk and agency intervention required has been established as a method of examining differential decisions based on race (Shapira & Benbenishty, 1993; Giovanno ni & Bacerra, 1979; Roberts, 2002). The vignettes used in this study were bo rrowed from Giovannoni a nd Bacerra (1979) and adapted to suit the purposes of the study. A total of 20 vignettes were utilized in the current

PAGE 19

19 study. Four of the vignettes we re adapted with both an Af rican American version and a European American version. The two versions were identical except for race of the child depicted in the vignette. These vignettes operated as between-subjects items in order to examine if differential decisions were made based on race. The remaining 16 vignettes included 4 African American, 4 European American, 4 Hisp anic American and 4 Asian American race designations. These 16 vignettes were rated by a ll participants and were combined to create various indices. Every effort was made to ensure that vignettes assigned to various races during the adaptations for the current study were rated at equivalent levels of severity based upon results of the Giovannoni and Bacerra (1979) study. An example of a vignette to be presented to pa rticipants was; The parents of an African American child ignored their childs complaint of an earache and chr onic ear drainage. The child was found to have a serious in fection and damage to the inner ear. (A complete list of the vignettes to be used can be found in the appe ndix to this paper) Limited information was provided in the vignette; surely not enough information to make a well-informed decision in actual practice. However, as this study constituted an initial step in this line of research and was presented to participants as An examination of decisions made in the context of child protective services based on limited information, using simp le vignettes such as these is justified. In addition, to deter participants from becoming sk eptical, they were presented with the following statement before readi ng any of the vignettes: The following statements were taken from real child abuse or neglect report summaries. Following each statement, you will be asked to respond to three items regarding your initial appraisal of the situation. Please read each statement carefully and respond to each item to the best of your professional ability based on the limited information provided.

PAGE 20

20 Regardless of the justifications, participants offered feedback such as, It is hard to use just a one or two sentence story to see what responses would be. We include a whole lot more information that may sway what we do. And, Sometimes, I felt that there wasn't enough case information provided regarding wh at decisions to make. Appraisal Rating Questions Following each vignette, participants were as ked to provide responses to three items related to the vignette. The first item was relate d to substantiation and asked, To what extent to do agree that the situation desc ribed above provides evidence of actual abuse or neglect? Responses were measured on a 4point scale (ranging from 1--Strongly Disagree to 4--Strongly Agree) that participants c hose using a drop-down box on the su rvey. The second item asked participants to rate th e degree to which you believ e this child to be at risk. Responses were measured on a 7-point scale (ranging from 1No ri sk to 7Extreme risk). The third item asked participants to choose a recommended level of in tervention for each case from 6 options (ranging from 1-No further intervention to 6-Removal of the child from the home.) The purpose of using different ranges on the rating scales was so that participants woul dnt simply mark the same rating for all three items for the sake of consistency. Multicultural Counseling Inventory The MCI is intended to measure the compet encies of any counselor working with a minority or culturally different client (Sowdows ky et al., 1994, p. 140). The MCI consists of 40 self-report items that ask respondent s to rate their level of agreemen t, on a 4-point scale (1 very inaccurate to 4 very accurate) with statements related to wo rking with minority clients. Restrictions on the use of the MCI in research prohibited the current researchers from adapting any of the items to reflect language relevant to so cial work. Instead, part icipants were presented with the following instructions precedi ng the standard MCI instructions:

PAGE 21

21 Now, read the following directions a nd statements and apply them to you as a social worker, caseworker, or st udent in a social work or social services training program. No feedback was offered by participan ts about this aspect of the study. In addition to initial tests by the developers of the MCI, this measure has been investigated numerous times to establish validity, reliabilit y, and factor structure (Constantine and Ladany, 2000; Constantine, Gloria, and Ladany, 2002; Wo rthington et al., 2000, Pope-Davis & Dings, 1995). Constantine, Gloria, and Ladany (2002) re ported evidence of the MC Is content validity derived through the accuracy of e xpert raters classification of items into their appropriate subscale categories and expert rati ngs of item clarity. They also provided evidence of criterionrelated validity (i.e., individuals with multicultu ral training or more experience working with culturally diverse populations obtaine d higher MCI scores). Adequate construct validity has also been established through exploratory factor anal ysis and confirmatory factor analysis (PopeDavis & Dings, 1994; Sodowsky, 1998; Sodowsky et al., 1994). High to moderate reliability coefficients (e.g., Cronbachs alphas = .81 fo r the Skills subscale, .80 for the Awareness subscale, .67 for the Relationship subscale,.80 fo r the Knowledge subscale, and .86 for the full scale) have also been reported for the MCI (Pope-Davis & Dings, 1994; Sodowsky, 1998; Sodowsky et al., 1994). Although the MCI was developed directly fr om the Sue et al. model of cultural competence, the authors found four factors to be the best fit for the items. Aside from the three domains of the Sue et al. model, a Relati onship subscale is included on the MCI. The Relationship subscale refers to the counselors in teractional processes wi th the minority client, including stereotypes of the mi nority client and worldview.

PAGE 22

22 Marlowe-Crowne Social Desirability Scale A majority of the literature regarding the development and utilization of measures of multicultural competence discuss the relationship of social desirability to such measures. The author of the MCI specifically recommends th at the completion of multicultural competence measures be accompanied by an index of so cial desirability (Sodowsky, 1994). Social desirability is typically operat ionalized using the Marlowe-Crow ne Social Desirability Scale (MCSDS) in psychological litera ture (Worthington, Mobley, & Fr anks, 2000). The MCSDS is a 33 item true/false scale designed to measure the degree to which rese arch participants attempt to make a good impression when comp leting research instruments (W orthington et al., 2000). The scale was originally designed to be used with m easures of personality, but has since been used numerous times in connection with measures of other constructs, including multicultural competence. Internal consistency of the MCSDS, calculated using the Kuder-Richardson formula 20, was found to be .88. The authors report a test-retest correlati on of .89 for the scale (Crowne & Marlowe, 1960). Procedure When participants clicked the link to the study website, they were presented with the informed consent form (see appendix A). Particip ants were informed that clicking the link at the bottom of the page to begin the study would be accepted as their consent to participate. They were randomly assigned to either form A or form B when they clicked the Take me to the study button by means of a randomization proce ss built into the online survey. Form A included the European American versions of the 4 between-subjects vign ettes (Items 1, 6, 11, & 16). Form B included the African American vers ion of the 4 between-subj ects vignettes (Items 1, 6, 11, & 16). No other differences exist between Form A and Form B.

PAGE 23

23 Participants were asked to complete a demographic questionnaire which included questions about their gender, et hnicity, current occupation, year s of experience working with children and whether or not they are curren tly involved in making decisions about child abuse/neglect. Before being presented with the first vignette, participants were instructed as follows: The following statements were taken from real child abuse or neglect report summaries. Following each statement, you will be asked to respond to three items regarding your initial appraisal of the situation. Please read each statement carefully and respond to each item to the best of your professional ability based on the limited information provided. For each vignette, participants were asked to re ad the information and complete the appraisal rating questions. One vignette and the three apprai sal rating questions appe ared together on each page (i.e., screen). (Both versions of the form included in Appendix A). Once participants read all vigne ttes and recorded their ratings, they were asked to complete the MCI and the Marlowe-Crown Social Desirabi lity Scale. Upon comple tion of all measures, participants were presented with a debriefing form informing them of the complete purpose of the study (see appendix A for debriefi ng form). At that point, part icipants were asked to indicate (by way of a button on the webpage) that they had read and understood the debriefing. They were also given the opportunity to indicate if they did not want thei r data to be used in the final analyses. At that point, participants c ould submit their responses. Clicking submit simultaneously uploaded their responses to a databa se and navigated participants to a page that thanked them for their participation, presented li nks to numerous online re sources and a feedback form asking them about their experience as a pa rticipant in the study a nd offering a chance to include any written feedback they wished. Res ponses to the feedback form were not linked in any way to responses from the survey.

PAGE 24

24 CHAPTER 3 RESULTS Before any analyses were run, ratings from the vignettes were summed to form specific indices. Specifically, ratings on the question regarding substan tiation agreement from the four between subjects vignettes (I tems 1, 6, 11, & 16) were summed to form the Combined Substantiation Index (Cronbachs alpha = .43, Sp earman-Brown = .42). Similarly, a Combined Risk Index (Cronbachs alpha = .52, Spearman-Br own = .55) and Combined Intervention Index (Cronbachs alpha = .43, Spearman-Brown = .49) wa s formed. All subsequent data exploration and analyses were conducted using SPSS 11.5. All relevant va riables were examined for normality by examining histograms and skewness and kurtosis statistics. Other than evidence of mild kurtosis (zkurtosis = 2.22) in the distribution of the Co mbined Substantiation Index, all other variables appeared to be normally distributed. Means, standard deviations, and correlation coefficients for all relevant variables are provided in Table 3-1 (below). The mean score on the MCI pr ovided by the graduate students and caseworkers ( M = 121.65, SD = 11.66 ) is roughly equivalent to the mean scores on the MCI provided in previous literature for gra duate students and other professionals ( M = 122.73, SD = 12.09 ) (Worthington, Mobley, Franks, & Tan, 2000; al so see Bellini, 2003). The mean score on the MCI for undergraduates ( M = 113.55, SD = 11.28 ) was significantly lower (t (219) = 4.70, p < .001) than the mean score for graduate students and caseworkers (no mean scores on the MCI Total Scale for undergraduates could be located in the literature). This is to be expected considering the lack of multicul tural coursework and training provided at the undergraduate level. In order to test Hypothesis 1, variables were entered into a MANO VA to test for main effects of race of the child (i.e., European Amer ican or African American) on all three appraisal

PAGE 25

25 ratings. A MANOVA was chosen because th e dependent variables (i.e., Combined Substantiation, Combined Risk, and Combined Intervention) are inte rrelated (Pearson Correlations of .524, .634, and .657, p < .001). Levenes Test for equality of variances was nonsignificant ( p > .05) for all DVs suggesting that the main effects can be interpreted with confidence. Table 3-2 (below) provides the results of th e MANOVA showing a main effect for race of the child on ratings of Risk (F (1,218) = 4.55, p < .05 ) and on ratings of Intervention (F (1,218) = 3.31, p < .05) This suggests that participants percei ved higher levels of risk and recommended more severe (i.e., intrusive) in terventions when the child depict ed in the vignette was African American (mean rating of risk = 5.74; mean ra ting of intervention = 4.44) than when the child was European American (mean rating of risk = 5.54 ; mean rating of intervention = 4.26). The main effect of race of the child on ra tings of Substantiation (F (1,218) = .930, p = .168) was nonsignificant. (Mean African American rating of s ubstantiation = 3.51; Mean European American rating of substantiation = 3.47) These results pr ovide partial confirmation for Hypothesis 1 and lend credence to the experimental manipulation of race in case vignettes as a useful tool in examining decision-making bias. In order to test Hypothesis 2, three separa te hierarchical regr ession analyses were conducted. The three combined rating indices (Substantiation, Risk, and Intervention) were entered as dependent variables and social desira bility (operationalized as scores on the MCSDS) and multicultural competence (operationalized as the MCI total-scale score) were entered stepwise as independent variables. Table 33 (below) provides the re sults of each of the appraisal rating indices regressed upon multicultural competence (results of rating indices regressed upon social desirabil ity not shown). After controlling for social desirability,

PAGE 26

26 multicultural competence was a significant predictor of mean ratings of substantiation (t(219) = 2.43, p = .016 b = .019 ), mean ratings of risk (t(219) = 2.08, p = .039, b = .034 ) and mean ratings of intervention (t(219) = 2.79, p = .006, b = .047). These results fully support Hypothesis 2. Specifically, the results suggest that higher levels of multicultural competence significantly predict higher ratings on each of the rating i ndices. The regression models explain small amounts of variance however, (r2 = .033, r2 = .020, and r2 = .036 respective ly) according to Cohens standards for effect size s in regression analyses ; although virtually all of the variance in the model is accounted for by multicultural competence (r2 = .026, r2 = .020, and r2 = .035 respectively). If the possibility of family-w ise error by running three regres sion analyses is taken into account, the Bonferroni adjusted p-va lue for significance would become p .016. Thus, the relationship between multicultura l competence and Substantiation would still be significant ( p = .016) as would be the relationship between multicul tural competence and ratings of Intervention ( p = .006). However, the relationship between multi cultural competence and ratings of Risk would no longer be significant ( p = .039). The final stage in these analyses was to check assumptions of the model(s). Collinearity statistics, based on tolerance and VIF suggest that the assumption of collinearity is not violated in any of the models. Durban-Watson was used to check whether residuals in the model are independent. Field (2000) suggests that Durban-Wat son values less than 1 or greater than 3 are cause for concern. Durban-Watson values for th e regression models used to test hypothesis 2 were all between 1.84 and 1.90 suggesting that re siduals in each model are independent. Scatterplots of residuals were examined for possible heteroscedastici ty and linearity. No

PAGE 27

27 indication of heteroscedasticity or non-linearity was evid ent in any of the models suggesting that the assumptions of homoscedasticity and linea rity in the data have been met. In order to test Hypothesis 3 regarding mode rating effects, methods reported by Baron and Kenny (1986) were utilized. They suggest running regression analyses a nd using the resulting bweights rather than correlation coefficients to test for differences because regression coefficients are not affected by differences in the variances of the independent variable or differences in measurement error of the dependent variable. (p. 1175). Therefore, two separate hierarchical regression analyses were run for each appraisal rating (i.e., substantiation agreement, level of perceived risk, and recommended intervention. In the first analysis of each rating, the mean score on the African American (Substantiation) index was the dependent variab le and social desirability and multicultural competence were entered hierarchically as predic tors. The resulting b-weight of multicultural competence on the African American Substantia tion index was b = .016. The second analysis included the European American (Substantiation) index as the dependent variable and social desirability and multicultural competence as predic tors. The resulting b-weight of multicultural competence on the European American index was b = .021. A z test was performed to test for a difference between the two b-weights of the indepe ndent variable (MCI full-scale score). A significant difference in the b-weights would be ev idence of the moderating effects of race of the child presented in the vignette on multicultural co mpetence as a predictor of appraisal ratings. As recommended by Paternoster, Brame, Mazero lle, and Piquero (1998) a more conservative denominator was used in the equation for the z statistic, thereby reducing the chances of rejecting the null when there truly is no difference between the two b-weights (see equation below).

PAGE 28

28 Table 3-4 (below) shows the unstandardized co efficients resulting from the hierarchical regression analyses of the three appraisal ratings on multicultural competence as well as the zstatistic for the test of modera tion of race of the child upon the relationship between multicultural competence and appraisal ratings. For the test of the moderation effects of race of the child on the relationship between multicultural competence and ratings of substantiation agreement, the resulting z-statistic (z = .33) is not significant. These findi ngs do not provide support for the hypothesis that the race of the child modera tes the relationship between multicultural competence and levels of substant iation agreement. This pro cedure was repeated to examine ratings regarding level of risk and recommended intervention. The resulting z-statistics for level of perceived risk (z = .09) and recommended inte rvention (z = .30) were non-significant as well, suggesting that race of the child depicted in the vignette did not moderate the relationship between multicultural comp etence and appraisal ratings. In addition, when the rating indices were regressed upon multicultural competence sepa rately for the European American version and the African American version, multicultura l competence was found to only be a significant predictor of ratings of Interv ention when a European American child was depicted in the vignette. The lack of a significant relationship be tween multicultural competence and the racespecific rating indices, and subsequent lack of support for hypothesis 3 could be due to the smaller sample size garnered in this study. Our in itial thought that future studies utilizing larger sample sizes may be successful in finding suppo rt for the moderation effect described in hypothesis 3. Using G*Power, the effect sizes fo r the African American and European American Intervention indices were entered to estimate the sample size needed to find a significant relationship in a regression model with two pred ictors at a power of 0.80. The results suggest

PAGE 29

29 that a sample size of 480 (total) participants should provide the power needed to find a significant relationship between multicultural co mpetence and both the African American and European American Intervention indices. Howe ver, when the product term for the interaction effect of multicultural competen ce and race of the child is ente red into the regression equation, the resulting (orthogonalized) eff ect size of the interaction is extremely small by Cohens standards (r2 added = .002) and G*Power estimates th at nearly 4000 partic ipants would be needed to find a significant moderation effect of race of the child on the relationship between multicultural competence and appraisal ratings of intervention! Thus, it seems that a true moderation effect of race of th e child on the relationship between multicultural competence and ratings of (in this case) Inte rvention is highly unlikely. Table 3-1. Means, standard deviations, and correlations among variables (N = 220) Variable 1 2 3 4 5 1. MCI Total Score .295.132.138* .167* 2. Social Desirability -.080.015 -.037 3. Combined Substantiation Index .657** .524** 4. Combined Risk Index .634** M 115.8716.18 3.49 5.64 4.53 SD 12.10 5.16 .33 .70 .72 p < .05. ** p < .01

PAGE 30

30 Table 3-2. Results of MANOVA with combined ra ting indices as DVs and version (race of the child depicted) as IV Index SSx dfF p SST Combined Substantiation 1.624 1.930 .168 382.450 Combined Risk 35.312 14.548*.017*1727.977 Combined Intervention 27.522 13.312*.035*1838.995 p < .05. Table 3-3. Summary of hierarchical regression analysis of appraisal rating indices on multicultural competence Index B SE t p Combined Substantiation .019 .008.1702.43* .016 Constant 12.338 .850---Combined Risk .034 .016.1462.08* .039 Constant 18.889 1.819---Combined Intervention .047 .017.1952.79**.006 Constant 12.857 1.861---* p < .05. ** p < .01

PAGE 31

31 Table 3-4. Summary of hierarchical regression analyses of multicultural competence on african american and european american appraisal rating indices and Z-statistics of race of the child as a moderator Index B SE B t Df Z African American Substantiation .016.010.1531.544108 European American Substantiation .021.011.1821.832110 Test of Moderation 0.33 African American Risk .031.020.1521.532108 European American Risk .034.026.1301.310110 Test of Moderation 0.09 African American Intervention .040.022.1821.828108 European American Intervention .050.025.1971.982110 Test of Moderation 0.30

PAGE 32

32 CHAPTER 4 DISCUSSION This study was to serve as an initial investigat ion into the relationships between race of the child, multicultural competence, and decision-ma king bias in the child protective services context. Previous successful use of short case vign ettes are presented in the literature (Shapira & Benbenishty, 1993; Giovanna & Bacerra, 1979; Robert s, 2002), however no evidence exists in the literature of using these short vignettes to dr aw out biases based on race. This study aimed to further develop the use of case vignettes in ch ild welfare research. By having participants randomly assigned to complete either a version with the target vignette s depicting European American children or a version using the identical target item s depicting African American children, differences in mean rati ngs of the vignettes that could be directly attributed to the experimental manipulation of race of the child could be investigated. In the sample of data used in the curren t study, it appears that when making decisions about perceived levels of risk and recommended levels of inte rvention in the child protective context, the race of the child presented in the vi gnette did make a difference. When the child presented in the vignette was African America n, ratings of risk and recommended intervention were higher than when the child presented in th e vignette was European American. This finding provides experimental support to previous inves tigations using field data (Eckenrode, Powers, Doris, Munsch, and Bolger 1988; and Lamb, 1979; NIS-3, 19 93) and qualitative data (Chibnall, Dutch, Jones-Harden, Brown, and Gourdine, 2003) that suggest the racial biases do impact child protective service provision. As an initial investigation into the rela tionship between multicultural competence and decisions made in the context of child protectiv e service investigations, this study offers support for the consideration of multicultural competen ce as a valid construct in examining such

PAGE 33

33 decisions. It is unclear what aspect of the decision-making process multicultural competence is tapping and it is uncertain if the MCI is the best measure to use in such investigations. Answers to these questions may be useful in guiding future research. Although Hypothesis 2 was supported by the resu lts of the regression analyses, neither theory nor past research suggests an expl anation for the positive relationship between multicultural competence and severity of appraisals. Perhaps those with more multicultural competence are more globally aware of the inhere nt dangers presented to a child in any given situation compared to those who are less multicultu rally competent. Or perhaps the distribution of demographic variables in this sample ha d an unexpected impact on the results. Current Limitations Currently there are various limitations to the study. For one, multicultural competence was measured using a self-report instrument. This is a limitation because a self-report measure is designed to capture perceived levels of mu lticultural competence and it is unknown how well perceived multicultural competence reflects how ones level of multicultural competence may impact his/her performance. Unfortunately, self-report measures are the only measures of multicultural competence currently available. Undergraduate students made up the majority (7 1%) of the sample in this study. Initially, this is troublesome on a research design level. Perhaps the variability in scores on the outcome variables is due to ut ilizing a sample that has not had a ny training or experience in making decisions related to child welfare investigations Nonetheless, outcome measures of risk and intervention were still sign ificantly higher, at the p < .05 level, for vignettes depicting African American children than for vignettes depicti ng European American children, even when undergraduate vs. graduate/professional status was c ontrolled for by entering it as a covariate. The same level of support was found for hypothesi s 2 when controlling for undergraduate vs.

PAGE 34

34 graduate/professional status. However, the undergra duate status of the majority of the sample may have an impact on how well the sample represents the populat ion of interest. Undergraduate students in the psychology rese arch pool may or may not be studying psychology or any related social science field. The inclusi on of students who are not social science majors might decrease the probability that the majority of the sample has a ny interest in, or any awareness of, making child welfare-related decisions Thus, the question could be raised as to whether the results from the current sample are representative of the population of people who might be expected to make child welfare-related decisions. The exploratory nature of the study limits both the scope and focus of questions asked. Moreover, the model used to examining ch ild protective investigation outcomes is underspecified. Exploratory analyses on the data may lead to suggestions for a more specified model to be tested in future research. But this is beyond the scope of the current paper. Despite the apparent success of using these case vignettes in this study, the ecological validity of making determinations of actual chil d abuse/neglect, risk, and intervention based upon the case vignettes is open to de bate. Although similar case vign ettes have been utilized in past research (Giovanna & Bacerra, 1979; Sh apira & Benbenishty, 1993), feedback from participants suggested that in many instances, a determination would not be based upon such a limited amount of information. The online nature of the study may pose anot her limitation. Although both the survey and the resulting data were housed on a secure serv er, there were no safeguards in place to ensure that each participant only completed the survey one time. A participant could, if he/she chose, complete the survey numerous times and if su ch a person had a specific agenda in mind, this could alter the results. There is no indication and the re searcher has no reason to believe that this

PAGE 35

35 posed a true limitation of this study. The length of tim e necessary to complete the survey and the lack of compensation for participants make numerous submissions of survey responses unlikely. It is also possible that a participant coul d change his/her answers to any of the appraisal rating questions after he/she became aware of or completed the section measuring multicultural competence. However, due to the between subject s nature of the design, it seems unlikely that a participant would be aw are of which specific items to change. A potential limitation of the data itself may be the presence of outliers. A brief look at scatterplots of the data sugge sts that there are low ratings on each of the indices (the substantiation index in particula r) that may be considered outlie rs and may be having a negative effect on our data analyses. Currently, data from all participants who completed the entire survey were included in our an alyses. A closer examination of the data utilizing casewise diagnostics may offer substantial st atistical reasons to eliminate th ese scores from our analyses. Specifically, a close look at the standardized resi duals of our regression an alyses and utilization of Cooks Distance values can offer informati on regarding the impact these potential outliers may be having on our analyses. A quick look at th e scatterplots suggest that these outliers may have high discrepancy and moderate leverage with which to disrupt our results. A final limitation to be discussed in this pa per is a limitation related to the specific measure of multicultural competen ce used in this study. Although the MCI has been reported to be a valid instrument in previous literature, because the MCI had to be used in its original form and could not be altered to reflect language rele vant to social workers, it is possible that a measure of multicultural competence developed speci fically in the social work or child welfare field, or a measure adapted to re flect language relevant to social workers might be more useful and tap into competencies which could explain mo re variance in our models of analyses.

PAGE 36

36 On a related theoretical level, a question of what is truly being measured by the MCI and MCSDS must be raised. Multicu ltural Competence is a somewhat ambiguous concept and it is therefore difficult to ascertain wh ether the MCI is accurately capt uring aspects of working with people from diverse cultures that are important in making decisions about child abuse/neglect. Other constructs may more accurately reflect culturally-related asp ects of child welfare investigation. For example, Racial Identity is a construct that has received moderate coverage in the literature. Utilizing the concept of racial identity may allow for more flexibility when examining the role of ethnicity in child welfare-re lated decisions. White R acial Identity has been found to clearly be differentially re lated to racism in such as way that different racial identity schema profiles relate to racism in different ways (Carter, Helms, & Juby, 2004). Perhaps examining racial bias in decisi on making could benefit from util izing the construct of racial identity as opposed to mu lticultural competence. The method of controlling for social desira bility in this study is also potentially problematic. Although a standard of measure soci al desirability was used in this study, the appropriateness of the MCSDS is unclear. One might argue that social desirability is adaptive to different generations. Thus, the measurement of so cial desirability must be continually redefined and recreated; which has not been sufficiently il lustrated in the literature. Sodowsky (1996) argued the general construct of social desi rability might not be useful when studying multicultural competence. Perhaps there is a differe nt type of social desirability in relation to multicultural competence. Recommendations for Future Research The results of this study lend promise to the valu e of future research in this area. Utilizing the case vignettes in an experimental design fash ion appeared to be successful. Thus, future

PAGE 37

37 research may benefit from utilizing case vignettes in such a way; particularly to investigate differential decisions based on specific demographi c characteristics presented in the vignettes. It appears that multicultural competence may be a valid construct in examining decisions made in the context of child abuse/neglect servic es provision. However, as noted above, it is recommended that future studies examining this issue utilize either an adapted measure that reflects language relevant to social workers or a measure developed specifically based upon multicultural competencies related to child welfare services. At this time, it does not appear that measures exist that would be in line with this recommendation. Future research may also benefit from utilizing different theoretical constructs in explaining the variance in scores of child welfare investigation outcome measures. Future research in this area may also bene fit from more effective recruitment methods aimed at accessing a larger sample of graduate st udents of social work and current child welfare caseworkers. There appears to be a different pattern of results among these groups based upon initial eyeballing of the appraisal rating means from these groups. Higher numbers of participants could also increase the power of fu ture studies to find si gnificant effects. A final suggestion for future research is a direct result of feedback from participants of this study and other professionals who we re consulted while the study was act ive. It is suggested that racial biases and multicultural competence may play a larger role in the treatment phase of the child welfare process. Some professionals note d that abuse/neglect inve stigation decisions are driven more by state and federal laws than by individual judgments and decisions. Perhaps individual judgments and decisions, and therefore racial bias and multicultural competence, have a greater impact when it comes to treatment and re source allocation decisions than substantiation and risk decisions. This idea was partially suppor ted by the results of this study in considering

PAGE 38

38 that significant impacts of multicultural competence and race were seen more when it came to appraisals of recommended intervention than subs tantiation and risk apprai sals. Future research might benefit from a focus on the treatment and resource allocation phase of the child welfare service process rather than the investigative and substantiation phase of the process. Implications Due to the exploratory nature of this study and the first step this study has taken to investigate the relationships between race of th e child, multicultural competence of the worker, and decisions made in the context of child welf are service provision, a ny implications discussed are tentative and should be further ex plored with future research. One implication of the findings of this study is that racial bias in making decisions about child abuse/neglect appears to be a robust effect. When case vi gnettes presented on a computer screen can elicit differential deci sions based upon the race of the ch ild presented in the vignette, it may be reasonable to think that real-world situ ations in which a worker is confronted with racial differences may have a similar, if not greater impact on his/ her perceptions, thereby leading to bias in the provision of services. Regarding the use of the multicultural competence construct, the results of this study seem to provide the impetus to apply th is construct to future investiga tions of child welfare and other social work processes. Newly created measures incorporating specific competencies relevant to social work are missing from the current literat ure and the results of this study may provide additional support for the viability and usef ulness of creating such measures.

PAGE 39

39 APPENDIX FORMS AND MEASURES Informed Consent Protocol Title: Child Protective D ecisions Based on Limited Information Please read this consent document carefully befo re you decide to participate in this study. Purpose of the research study: The purpose of this study is to examine decision s made in the child protective services context based on limited amounts of information. What you will be asked to do in this study: You will be asked to read a series of twenty st atements from separate case report summaries and make three separate ratings based on the informa tion provided in each statement. Once you have read all twenty statements and provided ratings for each, you will be asked to respond to a number of items based on the extent to which you believe the item is a reflection of you. Some of the items ask you to respond with either true or false and some of the items ask you to respond by rating how accurate the item is as it re lates to you as a professi onal on a 4-point scale. Time required: Approximately 45 minutes Risks and Benefits: There are no anticipated risks to participating in this study. Anticipated benefits are minimal, but you may gain an increased awareness of your own decision making style in instances of limited information. Compensation: You will not be compensated for your participation in this study. Confidentiality: All information provided in this study will be co mpletely anonymous. You will not be asked to provide any identifying information, nor will your location or internet address be tracked while you are participating in the study. No attempts will be made to identify you in any way.

PAGE 40

40 Voluntary participation: Participation in this study is completely volunta ry. There is no penalty for not participating. Right to withdrawal from the study: You have the right to discontinue participation in the study at a ny time, for any reason, without consequence. Whom to contact if you have questions about the study: Michael N. Ghali, Graduate Student, Departme nt of Psychology (Counseling) University of Florida, Gainesville, FL 32611 Whom to contact about your rights as a research participant in the study: UFIRB Office, Box 112250, University of Florida, Gainesville, FL 32611-2250; ph 392-0433. Agreement: By checking the box below, you agree that you ha ve read the procedur e described above and voluntarily agree to participate in the study. If you do not agree, pl ease select the Quit button below and this window in your web browser will be closed. If you do agree to participate, please check the box marked I agree below, print this page for your records, and select the Continue button.

PAGE 41

41 Vignettes: Form A The following statements were taken from ch ild abuse or neglect report summaries. Following each statement, you will be asked to respond to three items regarding your initial appraisal of the situation. Please read each st atement carefully and respond to each item to the best of your professional abilit y based on the information provided. (Bold items will be used as between subjects items) 1. An 8 year old European American girl was brought to the hospital three times for being underweight. Each time the child gai ned weight during the hospital stay. Upon return home, the child lost weight again. 2. A 12 year old Hispanic American boy was stru ck by his parent with a wooden stick. The child suffered a concussion. 3. The parents of an 8 year old European Amer ican boy ignored their childs complaint of an earache and chronic ear drainage. The ch ild was found to have a serious infection and damage to the inner ear. 4. A 12 year old African American girl was regularly left home alone all night. 5. The parents of an Asian American child regular ly fail to feed their 8 year old for periods of at least 24 hours. The young girl was hos pitalized for six weeks for being seriously underweight. 6. The parents of a 12 year old European American girl have repeatedly failed to keep medical appointments for their child. The child has a congenital heart defect. 7. The parents of a 12 year old Asian American boy regularly leave th eir child with their neighbors, without knowing who w ould assume responsibility a nd be in charge. On one occasion, the child was seen at a store four blocks away. 8. The parents of a 12 year old Hispanic Am erican girl have not given their child medication prescribed by a physician. The child has a throat infection. 9. The parents of an 8 year old Hispanic Ameri can boy regularly leave the child alone inside the house after dark. Often, they do not return until midnight. On one occasion, the child started a small fire. 10. A 12 year old European American child is no t given any healthcare by her parents. The child complains of physical ailments. 11. The parent of a 12 year old European Am erican girl hit the child in the face, striking her with the fist. The child suffered a black eye and a cut lip.

PAGE 42

42 12. The parent of a 12 year old African Ameri can boy banged the child against the wall while shaking him by the shoulders. The child suffered large bruises. 13. The parents of an 8 year old Hispanic Amer ican child fail to prepare regular meals for her. The child often has to fix her own supper. 14. The parents usually punish th eir 12 year old European American boy by spanking him with a leather strap, leaving re d marks on the childs skin. 15. On several occasions, an 8 year old African American boy was left with his grandmother for periods of time up to ten days without pr oviding any means of contacting the parents. The child became ill during the most recent instance. 16. An 8 year old European American boy was burned on the chest and buttocks with a cigarette by a parent. The child has second degree burns. 17. A 12 year old Asian American boy is regularly left alone outside during the day until almost dark. Neighbors have spotted the child wandering five bl ocks from home. 18. The parents of an 8 year old European Ameri can girl feed only milk to the child. The child has an iron deficiency. 19. An 8 year old Asian American girl was imme rsed in a tub of hot water. The child suffered second-degree burns. 20. The parents of an 8 year old African American boy regularly leave the child alone inside the house during the day. Often, they did not return home until almost dark. On one occasion, the parents returned home to find broken glass on the floor.

PAGE 43

43 Vignettes: Form B The following statements were taken from ch ild abuse or neglect report summaries. Following each statement, you will be asked to respond to three items regarding your initial appraisal of the situation. Please read each st atement carefully and respond to each item to the best of your professional abilit y based on the information provided. (Bold items will be used as between subjects items) 1. An 8 year old African American girl was brought to the hospit al three times for being underweight. Each time the child gai ned weight during the hospital stay. Upon return home, the child lost weight again. 2. A 12 year old Hispanic American boy was stru ck by his parent with a wooden stick. The child suffered a concussion. 3. The parents of an 8 year old European Amer ican boy ignored their childs complaint of an earache and chronic ear drainage. The ch ild was found to have a serious infection and damage to the inner ear. 4. A 12 year old African American girl was regularly left home alone all night. 5. The parents of an Asian American child regular ly fail to feed their 8 year old for periods of at least 24 hours. The young girl was hos pitalized for six weeks for being seriously underweight. 6. The parents of a 12 year old African Ameri can girl have repeatedly failed to keep medical appointments for their child. The child has a congenital heart defect. 7. The parents of a 12 year old Asian American boy regularly leave th eir child with their neighbors, without knowing who w ould assume responsibility a nd be in charge. On one occasion, the child was seen at a store four blocks away. 8. The parents of a 12 year old Hispanic Am erican girl have not given their child medication prescribed by a physician. The child has a throat infection. 9. The parents of an 8 year old Hispanic Ameri can boy regularly leave the child alone inside the house after dark. Often, they do not return until midnight. On one occasion, the child started a small fire. 10. A 12 year old European American child is no t given any healthcare by her parents. The child complains of physical ailments. 11. The parent of a 12 year old African American girl hit the child in the face, striking her with the fist. The child suff ered a black eye and a cut lip.

PAGE 44

44 12. The parent of a 12 year old African Ameri can boy banged the child against the wall while shaking him by the shoulders. The child suffered large bruises. 13. The parents of an 8 year old Hispanic Amer ican child fail to prepare regular meals for her. The child often has to fix her own supper. 14. The parents usually punish th eir 12 year old European American boy by spanking him with a leather strap, leaving re d marks on the childs skin. 15. On several occasions, an 8 year old African American boy was left with his grandmother for periods of time up to ten days without pr oviding any means of contacting the parents. The child became ill during the most recent instance. 16. An 8 year old African American boy was burned on the chest and buttocks with a cigarette by a parent. The child has second degree burns. 17. A 12 year old Asian American boy is regularly left alone outside during the day until almost dark. Neighbors have spotted the child wandering five bl ocks from home. 18. The parents of an 8 year old European Ameri can girl feed only milk to the child. The child has an iron deficiency. 19. An 8 year old Asian American girl was imme rsed in a tub of hot water. The child suffered second-degree burns. 20. The parents of an 8 year old African American boy regularly leave the child alone inside the house during the day. Often, they did not return home until almost dark. On one occasion, the parents returned home to find broken glass on the floor.

PAGE 45

45 Appraisal Rating Questions To what extent to do agree that the inform ation provided in the preceding case summary is evidence of actual child abuse or neglect? ____ 1. Strongly disagree ____ 2. Disagree ____ 3. Agree ____ 4. Strongly agree In your best professional opini on, based on information given in the preceding case summary, please rate the degree to which you be lieve this child to be at risk: ____1. No risk ____2. Low risk ____3. Some risk ____4. Intermediate level of risk ____5. High risk ____6. Very high risk ____7. Extreme risk In your best professional opini on, based on the information in the preceding case summary, please choose ONE of the following options for intervention below: ____1. No further intervention. ____2. Indirect interven tion through professiona ls already in contact with the family. ____3. Direct intervention with this family without removi ng the child from the home. ____4. Direct intervention with the family, place the child in day placement until the evening.

PAGE 46

46 ____5. Removal of the child from the home for an extended period of time, with the parents agreement, while continuing to work with the family. ____6. Removal of the child from the home for an extended period of time, even without the parents agreement (with court order, if necessary ), while continuing to work with the family.

PAGE 47

47 Multicultural Counseling Inventory (MCI) The following statements cover counselor practic es in multicultural counseling. Indicate how accurately each statement describes you as a coun selor, psychologist, or student in a mental health training program when working in a multicu ltural counseling situation. Give ratings that you actually believe to be true rather than those you wish were true. The scale ranges from 1 (very inaccu rate) to 4 (very accurate). Th e scale indicates the following: 1 very inaccurate 2 somewhat inaccurate 3 somewhat accurate 4 very accurate When working with minority clients 1. I perceive that my race causes the clients to mistrust me. 2. I have feelings of overcompe nsation, oversolicitation, and gu ilt that I do not have when working with majority clients. 3. I am confident that my conceptualizati on of client problems does not consist of stereotypes and value-oriented biases. 4. I find that differences between my worldvi ews and those of the clients impede the counseling process. 5. I have difficulties communicating with clie nts who use a perceptual, reasoning, or decision-making style that is different from mine. 6. I include the facts of age, gender roles, and socioeconomic status in my understanding of different minority cultures. 7. I use innovating concepts and treatment methods. 8. I manifest an outlook on life that is best described as world-minded or pluralistic. 9. I examine my own cultural biases. 10. I tend to compare client behaviors w ith those of majority group members. 11. I keep in mind research findings about minor ity clients preferences in counseling.

PAGE 48

48 12. I know what are the changing prac tices, views, and interests of people at the present time. 13. I consider the range of behaviors, values, and individual differences within a minority group. 14. I make referrals or seek consultations based on the clients minority identity development. 15. I feel my confidence is shaken by the se lf-examination of my personal limitations. 16. I monitor and correct my defensiveness (e. g., anxiety, denial, anger, fear, minimizing, overconfidence). 17. I apply the sociopolitical history of the clie nts respective minority groups to understand them better. 18. I am successful at seeing 50% of the clie nts more than once, not including intake. 19. I experience discomfort because of the clie nts different physical appearance, color, dress, or socioeconomic status. 20. I am able to quickly recognize and recover fr om cultural mistakes or misunderstandings. 21. I use several methods of assessment (includi ng free response questions, observations, and varied sources of information and excluding standardized tests). 22. I have experience at solving problems in unfamiliar settings. 23. I learn about clients different ways of acculturation to the dominant society to understand the clients better. 24. I understand my own philo sophical preferences. 25. I have a working understanding of certain cult ures (including African American, Native American, Hispanic, Asian American, new Th ird World immigrants, and international students. 26. I am able to distinguish between those w ho need brief, problem-solving, structured therapy and those who need long-term, pr ocess-oriented, unstructured therapy. 27. When working with international students or immigrants, I unders tand the importance of the legalities of visa, passport, green card, and naturalization. Evaluate the degree to which the following mult icultural statements can be applied to you. 28. My professional or collegial interactions with minority individuals are extensive.

PAGE 49

49 29. In the past year, I have had a 50% increase in my multicultural case load. 30. I enjoy multicultural interactions as much as interactions with people of my own color. 31. I am involved in advocacy efforts against institu tional barriers in ment al health services for minority clients (e. g., lack of bilingua l staff, multiculturally skilled counselors, minority professional leadership, and outpatient counseling facilities). 32. I am familiar with nonstandard English. 33. My life experiences with minority individua ls are extensive (e. g., via ethinically integrated neighborhoods, marri age, and friendship). 34. In order to be able to work with minority clients, I frequently seek consultation with multicultural experts and attend multicultural workshops or training sessions. When working with all clients, 35. I am effective at crisis interventions (e. g., suicide attempt, tragedy, broken relationship). 36. I use varied counseling techniques and skills. 37. I am able to be concise and to the po int when reflecting, clarifying, and probing. 38. I am comfortable with exploring sexual issues. 39. I am skilled at getting a client to be sp ecific in defining and clarifying problems. 40. I make my nonverbal and verb al responses congruent.

PAGE 50

50 The Marlowe-Crowne Social Desirability Scale Personal Reaction Inventory Listed below are a number of statements cancerning personal attitudes and tr aits. Read each item and decide whether the statement is True or False as it pertains to you personally. 1. Before voting I thoroughly investigate the qualifications of all the candidates. (T) 2. I never hesitate to go out of my wa y to help someone in trouble. (T) 3. It is sometimes hard for me to go on with my work, if I am not encouraged. (F) 4. I have never intensel y disliked anyone. (T) 5. On occasion I have had doubts about my ability to succeed in life. (F) 6. I sometimes feel resentful when I don't get my way. (F) 7. I am always careful about my manner of dress. (T) 8. My table manners at home are as good as when I eat out in a restaurant. (T) 9. If I could get into a movie w ithout paying and be sure I was not seen, I would probably do it. (F) 10. On a few occasions, I have given up doing some thing because I thought too little of my ability. (F) 11. I like to gossip at times. (F) 12. There have been times when I felt like rebe lling against people in authority even though I knew they were right. (F) 13. No matter who I'm talking to, I'm always a good listener. (T) 14. I can remember "playing sick" to get out of something. (F) 15. There have been occasions when I took advantage of someone. (F) 16. I'm always willing to admit it when I make a mistake. (T) 17. I always try to prac tice what I preach. (T) 18. I don't find it particularly difficult to get along with loud-mouthed, obnoxious people. (T)

PAGE 51

51 19. I sometimes try to get even rather than forgive and forget. (F) 20. When I don't know something I don't at all mind admitting it. (T) 21. I am always courteous, even to people who are disagreeable. (T) 22. At times I have really insisted on having things my own way. (F) 23. There have been occasions when I felt like smashing things. (F) 24. I would never think of letting someone el se be punished for my wrongdoings. (T) 25. I never resent being aske d to return a favor. (T) 26. I have never been irked when people expresse d ideas very different from my own. (T) 27. I never make a long trip without ch ecking the safety of my car. (T) 28. There have been times when I was quite jealous of the good fortune of others. (F) 29. I have almost never felt the urge to tell someone off. (T) 30. I am sometimes irritated by people who ask favors of me. (F) 31. I have never felt that I was punished without cause. (T) 32. I sometimes think when people have a mistort une they only got what they deserved. (F) 33. I have never deliberately said someth ing that hurt someone's feelings. (T)

PAGE 52

52 Debriefing By now, you may be curious about the questions being addressed in this study. This page will serve to disclose to you th e specific aims of this study, ask you to acknowledge that you have been made aware of and understand the aims of this research, and direct you to a page of links that provide additional info rmation about the topic on which this research is focused. The aim of this study was to examine th e relationship between multicultural competence and decision making in the context of child pr otective services. As populations of countries across the globe become more r acially/ethnically diverse, multic ultural competence is becoming increasingly important in all areas of human serv ice interaction. In the present study, we were interested in examining the relationship betw een multicultural competence and the assessment and intervention recommendations made by individuals who work with children and families from diverse racial and ethnic backgrounds. We hope to draw on this research to help develop and improve training programs aimed at enhancing multicultural competence among caseworkers and others who work with childre n and families who may become involved in the child welfare system. Fulfilling the aims of this research requires th at the exact nature of the research aims, and focus of the study, not be disclo sed until after participants have completed the survey. Your answers to all survey items remain completely anonymous and individu al responses cannot be linked to you in any way. In a moment, you will be directed to a page that includes various links to websites which can provide information, training, and further link s regarding multicultural competence, as well as a short questionnaire regarding your experien ce with participating in this study. Utilization (or non-utilization) of the links provided and fee dback questions cannot be linked to answers to previous items in this study and remain anonymous. Please check the box or sign below to signi fy that you have read this form and understand the aims of this study and the need fo r this information to be provided you after you have completed the study. Whom to contact if you have questions about the study: The principal investigator for this st udy can be contacted using the following information: Michael Ghali M.A. Doctoral Student in Counseling Psychology University of Florida Department of Psychology Room 114, P.O. Box 112250 Gainesville, FL 32601 Phone: (352) 392-0601 Fax: (352) 392-7985

PAGE 53

53 Email: ghali@ufl.edu Additional questions can be directed to: UFIRB Office, Box 112250, University of Fl orida, Gainesville, FL 32611-2250; ph 3920433. I have read the above debriefing form and have been made aware of the full purpose of this research. Any attempts at deception have been di sclosed and explained to me and I am aware of whom to contact should I have a ny questions or concerns regardi ng the research in which I have participated. ___________________________________ ______________ Signature of participant Date Please initial if you would like your data to NOT be used in this study: __________

PAGE 54

54 Feedback Form Please share your feedback regarding your experi ence participating in this research study. Your answers are complete anonymous, so please be honest and forward with your responses. Your feedback can help in the design and admi nistration of improved re search studies in the future. Completion of this feedback form is op tional and will in no way effect your responses to previous items in this study. 1. How would your rate your overa ll experience as a pa rticipant in this research study? a. Excellent b. Good c. Okay d. Poor 2. The time required to participate in this study was: a. Much too long b. Somewhat long c. Adequate d. Too short 3. Despite the original description of th e study, the true aims of the study were: a. Absolutely obvious b. Not too hard to figure out c. Not clear d. I had no idea 4. The instructions for participating in the study were: a. Very clear and easy to follow b. Somewhat clear and easy to follow c. At times hard to follow d. Not at all clear or easy to follow 5. Disclosing all of the true ai ms of the study after you comp leted your participation was: a. Completely justified b. Probably unnecessary, but it didnt bother me c. Probably necessary, but it bothered me that I was initially deceived d. Completely unjustified 6. Please share any other comments you may have about your experi ence in part icipating in this study: __________________________________________________________________ ________________________________________________________________________ _______________________________________

PAGE 55

55 REFERENCES Bellini, J. (2003). Counselors multicultural competencies and vocational rehabilitation outcomes in the context of counselor-clien t racial similarity and difference. Rehabilitation Counseling Bulletin, 46(3), 164-173. Bernatovicz, F. (n.d.). Retention of child welfare caseworkers: A report National Child Welfare Resource Center for Organizationa l Improvement. Accessed 7/19/06 at: www.muskie.usm.maine.edu/help kids/pubstext/retention.html Carter, R. T., Helms, J. E., & Juby, H.L. (2004) The relationship between racism and racial identity for white americans: A profile analysis. Multicultural Counseling and Development, 32, 2-17. Chibnall, Dutch, Jones-Harden, Brown, & Gourdine (2003). Children of color in the child welfare system: Perspectives from the child welfare community. National Clearninghouse o f Child Abus e and Neglect Information. Accessed 11/14/05 at http://nccanch.acf.hhs.gov/pubs/o therpubs/children/index.cfm Child Welfare League of America. (1999). Mini mum education required by state child welfare agencies, percent, by degree type, 1998. State Child Welfare Agency Survey. Cohen, E. (2003). Framework for culturally co mpetent decision-making in child welfare. Child Welfare. 82(2), 143-155 Constantine, M. G., Gloria, A., & Ladany, N. (2002). The factor structure underlying three selfreport multicultural counseling competence scales. Cultural Diversity and Ethnic Minority Psychology, 8(4), 334-345. Constantine, M.G., & Ladany, N. (2001). New vi sions for defining and assessing multicultural counseling competence. Handbook of multicultural counseling (2nd Ed.), (pp.482-498). Sage Publications: Thousand Oaks, CA. Constantine, M. G., & Ladany, N. (2000). Sel f-report multicultural comp etence scales: Their relation to social desirability attitudes and multicultural case conceputalization ability. Journal of Counseling Psychology, 47(2), 155-164. Crowne, D., & Marlowe, D. (1960). A new scal e of social desirabi lity independent of psychopathology. Journal of Consulting Psychology, 24, 349-354. DAndrea, M., Daniels, J., & Heck, R. ( 1991). Evaluating the impact of multicultural counseling training. Journal of Counseling and Development, 70, 143-150. Davidson-Arad, B. (2001). Predicted changes in childrens quality of life in decisions regarding the removal of children at risk from their homes. Children and Youth Services Review, 23(2), 127-143.

PAGE 56

56 Eckenrode, J., Powers, J., Doris, J., Munsch, J ., & Bolger, N. (1988). Substantiation of child abuse and neglect reports. Journal of Consulting and Clinical Psychology, 56(1), 9-16. Field, A. (2000). Discovering statistics us ing SPSS for Windows. London: Sage Publications. Giovannoni, J., & Becerra, R. (1979). Defining child abuse. New York: The Free Press. Harran, E. (2002). Barriers to effective ch ild protection in a multicultural society. Child Abuse Review, 11, 411-414 Hong, G., & Hong, L. (1991). Comparative persp ectives on child abus e and neglect: Chinese versus hispanics and whites. Child Welfare, 70(4), 463-475. LaFromboise, T., Coleman, H., & Hernandez, A. (1991). Development and factor structure of the cross-cultural counse ling inventoryrevised. Professional Psychology: Research and Practice. 22(5), 380-388 Lamb, K. L. (1979). Outcomes of child abuse complaints: Caseworkers as predictors. (Doctoral Dissertation, Universi ty of Arizona, 1979). Dissertation Abstracts International, 40 (2A), pp. 1109-1110. McPhatter, A. (1997). Cultural competence in child welfare: What is it? How do we achieve it? What happens without it? Child Welfare, 76(1), 255-288 Paternoster, R., Brame, R., Mazerolle, P., & Piquer o, A. (1998). Using the correct statistical test for the equality of regression coefficients. Criminology, 36(4), 857-866. Pierce, R. L. & Pierce, L. H., (1996). Moving to ward cultural competence in the child welfare system. Children and Youth Services, 18(8), 713-731 Ponterotto, J., & Casas, M. (1991). Handbook of racial/ethnic minority counseling research Springfield, IL: Charles C. Thomas. Ponterotto, J., Gretchen, D., Ut sey, S., Rieger, B., & Austin, R. (2002). A revision of the multicultural counseling awareness scale. Multicultural Counseling and Development, 30, 153-180. Pope-Davis, D., & Dings, J. (1995). The assessme nt of multicultural co unseling competencies. In Joseph G. Ponterotto & Manuel J. Casas (Eds.) Handbook of multicultural counseling. Thousand Oaks, CA: Sage Publishing. Pope-Davis, D., & Dings, J. (1994). An empi rical comparison of two self-report multicultural counseling competency inventories. Measurement and Evaluation in Counseling and Development, 27(2), 93-102.

PAGE 57

57 Pope-Davis, D., Liu, W., & Nevitt, J. (2000). The development and initi al validation of the multicultural environmental inventor y: A preliminary investigation. Cultural Diversity and Ethnic Minority Psychology, 6(1), 57-64 Pope-Davis, D. B., Liu, W. M., Toporek, R. L., & Brittan-Powell, C. S. (2001). Whats missing from multicultural competency research: Re view, introspection, and recommendations. Cultural Diversity and Ethnic Minority Psychology. 7(2), 121-138 Roberts, D. (2002). Shattered bonds: The color of child welfare. New York, NY: Basic Civitas Books. Shapira, M., & Benbenishty, R., (1993). Modeling judgments and decisions in cases of child abuse and neglect. Social Work Research and Abstracts, 29(2), 14-19. Sodowsky, G. R. (1996). The multicultural counse ling inventory: Validity and applications in multicultural training. In G. R. Sodowsky and J.C. Impara (Eds.), Multicultural assessment in counseling and clinical psychology (pp. 283 324). Lincoln, NE: Buros Institute of Mental Measurements. Sodowsky, G, Kuo-Jackson, P., Richardson, M., & Corey, A. (1998). Correlates of selfreported multicultural competencies: Counselor multicultural social desirability, race, social inadequacy, locus of control r acial ideology, and multicultural training. Journal of Counseling Psychology. 45 (3), 256-264 Sodowsky, G., Taffe, R., Gutkin, T., & Wise, S. (1994). Development of the multicultural counseling inventory: A self-report m easure of multicultural competencies. Journal of Counseling Psychology, 41(2), 137-148. Sue, D. W., Bernier, J. E., Durran, A., Feinber g, L., Pedersen, P., Smith, E. J., Vasquez-Nutall, E. (1982). Position paper: Cross-cultural counseling competencies. The Counseling Psychologist, 10(2), 45-52. Sue, D. W., Arredondo, P, & McDavis, R. (1992). Multicultural counseling competencies and standards: A Call to the Profession. Journal of Multicultural Counseling and Development. 20(2), 64-68 Taussig, H. (2002). Risk behavi ors in maltreated youth placed in foster care: A longitudinal study of protective and vu lnerability factors. Child Abuse and Neglect. 26, 1179-1199 Third National Incidence Study of Child Abuse and Neglect (1993). Re trieved 11-4-04 at: http://www.healthieryou.com/cabuse.html United States Census Bureau (2004). P opulation by race 2003. Retrieved 11-4-04 at: http://www.census.gov/popest/state s/asrh/tables/SC-EST2003-04.xls

PAGE 58

58 Walker, S. (2002). Culturally competent protection of children s mental health. Child Abuse Review, vol. 1, 380-393. Walters, K.L., & Wheeler, D. P. (2002). Assess ment of multicultural competence scales for social work practitioners Retrieved 10/14/05 at : http://www.columbia.edu/cu/csswp/ research/desriptions/Walters.htm Wierzbicki, M., & Pekarik, G. (1993). A meta-analysis of psychotherapy dropout. Professional Psychology: Research and Practice, 24(2), 190-195. Worthington, R., Mobley, M., Franks, R., & Tan, J.A. (2000). Multicultural counseling competencies: Verbal content, counselo r attributions, and social desirability. Journal of Counseling Psychology, 47(4), 460-468 Zuravin, S., Orme, J., & Hegar, R. (1994). Predic ting severity of child a buse with ordinal probit regression. Social Work Research, 18(3), 131-138.

PAGE 59

59 BIOGRAPHICAL SKETCH Michael Ghali was born in Fort Thomas, KY, in 1975. The second of three children, Michael graduated from Newark High School in Newark, OH, in 1993. During his high school years, he was active on the soccer, swimming, and tennis teams as well as a singer in numerous choruses. Upon acceptance to the University of MissouriColumbia School of Journalism, Michael moved to Missouri, but quickly realiz ed his career passion for communication would be better served in helpi ng others, rather than simply reporting news. Michael graduated from Mizzou with a Bachelor of Scien ce degree in psychology in 1997 and spent the next few years working for the Fam ily Counseling Center of Missouri, Inc. as a case manager and program director. During this time, Michael met his future wife, Chris, through some mutual friends, began playing guitar, and participated on numerous city league softball teams. Satisfied in his chosen field, Michael moved to Colorado with Chris to pursue graduate studies at the University of Northern Colorado. Completion of a two-year program culminated in the award of a Master of Arts degree in community counseling. Immediately following completion of this program, Michael and Chris m oved to Florida with the hopes of being closer to family, starting a family of their own, and finding a location for Michael to continue his education. After providing individual, fa mily, and group counseling serv ices for adolescents and adults, Michael was accepted into the docto ral program in counseling psychology at the University of Florida. He currently resides in Gainesville, FL, with Ch ris and their 2-year-old son, Jace. Upon completion of a Ph.D., Michael intends to teach undergraduate psychology at a college and pursue private practice work ing with children and families.


Permanent Link: http://ufdc.ufl.edu/UFE0017949/00001

Material Information

Title: Multicultural Competence and Decision-Making Bias in Child Protective Services
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0017949:00001

Permanent Link: http://ufdc.ufl.edu/UFE0017949/00001

Material Information

Title: Multicultural Competence and Decision-Making Bias in Child Protective Services
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0017949:00001


This item has the following downloads:


Full Text





MULTICULTURAL COMPETENCE AND DECISION-MAKING BIAS IN CHILD
PROTECTIVE SERVICES




















By

MICHAEL N. GHALI


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE

UNIVERSITY OF FLORIDA

2006

































Copyright 2006

by

Michael N. Ghali












TABLE OF CONTENTS page


LI ST OF T ABLE S ................. ...............4................


AB S TRAC T ......_ ................. ............_........5


CHAPTER


1 INTRODUCTION ................. ...............6.......... ......


Overrepresentation of African-American Children in Child Welfare: A Closer Look ............7
A Need for Cultural Competence .............. ...............10....
M multicultural Competence .................. .............. ...............11......
Multicultural Competence and Child Welfare ................. ...............14........... ...

2 M ETHODS ................. ...............17.......... .....


Participants .............. ...............17....
Instrum ents .............. .. ...............18...

Demographic Questionnaire ................. ...............18.................
Case Vi gnettes ................. ................. 18..............
Appraisal Rating Questions ................. ...............20........... ....
Multicultural Counseling Inventory .............. ...............20....
Marlowe-Crowne Social Desirability Scale .............. ...............22....
Procedure .............. ...............22....


3 RE SULT S .............. ...............24....


4 DI SCUS SSION ............ ..... ..__ ............... 2...


Current Limitations ............... ... ....._ ...............33....
Recommendations for Future Research ............. .....___ ...............36..

Im plications .............. ...............38....

APPENDIX.....FORMS AND MEASURES ................. .........___.....__................38


Informed Consent .............. ...............39....

Vignettes: Form A .............. ...............41....
Vignettes: Form B............... ...............43...
Appraisal Rating Questions ....................... ...............45
Multicultural Counseling Inventory (MCI) .............. ...............47....
The Marlowe-Crowne Social Desirability Scale .............. ...............50....
Debriefing ............. ...... ._ ...............52....
Feedback Form .............. ...............54....


REFERENCES .............. ...............55....


BIOGRAPHICAL SKETCH ............. .............59......














LIST OF TABLES
Table Page

3-1 Means, standard deviations, and correlations among variables (N = 220) ................... .....29

3-2 Results of MANOVA with combined rating indices as DV' s and version (race of the
child depicted) as IV ................. ...............30................

3-3 Summary of hierarchical regression analysis of appraisal rating indices on
multicultural competence............... ...............3

3-4 Summary of hierarchical regression analyses of multicultural competence on african
american and european american appraisal rating indices and Z-statistics of race of
the child as a moderator ................. ...............3.. 1......... ...












Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science

MULTICULTURAL COMPETENCE AND DECISION-MAKING BIAS IN CHILD
PROTECTIVE SERVICES

By

Michael N. Ghali

December 2006

Chair: Mark Fondacaro
Major Department: Psychology

African-American children are overrepresented in the American Child Welfare system.

Research has concluded that this overrepresentation is not due to higher rates of abuse and

neglect in African-American families, but more likely due to differential treatment of African-

Americans somewhere along the process of reporting, investigation, and service allocation. This

study serves as an initial examination into the relationships between multicultural competence,

race, and decision-making bias in the investigation phase of the child protective service process.

Participants were asked to read a series of case vignettes and make decisions regarding 1) the

extent to which they believe the information is evidence of actual child abuse or neglect, 2) level

of risk to the child and 3) level of recommended intervention, based upon the information

presented in the vignettes. Multivariate Analysis of Variance (MANOVA) and regression

analyses were used to analyze the data. Results from this sample of data suggest that ratings of

perceived risk and recommended intervention do vary based on the race of the child and that the

multicultural competence of the investigators does significantly predict ratings of substantiation

agreement, level of perceived risk, and recommended intervention. Results and limitations are

discussed and recommendations for future research are presented.









CHAPTER 1
INTTRODUCTION

Minority children and families are, at times, poorly served by available social and human

services. This includes counseling, social work, health care, and other services. For example,

African-American children are overrepresented in the Child Welfare system in America. This

means that African American, or Black, children are represented in the Child Welfare system in

higher proportions than they are in the general population of children in the United States.

Although Blacks account for only 15% of all children in the United States, they account for 25%

of substantiated maltreatment victims and comprise 45% of the total number of children in foster

care (Chibnall, Dutch, Jones-Harden, Brown, Gourdine, 2003). In addition, these minority

children spend longer in the child welfare system (Roberts, 2002). These disproportions are

evidence of the differential treatment of African-American and Black children and families in the

child welfare system.

Current trends in the general population give more cause for alarm. By the year 2050, the

U.S. Census Bureau (2004) expects that people who report to be "White alone, Not Hispanic"

will make up only 50. 1% of the population. As the proportion of minorities in the general

population gets closer to, and will ultimately overtake, the maj ority population, these and other

problems in social/human services may compound themselves if adjustments are not made in the

way human service workers are trained, the way social services are delivered, and in the way the

systems themselves are organized and operated.

One model that shows promise in promoting positive changes in training professionals

who serve minority groups is a model of Multicultural Counseling Competence developed in the

field of counseling psychology (Sue & Arredondo, 1992). Perhaps this model can help to

explain the problem of African-American over-representation in child welfare by culturally less-









aware, less-knowledgeable, or less-skilled decisions made by child welfare workers. This paper

will outline the problem of African-American over-representation in Child Protective Services

(CPS), offer rationale for applying multicultural competence concepts to the problem, and report

and discuss results from a study which aims to explore how those concepts are related to or

might help explain racial disparities in child protective service delivery.

Overrepresentation of African-American Children in Child Welfare: A Closer Look

African-American families were, and are, receiving child welfare services, often uninvited,

at a proportionally greater rate when compared to white families. Roberts (2002) reports that

black children make up 17% of the children' s general population yet 42% of the child welfare

population. In addition, once black children enter the child welfare system, they often experience

different paths through the system than maj ority children. For example, black children end up in

foster care more often than white children and are reunified with their natural families at a slower

and lower rate (Roberts, 2002). Although the intent of the child welfare system is to make

decisions in the best interest of the child, entering the child welfare system, and foster care in

particular, is correlated with other negative life experiences, such as delinquency and entering

the juvenile justice system (Taussig, 2002), and can set the child up for a difficult road through

life. One typical entry point into the child welfare system is through Child Protective Services

(CPS). Generally speaking, suspected abuse or neglect is reported to a child abuse and neglect

hotline. When the report is made, the hotline operator gathers demographic information

regarding the family in question and information about the abuse or neglect and how it was

identified, as reported by the often anonymous caller. The hotline operator then decides whether

or not to accept the report as valid. If the report is accepted, the reported information is sent to

Protective Investigation.









A Protective Investigator (PI) is sent to investigate the report. The PI has only a few hours

to gather any information on which to make two crucial decisions. First, the PI must decide

whether or not the abuse/neglect actually happened. If it is determined that the abuse/neglect has

indeed occurred, the report is said to be 'substantiated'. If no evidence of abuse/neglect is found,

the report is 'unsubstantiated'. Although states have laws to define what is abuse and neglect,

these laws can be interpreted differently by different people and there may be gray areas that

remain unclear despite the laws. If evidence of abuse/neglect is found by the PI and the report

is substantiated, the PI must then make a decision to either, leave the child home and provide

services in that context, or remove the child from the home due to the imminence of future harm

and place the child in a foster care setting. An investigator often makes an assessment of the

level of immediate and future risk to the child to help make these decisions (National

Association of Child Welfare Administrators, 1999).

Once the initial investigation is complete and those initial decisions made, the case is

transferred to another worker, considered to be Protective Supervision (PS). This worker

conducts an ongoing assessment to determine the safety of the child and needs to the family.

The PS worker also decides what services to offer to the family and the manner in which those

services will be provided (National Association of Child Welfare Administrators, 1999).

Clearly, there are many important decisions made during the entire CPS process and it may be

reasonable to think that cultural differences, and perhaps misconceptions based on those

differences, may impact how those decisions are made.

Before we examine reasons for the overrepresentation of black children in the system, we

can rule out some other possible explanations. One possible explanation for this

overrepresentation of black children in the child welfare system is that they actually experience









higher rates of abuse/neglect. One study has investigated this claim, but could find no support

for it. The Third National Incidence Study of Child Abuse and Neglect (NIS-3) conducted by

the Department of Health and Human Services in 1993 found that black children were not

abused or neglected at a rate that differed from that of other racial groups. The report concluded

that "differential representation of minorities in the child welfare population does not derive

from inherent differences in the rates at which they are abused or neglected."

If rate of maltreatment is not the cause, perhaps severity of maltreatment is the reason more

African American children are reported to be abused, subsequently investigated, and entered into

the system than majority children. Various researchers have examined this factor, but

unfortunately have been unable to arrive at a consensus. For example, Hampton (1987) found

that, in regards to physical abuse, severity of injury adds statistically significant power in

discriminating white and black families at both low and high SES. In contrast, in the context of a

study controlling for more variables, Zuravin and Orme (1994) found that the child' s race did not

have a significant main effect on severity of abuse.

Another possible explanation is that because the child welfare system is designed to

address issues relating to poor families, and because black families are disproportionately poor,

this is the reason more black children are entered into the system. However, Roberts (2002)

points out that there is good reason to suspect that poverty cannot completely explain the

system's racial disparity. Roberts (2002) is referring to findings that approximately the same

percentage of Latino and Black households earn less than $15,000, the income level most highly

associated with child maltreatment (NIS-3, 1993), yet Latino children are placed in foster care at

a rate identical to their proportion of the population.









After setting aside these possible explanations, it seems likely that the child's race

influences the overrepresentation of African American children in the child welfare system. In

fact, prior research has revealed numerous references to race as a contributing factor in disparate

decisions in child welfare processes. For example, Lamb (1979) found that of 17 attribute

variables entered into an analysis of variance with case outcome as the DV, only 4 of those

attributes significantly increased the explained variance. Race of the child was one of those four

variables and contributed significantly to the total explained variance. Eckenrode, Powers,

Doris, Munsch, and Bolger (1988) conducted a study in which they examined the child welfare

records in the state of New York from the year 1985. Their data suggested that "for physical

abuse reports, no background characteristics other than ethnicity had an effect on substantiation,

but this effect was quite strong, accounting alone for 8% of the variance in substantiation" (p.

15). More recently, the NIS-3 (1993) report reveals findings that "suggest that the different races

receive differential attention somewhere during the process of referral, investigation, and service

allocation."

A Need for Cultural Competence

Some within the child welfare system are realizing the need to diagnose and make needed

adjustments in regards to the problem of over-representation. In a study authored by Chibnall,

Dutch, Jones-Harden, Brown, and Gourdine (2003), the authors concluded that the need for

continuing research in the area of overrepresentation is evident.


Researchers, policy-makers, and practitioners have divergent views on the
causes of minority over-representation. This phenomenon may be the
result of a disproportionate need for services or of systematic racial
influences on decision-making at any number of points along the
continuum of child welfare services, including reporting, investigation,
substantiation, and placement. Researchers have attempted to explore
levels of need and to examine how race affects children's experiences at
each of these points, but Eindings have been inconsistent. Where racial









differences have been found, the reasons for these differences remain
unclear. (pg. 15)

The Chibnall et al. (2003) study involved a focus group of agency administrators, field workers,

and policy makers in the field of child welfare from across the country. The participants

"identified racial bias as a common problem that frequently interfered with good decision

making" (Executive Summary, p. iii). Participants also mentioned the need for "culturally

competent and experienced staff" as a "strategy to improve services to families of color and

families in general" (p. 34), noting that staff often lack exposure to different cultures and had no

context for understanding the norms and practices of other cultures.

Others have also reported the need for cultural competence in the child welfare system.

Pierce and Pierce (1996) noted that "workers must recognize that their culture determines how

they define family, what values are emphasized and even how to greet a person properly" (p.

719). Cohen (2003) referenced a specific dimension of cultural competence when she referred to

attitudes (i.e., awareness) of the child welfare staff as an important variable in decision-making.

Finally, one recommendation of the Chibnall et al. (2003) study was that future research should

examine the effects of cultural competency and cultural sensitivity training on actual child

welfare practice.

Multicultural Competence

Researchers in the field of social work and child welfare have begun to develop models of

cultural competence. For example, McPhatter (1997) presented a Cultural Competence

Attainment model. This model consists of three interrelated components: Enlightened

Consciousness, Grounded Knowledge Base, and Cumulative Skill Proficiency. Despite available

models, such as the Cultural Competence Attainment model, research regarding the assessment

of practitioner and student levels of cultural competence and how these levels affect practice,









interventions, and outcomes in child welfare has been lacking. Some believe that the reason for

this lack of documented research is due to "insufficient development of multicultural competence

practice measures" (Walters & Wheeler, 1998). It seems that until valid measures of cultural

competence have been developed in the Hield of social work, cultural competence research based

on existing models in the Hield will be stalled.

Perhaps the most well-developed and empirically investigated model of cultural

competence exists in the Hield of counseling psychology. In 1980, Division 17 (Counseling) of

the APA put their Educational and Training Committee to work on developing minimal cross-

cultural counseling competencies for training programs to incorporate into their training and

curricula (Sue et al., 1982). The result was the first formally documented "characteristics of the

culturally skilled counselor" posited in Sue et al.'s position paper. This model has been

developed over the past two decades and has come to be known as a model of Multicultural

Counseling Competencies. The model is built on three main dimensions. The first dimension,

Awareness, refers to the process of becoming aware of one' s own assumptions about human

behavior, values, biases, preconceived notions, personal limitations, and so forth. The second

dimension, Knowledge, refers to attempts to understand the worldview of culturally different

clients without negative judgment. The third dimension, Skills, refers to actively developing and

practicing appropriate, relevant, and sensitive intervention strategies and skills in working with

culturally different clients. These three dimensions are conceptually very similar to the three

dimensions contained in the Cultural Competence Attainment model described above.

Considering the findings, assertions, and recommendations discussed in the literature, it seems

that the dimensions of awareness, knowledge, and skills might play a role in how decisions are

made in the provision of child welfare and child protective services.









In fact, some research has already found a connection between these dimensions and

differential decisions when race is involved. A study of Chinese, Hispanic, and White parents

found that the three cultural groups demonstrated differences in determinations of both the

presence and severity of abuse and neglect (Hong and Hong, 1991). Specifically, the Chinese

raters in their study "tended to judge parental conduct less harshly and recommended agency

intervention less frequently than Hispanics and whites" (pg. 5). The authors discuss the findings

in terms of cultural explanations and note that cultural sensitivity is called for when subj ects'

reactions and response patterns can be attributed to characteristics of their culture. Cohen (2003)

notes that child welfare workers ultimately "make decisions on family functioning, parent-child

relationships, risks, and child safety based more on their perceptions, attitudes, and judgments

than on cultural factors that influence child rearing" (p. 149). A possible implication of this

statement is that child welfare workers do not possess enough cultural awareness, knowledge,

and skill to make decisions based on cultural factors. As a consequence, they rely on their

perceptions, attitudes, and judgments rather than cultural factors to make decisions.

One advantage of the Sue et al. model of multicultural competence is that various measures

of multicultural competence have been developed from this model and used in related research.

The most intensive effort to develop such tools occurred in the early 1990's (D'Andrea, Daniels,

& Heck, 1991; LaFromboise, Coleman, & Hernandez, 1991; Ponterotto & Casas, 1991;

Sodowsky, Taffe, Gutkin, & Wise, 1994). The resulting measures were all initially created from

the three dimensions in the Sue et al. model: Awareness, Knowledge, and Skills, and are

organized into two (Awareness & Knowledge/Skills), three (Awareness, Knowledge & Skills),

or four (Awareness, Knowledge, Skills, & Relationship) factors based on the items contained in

the measure.









Multicultural Competence and Child Welfare

Conceptual links can be made between decisions at different stages of the CPS process and

specific items on these measures. For example, because PI's do not usually have more than

essentially a few of hours to make such influential decisions, initial impressions could be

extremely influential in how situations are perceived and, subsequently, how decisions are made.

In this type of scenario, it may be important that the PI be "aware of differential interpretations

of nonverbal communication (e.g., personal space, eye contact, handshakes) within various

racial/ethnic groups (Ponterotto, Gretchen, Utsey, Rieger, & Austin, 2002, item from MCKAS).

Again, as a worker it may be crucial to your decisions to have an "understanding of how your

cultural background has influenced the way you think and act" (D'Andrea et al., 1991, from an

item on the MAKSS).

Davidson-Arad (2001) found empirical evidence that there is a "significant relationship

between the decision to remove children at risk from home .. and assessments of their parents

as uncooperative .. ." (pg. 127) Parental cooperation was measured by three questions

regarding parents' "cooperation with the child protection worker, motivation for change, and

acceptance of responsibility for the problem." (pg 132). This finding provides grounds for

another conceptual link between workers' decisions and items on measures of multicultural

competence in that it may be useful for the child protection worker to "have a working

understanding of certain cultures" or be "familiar with nonstandard English" (Sodowsky et al.,

1994, from items on the MCI) when working with minority families.

Roberts (2002) noted a discrepancy between the types of services usually offered to white

clients as compared to black clients. For example, white clients are more often referred for

housing assistance and black clients are more often referred for mental health services and

parenting classes. When a worker decides, for example, which services to provide for the









family, hopefully that worker is confident in his/her "ability to effectively assess the mental

health needs of a person from a cultural background different from [their] own (D'Andrea et al.,

1991, from an item on the MAKSS) or is knowledgeable of the relative importance of family,

community, and faith based on the clients' culture and variations within the clients' culture.

Considering the links, both empirical and conceptual, between racial disparities in the child

welfare system and multicultural competence, research directly examining the influence of

multicultural competence in racially disparate decisions is essential. Previous research has not

included multicultural competence and child welfare decision making as variables in the same

experimental paradigm; the proposed research aims to do so. Specifically, this author is

interested in exploring the following research questions: Are decisions made in the context of

child protective services different depending the race of the child? Does multicultural

competence relate to decisions made in the child protective service context? Does race of the

child moderate the relationship between multicultural competence and decisions in the context of

child protective services?

Exploratory in nature, this study will examine the relationship between multicultural

competence on the one hand and child-related investigative decisions (i.e., substantiation

agreement, assessed level of risk, recommended level of intervention) on the other. Based on

empirical evidence from field data linking race to differential decisions in the child protective

service context (Eckenrode et al., 1988; Lamb, 1979; Pierce & Pierce, 1996) and anecdotal

evidence from previous research citing the ability of case vignettes to elicit those differential

decisions (Roberts, 2002; Shapira & Benbenishty, 1993) the aim of the current study is to test the

following hypotheses:









HYPOTHESIS 1: Mean appraisal ratings of substantiation agreement, level of risk, and
recommended intervention will be higher when case vignettes depict African American
children compared to when vignettes depict Caucasian children.
A between-subj ects regression model will be the basis of testing the second and third hypotheses.

Drawing on recent work by Bellini (2003) showing a significant effect of multicultural

competence on client outcomes, a second hypothesis was developed:

HYPOTHESIS 2: After controlling for social desirability, multicultural competence will
account for unique variance in appraisal ratings of substantiation agreement, level of risk,
and recommended intervention, regardless of the race of the child depicted in the
vignette.

Multicultural competence of the counselor has been shown to relate to vocational outcomes

when the client is African American, but the relationship is weaker when the client is Caucasian

(Bellini, 2003). Similarly, we expect multicultural competence of the caseworker to relate to

ratings of substantiation, level of risk, and recommended intervention when the child is African

American, but we expect the relationship to be weaker when the child is Caucasian.

HYPOTHESIS 3: Race of the child depicted in the vignette will moderate the
relationship between multicultural competence and ratings of substantiation agreement,
level of risk, and recommended intervention, such that multicultural competence will
account for a greater amount of unique variance in ratings (i.e., substantiation agreement,
level of risk, recommended intervention) when examining ratings of vignettes depicting
African American children and will account for less variance when examining ratings of
vignettes depicting Caucasian children.









CHAPTER 2
IVETHOD S

This study was conducted via the internet. Participation was voluntary and participants

were not be reimbursed for their participation in the study. An online survey, including an

informed consent process and debriefing was posted on a secure server and two separate methods

of participant recruitment were used; the undergraduate participant pool at a large Southeastern

university was utilized, and graduate students of social work as well as caseworkers were

accessed through emails requesting participation. Detailed procedures are described below.

Following submission of their responses, participants were provided links to various resources

related to multicultural competence in social work and were also given the opportunity to

voluntarily complete a feedback form regarding their experience as a participant in the study.

Criteria for inclusion in the study included age (18 & over), current enrollment in a human

service-related program of study (e.g., sociology, children and family services, psychology,

social work) or current employment in a child welfare agency that conducts child protective

investigations, and completion of the entire survey.

Participants

Combined methods of participant recruitment resulted in 266 total survey response

submissions. Of these, 45 were not complete and were dropped from all data analyses. The

resulting final sample (N=221) was comprised of 157 (71%) undergraduate students, 25 (11.5%)

graduate students, 23 (10.5%) caseworkers, and 15 (7%) who marked "other", but whose

reported occupations directly or indirectly involve child welfare education or service provision.

Females made up the majority of the sample; 164 (74%) compared to 57 (26%) males. The

ethnic breakdown of the sample was also uneven, with Caucasians making up the maj ority with

162 (73.5%), compared to 15 (7%) African-Americans, 11 (5%) Asian-Americans, 18 (8%)










Hispanic-Americans, and 11 (5%) participants who marked "other". The distribution of

degree/job categories is of particular consideration. The high percentage of undergraduate

psychology students is of viable use because it is reported that caseworkers who are most likely

to retain their positions are those with undergraduate degrees (Bernotavicz, n.d.) and less than

30% of child welfare workers possess degrees specific to social work (CWLA, 1999).

Instruments

The instruments used in this study included a demographic questionnaire, a series of case

vignettes describing a potential case of abuse or neglect, three appraisal rating questions, the

Multicultural Counseling Inventory (MCI) (Sodowsky, Taffe, Gutkin, and Wise, 1994), and the

Marlowe-Crown Social Desirability Scale MCSDS (Crowne & Marlowe, 1960).

Demographic Questionnaire

The demographic questionnaire included basic demographic items as well as questions

related to child welfare work experience. Answers to these questions were used in follow-up

post hoc analyses. Participants were not asked for any identifying information (e.g., name, date

of birth).

Case Vignettes

Case vignettes were used in this study to give respondents a basis on which to make

decisions (i.e., ratings) regarding probability of substantiation, levels of risk, and levels of

recommended intervention they perceived based on the information presented in the vignette.

Using case vignettes to have participants make decisions about the level of risk and agency

intervention required has been established as a method of examining differential decisions based

on race (Shapira & Benbenishty, 1993; Giovannoni & Bacerra, 1979; Roberts, 2002).

The vignettes used in this study were borrowed from Giovannoni and Bacerra (1979) and

adapted to suit the purposes of the study. A total of 20 vignettes were utilized in the current










study. Four of the vignettes were adapted with both an African American version and a

European American version. The two versions were identical except for race of the child

depicted in the vignette. These vignettes operated as between-subj ects items in order to examine

if differential decisions were made based on race. The remaining 16 vignettes included 4

African American, 4 European American, 4 Hispanic American and 4 Asian American race

designations. These 16 vignettes were rated by all participants and were combined to create

various indices. Every effort was made to ensure that vignettes assigned to various races during

the adaptations for the current study were rated at equivalent levels of severity based upon results

of the Giovannoni and Bacerra (1979) study.

An example of a vignette to be presented to participants was; "The parents of an African

American child ignored their child's complaint of an earache and chronic ear drainage. The

child was found to have a serious infection and damage to the inner ear." (A complete list of the

vignettes to be used can be found in the appendix to this paper) Limited information was

provided in the vignette; surely not enough information to make a well-informed decision in

actual practice. However, as this study constituted an initial step in this line of research and was

presented to participants as 'An examination of decisions made in the context of child protective

services based on limited information,' using simple vignettes such as these is justified. In

addition, to deter participants from becoming skeptical, they were presented with the following

statement before reading any of the vignettes:

The following statements were taken from real child abuse or neglect
report summaries. Following each statement, you will be asked to respond
to three items regarding your initial appraisal of the situation. Please read
each statement carefully and respond to each item to the best of your
professional ability based on the limited information provided.










Regardless of the justifications, participants offered feedback such as, "It is hard to use just a one

or two sentence story to see what ...responses would be. We include a whole lot more

information that may sway what we do." And, "Sometimes, I felt that there wasn't enough case

information provided regarding what decisions to make."

Appraisal Rating Questions

Following each vignette, participants were asked to provide responses to three items

related to the vignette. The first item was related to substantiation and asked, "To what extent to

do agree that the situation described above provides evidence of actual abuse or neglect?"

Responses were measured on a 4-point scale (ranging from 1--Strongly Disagree to 4--Strongly

Agree) that participants chose using a drop-down box on the survey. The second item asked

participants to "rate the degree to which you believe this child to be at risk." Responses were

measured on a 7-point scale (ranging from 1--No risk to 7--Extreme risk). The third item asked

participants to choose a recommended level of intervention for each case from 6 options (ranging

from 1-No further intervention to 6-Removal of the child from the home....) The purpose of

using different ranges on the rating scales was so that participants wouldn't simply mark the

same rating for all three items for the sake of consistency.

Multicultural Counseling Inventory

The MCI is intended to measure the "competencies of any counselor working with a

minority or culturally different client" (Sowdowsky et al., 1994, p. 140). The MCI consists of 40

self-report items that ask respondents to rate their level of agreement, on a 4-point scale (1 very

inaccurate to 4 very accurate) with statements related to working with minority clients.

Restrictions on the use of the MCI in research prohibited the current researchers from adapting

any of the items to reflect language relevant to social work. Instead, participants were presented

with the following instructions preceding the standard MCI instructions:










Now, read the following directions and statements and apply them to you
as a social worker, caseworker, or student in a social work or social
services training program.

No feedback was offered by participants about this aspect of the study.

In addition to initial tests by the developers of the MCI, this measure has been investigated

numerous times to establish validity, reliability, and factor structure (Constantine and Ladany,

2000; Constantine, Gloria, and Ladany, 2002; Worthington et al., 2000, Pope-Davis & Dings,

1995). Constantine, Gloria, and Ladany (2002) reported evidence of the MCI' s content validity

derived through the accuracy of expert raters' classification of items into their appropriate

subscale categories and expert ratings of item clarity. They also provided evidence of criterion-

related validity (i.e., individuals with multicultural training or more experience working with

culturally diverse populations obtained higher MCI scores). Adequate construct validity has also

been established through exploratory factor analysis and confirmatory factor analysis (Pope-

Davis & Dings, 1994; Sodowsky, 1998; Sodowsky et al., 1994). High to moderate reliability

coefficients (e.g., Cronbach's alphas = .81 for the Skills sub scale, .80 for the Awareness

subscale, .67 for the Relationship subscale,.80 for the Knowledge subscale, and .86 for the full

scale) have also been reported for the MCI (Pope-Davis & Dings, 1994; Sodowsky, 1998;

Sodowsky et al., 1994).

Although the MCI was developed directly from the Sue et al. model of cultural

competence, the authors found four factors to be the best fit for the items. Aside from the three

domains of the Sue et al. model, a Relationship sub scale is included on the MCI. The

Relationship subscale refers to the counselor's interactional processes with the minority client,

including stereotypes of the minority client and worldview.












Marlowe-Crowne Social Desirability Scale

A maj ority of the literature regarding the development and utilization of measures of

multicultural competence discuss the relationship of social desirability to such measures. The

author of the MCI specifically recommends that the completion of multicultural competence

measures be accompanied by an index of social desirability (Sodowsky, 1994). Social

desirability is typically operationalized using the Marlowe-Crowne Social Desirability Scale

(MCSDS) in psychological literature (Worthington, Mobley, & Franks, 2000). The MCSDS is a

33 item true/false scale designed to measure the degree to which research participants attempt to

make a good impression when completing research instruments (Worthington et al., 2000). The

scale was originally designed to be used with measures of personality, but has since been used

numerous times in connection with measures of other constructs, including multicultural

competence. Internal consistency of the MCSDS, calculated using the Kuder-Richardson

formula 20, was found to be .88. The authors report a test-retest correlation of .89 for the scale

(Crowne & Marlowe, 1960).

Procedure

When participants clicked the link to the study website, they were presented with the

informed consent form (see appendix A). Participants were informed that clicking the link at the

bottom of the page to begin the study would be accepted as their consent to participate. They

were randomly assigned to either form A or form B when they clicked the "Take me to the

study" button by means of a randomization process built into the online survey. Form A

included the European American versions of the 4 between-subj ects vignettes (Items 1, 6, 1 1, &

16). Form B included the African American version of the 4 between-subj ects vignettes (Items

1, 1, 1).No other differences exist between Form A and Form B.









Participants were asked to complete a demographic questionnaire which included

questions about their gender, ethnicity, current occupation, years of experience working with

children and whether or not they are currently involved in making decisions about child

abuse/neglect. Before being presented with the first vignette, participants were instructed as

follows:

The following statements were taken from real child abuse or neglect
report summaries. Following each statement, you will be asked to respond
to three items regarding your initial appraisal of the situation. Please read
each statement carefully and respond to each item to the best of your
professional ability based on the limited information provided.

For each vignette, participants were asked to read the information and complete the appraisal

rating questions. One vignette and the three appraisal rating questions appeared together on each

page (i.e., screen). (Both versions of the form included in Appendix A).

Once participants read all vignettes and recorded their ratings, they were asked to complete

the MCI and the Marlowe-Crown Social Desirability Scale. Upon completion of all measures,

participants were presented with a debriefing form informing them of the complete purpose of

the study (see appendix A for debriefing form). At that point, participants were asked to indicate

(by way of a button on the webpage) that they had read and understood the debriefing. They

were also given the opportunity to indicate if they did not want their data to be used in the final

analyses. At that point, participants could submit their responses. Clicking submit

simultaneously uploaded their responses to a database and navigated participants to a page that

thanked them for their participation, presented links to numerous online resources and a feedback

form asking them about their experience as a participant in the study and offering a chance to

include any written feedback they wished. Responses to the feedback form were not linked in

any way to responses from the survey.









CHAPTER 3
RESULTS

Before any analyses were run, ratings from the vignettes were summed to form specific

indices. Specifically, ratings on the question regarding substantiation agreement from the four

between subjects vignettes (Items 1, 6, 11, & 16) were summed to form the Combined

Substantiation Index (Cronbach's alpha = .43, Spearman-Brown = .42). Similarly, a Combined

Risk Index (Cronbach's alpha = .52, Spearman-Brown = .55) and Combined Intervention Index

(Cronbach's alpha = .43, Spearman-Brown = .49) was formed. All subsequent data exploration

and analyses were conducted using SPSS 11.5. All relevant variables were examined for

normality by examining histograms and skewness and kurtosis statistics. Other than evidence of

mild kurtosis (zkurtosi = 2.22) in the distribution of the Combined Substantiation Index, all other

variables appeared to be normally distributed.

Means, standard deviations, and correlation coefficients for all relevant variables are

provided in Table 3-1 (below). The mean score on the MCI provided by the graduate students

and caseworkers (M~= 121. 65, SD = 11. 66) is roughly equivalent to the mean scores on the MCI

provided in previous literature for graduate students and other professionals (M~= 122. 73, SD =

12. 09) (Worthington, Mobley, Franks, & Tan, 2000; also see Bellini, 2003). The mean score on

the MCI for undergraduates (M~= 113. 55, SD = 11.28) was significantly lower (t (219) = 4.70, p

S.001) than the mean score for graduate students and caseworkers (no mean scores on the MCI

Total Scale for undergraduates could be located in the literature). This is to be expected

considering the lack of multicultural coursework and training provided at the undergraduate

level .

In order to test Hypothesis 1, variables were entered into a MANOVA to test for main

effects of race of the child (i.e., European American or African American) on all three appraisal









ratings. A MANOVA was chosen because the dependent variables (i.e., Combined

Substantiation, Combined Risk, and Combined Intervention) are interrelated (Pearson

Correlations of .524, .634, and .657, p < .001). Levene's Test for equality of variances was non-

significant (p > .05) for all DV' s suggesting that the main effects can be interpreted with

confi dence.

Table 3-2 (below) provides the results of the MANOVA showing a main effect for race of

the child on ratings of Risk (F (1,218) = 4.55, p < .05) and on ratings of Intervention (F (1,218) =

3.3 1, p < .05). This suggests that participants perceived higher levels of risk and recommended

more severe (i.e., intrusive) interventions when the child depicted in the vignette was African

American (mean rating of risk = 5.74; mean rating of intervention = 4.44) than when the child

was European American (mean rating of risk = 5.54 ; mean rating of intervention = 4.26). The

main effect of race of the child on ratings of Substantiation (F (1,218) = .930, p = .168) was non-

significant. (Mean African American rating of substantiation = 3.51; Mean European American

rating of substantiation = 3.47) These results provide partial confirmation for Hypothesis 1 and

lend credence to the experimental manipulation of race in case vignettes as a useful tool in

examining decision-making bias.

In order to test Hypothesis 2, three separate hierarchical regression analyses were

conducted. The three combined rating indices (Substantiation, Risk, and Intervention) were

entered as dependent variables and social desirability (operationalized as scores on the MCSDS)

and multicultural competence (operationalized as the MCI total-scale score) were entered

stepwise as independent variables. Table 3-3 (below) provides the results of each of the

appraisal rating indices regressed upon multicultural competence (results of rating indices

regressed upon social desirability not shown). After controlling for social desirability,









multicultural competence was a significant predictor of mean ratings of substantiation (t(219) =

2.43, p = .016, b = .019), mean ratings of risk (t(219) = 2.08, p = .039, b = .034), and mean

ratings of intervention (t(219) = 2.79, p = .006, b = .047). These results fully support Hypothesis

2. Specifically, the results suggest that higher levels of multicultural competence significantly

predict higher ratings on each of the rating indices. The regression models explain small

amounts of variance however, (r2 = .033, r2 = .020, and r2 = .036 respectively) according to

Cohen' s standards for effect sizes in regression analyses; although virtually all of the variance in

the model is accounted for by multicultural competence (r2 = .026, r2 = .020, and r2 = .035

respectively).

If the possibility of family-wise error by running three regression analyses is taken into

account, the Bonferroni adjusted p-value for significance would become p < .016. Thus, the

relationship between multicultural competence and Substantiation would still be significant (p =

.016), as would be the relationship between multicultural competence and ratings of Intervention

(p = .006). However, the relationship between multicultural competence and ratings of Risk

would no longer be significant (p = .039).

The final stage in these analyses was to check assumptions of the modelss. Collinearity

statistics, based on tolerance and VIF suggest that the assumption of collinearity is not violated

in any of the models. Durban-Watson was used to check whether residuals in the model are

independent. Field (2000) suggests that Durban-Watson values less than 1 or greater than 3 are

cause for concern. Durban-Watson values for the regression models used to test hypothesis 2

were all between 1.84 and 1.90 suggesting that residuals in each model are independent.

Scatterplots of residuals were examined for possible heteroscedasticity and linearity. No










indication of heteroscedasticity or non-linearity was evident in any of the models suggesting that

the assumptions of homoscedasticity and linearity in the data have been met.

In order to test Hypothesis 3 regarding moderating effects, methods reported by Baron and

Kenny (1986) were utilized. They suggest running regression analyses and using the resulting b-

weights rather than correlation coefficients to test for differences because "regression

coefficients are not affected by differences in the variances of the independent variable or

differences in measurement error of the dependent variable." (p. 1 175). Therefore, two separate

hierarchical regression analyses were run for each appraisal rating (i.e., substantiation agreement,

level of perceived risk, and recommended intervention.

In the first analysis of each rating, the mean score on the African American

(Substantiation) index was the dependent variable and social desirability and multicultural

competence were entered hierarchically as predictors. The resulting b-weight of multicultural

competence on the African American Substantiation index was b = .016. The second analysis

included the European American (Substantiation) index as the dependent variable and social

desirability and multicultural competence as predictors. The resulting b-weight of multicultural

competence on the European American index was b = .021. A z test was performed to test for a

difference between the two b-weights of the independent variable (MCI full-scale score). A

significant difference in the b-weights would be evidence of the moderating effects of race of the

child presented in the vignette on multicultural competence as a predictor of appraisal ratings.

As recommended by Paternoster, Brame, Mazerolle, and Piquero (1998) a more conservative

denominator was used in the equation for the : statistic, thereby reducing the chances of rejecting

the null when there truly is no difference between the two b-weights (see equation below).
bi -b
Z=----~--
j Jhl + SEb/I









Table 3-4 (below) shows the unstandardized coefficients resulting from the hierarchical

regression analyses of the three appraisal ratings on multicultural competence as well as the z-

statistic for the test of moderation of race of the child upon the relationship between multicultural

competence and appraisal ratings. For the test of the moderation effects of race of the child on

the relationship between multicultural competence and ratings of substantiation agreement, the

resulting z-statistic (z = .33) is not significant. These findings do not provide support for the

hypothesis that the race of the child moderates the relationship between multicultural

competence and levels of substantiation agreement. This procedure was repeated to examine

ratings regarding level of risk and recommended intervention. The resulting z-statistics for level

of perceived risk (z = .09) and recommended intervention (z = .30) were non-significant as well,

suggesting that race of the child depicted in the vignette did not moderate the relationship

between multicultural competence and appraisal ratings. In addition, when the rating indices

were regressed upon multicultural competence separately for the European American version

and the African American version, multicultural competence was found to only be a significant

predictor of ratings of Intervention when a European American child was depicted in the

vignette.

The lack of a significant relationship between multicultural competence and the race-

specific rating indices, and subsequent lack of support for hypothesis 3 could be due to the

smaller sample size garnered in this study. Our initial thought that future studies utilizing larger

sample sizes may be successful in finding support for the moderation effect described in

hypothesis 3. Using G*Power, the effect sizes for the African American and European American

Intervention indices were entered to estimate the sample size needed to find a significant

relationship in a regression model with two predictors at a power of 0.80. The results suggest










that a sample size of480 (total) participants should provide the power needed to find a

significant relationship between multicultural competence and both the African American and

European American Intervention indices. However, when the product term for the interaction

effect of multicultural competence and race of the child is entered into the regression equation,

the resulting (orthogonalized) effect size of the interaction is extremely small by Cohen' s

standards (r2 added = .002) and G*Power estimates that nearly 4000 participants would be

needed to find a significant moderation effect of race of the child on the relationship between

multicultural competence and appraisal ratings of intervention! Thus, it seems that a true

moderation effect of race of the child on the relationship between multicultural competence and

ratings of (in this case) Intervention is highly unlikely.



Table 3-1. Means, standard deviations, and correlations among variables (N = 220)
Variable 1 2 3 4 5

1. MCI Total Score .295 .132 .138* .167*

2. Social Desirability -.080 .015 -.037

3. Combined Substantiation Index .657** .524**

4. Combined Risk Index .634**

M 115.87 16.18 3.49 5.64 4.53

SD 12.10 5.16 .33 .70 .72

* p<.05. ** p<.01











Table 3-2. Results of MANOVA with combined rating indices as DV' s and version (race of the
child depicted) as IV
Index SSx df F p SST

Combined Substantiation 1.624 1 .930 .168 382.450

Combined Risk 35.312 1 4.548* .017* 1727.977

Combined Intervention 27.522 1 3.312* .035* 1838.995

p <.05.


Table 3-3. Summary of hierarchical regression analysis of appraisal rating ind
multicultural competence
Index B SE P t p

Combined Substantiation .019 .008 .170 2.43* .016

Constant 12.338 .850 -- ---

Combined Risk .034 .016 .146 2.08* .039

Constant 18.889 1.819 -- ---

Combined Intervention .047 .017 .195 2.79** .006

Constant 12.857 1.861 --

* p<.05. ** p<.01


lices on





Table 3-4. Summary of hierarchical regression analyses of multicultural competence on african
american and european american appraisal rating indices and Z-statistics of race of
the child as a moderator
Index B SEE B t Df Z

African American Substantiation .016 .010 .153 1.544 108

European American Substantiation .021 .011 .182 1.832 110


Test of Moderation

African American Risk

European American Risk

Test of Moderation

African American Intervention

European American Intervention

Test of Moderation


0.33


.031

.034


.020

.026


.152

.130


1.532

1.310


0.09


.040

.050


.022

.025


.182

.197


1.828

1.982


0.30









CHAPTER 4
DISCUSSION

This study was to serve as an initial investigation into the relationships between race of the

child, multicultural competence, and decision-making bias in the child protective services

context. Previous successful use of short case vignettes are presented in the literature (Shapira &

Benbenishty, 1993; Giovanna & Bacerra, 1979; Roberts, 2002), however no evidence exists in

the literature of using these short vignettes to draw out biases based on race. This study aimed to

further develop the use of case vignettes in child welfare research. By having participants

randomly assigned to complete either a version with the target vignettes depicting European

American children or a version using the identical target items depicting African American

children, differences in mean ratings of the vignettes that could be directly attributed to the

experimental manipulation of race of the child could be investigated.

In the sample of data used in the current study, it appears that when making decisions

about perceived levels of risk and recommended levels of intervention in the child protective

context, the race of the child presented in the vignette did make a difference. When the child

presented in the vignette was African American, ratings of risk and recommended intervention

were higher than when the child presented in the vignette was European American. This finding

provides experimental support to previous investigations using field data (Eckenrode, Powers,

Doris, Munsch, and Bolger, 1988; and Lamb, 1979; NIS-3, 1993) and qualitative data (Chibnall,

Dutch, Jones-Harden, Brown, and Gourdine, 2003) that suggest the racial biases do impact child protective

service provision.

As an initial investigation into the relationship between multicultural competence and

decisions made in the context of child protective service investigations, this study offers support

for the consideration of multicultural competence as a valid construct in examining such









decisions. It is unclear what aspect of the decision-making process multicultural competence is

tapping and it is uncertain if the MCI is the best measure to use in such investigations. Answers

to these questions may be useful in guiding future research.

Although Hypothesis 2 was supported by the results of the regression analyses, neither

theory nor past research suggests an explanation for the positive relationship between

multicultural competence and severity of appraisals. Perhaps those with more multicultural

competence are more globally aware of the inherent dangers presented to a child in any given

situation compared to those who are less multiculturally competent. Or perhaps the distribution

of demographic variables in this sample had an unexpected impact on the results.

Current Limitations

Currently there are various limitations to the study. For one, multicultural competence was

measured using a self-report instrument. This is a limitation because a self-report measure is

designed to capture perceived levels of multicultural competence and it is unknown how well

perceived multicultural competence reflects how one' s level of multicultural competence may

impact his/her performance. Unfortunately, self-report measures are the only measures of

multicultural competence currently available.

Undergraduate students made up the majority (71%) of the sample in this study. Initially,

this is troublesome on a research design level. Perhaps the variability in scores on the outcome

variables is due to utilizing a sample that has not had any training or experience in making

decisions related to child welfare investigations. Nonetheless, outcome measures of risk and

intervention were still significantly higher, at the p < .05 level, for vignettes depicting African

American children than for vignettes depicting European American children, even when

undergraduate vs. graduate/professional status was controlled for by entering it as a covariate.

The same level of support was found for hypothesis 2 when controlling for undergraduate vs.










graduate/professional status. However, the undergraduate status of the maj ority of the sample

may have an impact on how well the sample represents the population of interest.

Undergraduate students in the psychology research pool may or may not be studying psychology

or any related social science field. The inclusion of students who are not social science maj ors

might decrease the probability that the maj ority of the sample has any interest in, or any

awareness of, making child welfare-related decisions. Thus, the question could be raised as to

whether the results from the current sample are representative of the population of people who

might be expected to make child welfare-related decisions.

The exploratory nature of the study limits both the scope and focus of questions asked.

Moreover, the model used to examining child protective investigation outcomes is

underspecified. Exploratory analyses on the data may lead to suggestions for a more specified

model to be tested in future research. But this is beyond the scope of the current paper.

Despite the apparent success of using these case vignettes in this study, the ecological

validity of making determinations of actual child abuse/neglect, risk, and intervention based

upon the case vignettes is open to debate. Although similar case vignettes have been utilized in

past research (Giovanna & Bacerra, 1979; Shapira & Benbenishty, 1993), feedback from

participants suggested that in many instances, a determination would not be based upon such a

limited amount of information.

The online nature of the study may pose another limitation. Although both the survey and

the resulting data were housed on a secure server, there were no safeguards in place to ensure

that each participant only completed the survey one time. A participant could, if he/she chose,

complete the survey numerous times and if such a person had a specific agenda in mind, this

could alter the results. There is no indication and the researcher has no reason to believe that this










posed a true limitation of this study. The length of time necessary to complete the survey and

the lack of compensation for participants make numerous submissions of survey responses

unlikely. It is also possible that a participant could change his/her answers to any of the appraisal

rating questions after he/she became aware of or completed the section measuring multicultural

competence. However, due to the between subjects nature of the design, it seems unlikely that a

participant would be aware of which specific items to change.

A potential limitation of the data itself may be the presence of outliers. A brief look at

scatterplots of the data suggests that there are low ratings on each of the indices (the

substantiation index in particular) that may be considered outliers and may be having a negative

effect on our data analyses. Currently, data from all participants who completed the entire

survey were included in our analyses. A closer examination of the data utilizing casewise

diagnostics may offer substantial statistical reasons to eliminate these scores from our analyses.

Specifically, a close look at the standardized residuals of our regression analyses and utilization

of Cook' s Distance values can offer information regarding the impact these potential outliers

may be having on our analyses. A quick look at the scatterplots suggest that these outliers may

have high discrepancy and moderate leverage with which to disrupt our results.

A final limitation to be discussed in this paper is a limitation related to the specific

measure of multicultural competence used in this study. Although the MCI has been reported to

be a valid instrument in previous literature, because the MCI had to be used in it' s original form

and could not be altered to reflect language relevant to social workers, it is possible that a

measure of multicultural competence developed specifically in the social work or child welfare

field, or a measure adapted to reflect language relevant to social workers might be more useful

and tap into competencies which could explain more variance in our models of analyses.









On a related theoretical level, a question of what is truly being measured by the MCI and

MCSDS must be raised. Multicultural Competence is a somewhat ambiguous concept and it is

therefore difficult to ascertain whether the MCI is accurately capturing aspects of working with

people from diverse cultures that are important in making decisions about child abuse/neglect.

Other constructs may more accurately reflect culturally-related aspects of child welfare

investigation. For example, Racial Identity is a construct that has received moderate coverage in

the literature. Utilizing the concept of racial identity may allow for more flexibility when

examining the role of ethnicity in child welfare-related decisions. White Racial Identity has been

found to clearly be differentially related to racism in such as way that different racial identity

schema profiles relate to racism in different ways (Carter, Helms, & Juby, 2004). Perhaps

examining racial bias in decision making could benefit from utilizing the construct of racial

identity as opposed to multicultural competence.

The method of controlling for social desirability in this study is also potentially

problematic. Although a standard of measure social desirability was used in this study, the

appropriateness of the MCSDS is unclear. One might argue that social desirability is adaptive to

different generations. Thus, the measurement of social desirability must be continually redefined

and recreated; which has not been sufficiently illustrated in the literature. Sodowsky (1996)

argued the general construct of social desirability might not be useful when studying

multicultural competence. Perhaps there is a different type of social desirability in relation to

multicultural competence.

Recommendations for Future Research

The results of this study lend promise to the value of future research in this area. Utilizing

the case vignettes in an experimental design fashion appeared to be successful. Thus, future









research may benefit from utilizing case vignettes in such a way; particularly to investigate

differential decisions based on specific demographic characteristics presented in the vignettes.

It appears that multicultural competence may be a valid construct in examining decisions

made in the context of child abuse/neglect services provision. However, as noted above, it is

recommended that future studies examining this issue utilize either an adapted measure that

reflects language relevant to social workers or a measure developed specifically based upon

multicultural competencies related to child welfare services. At this time, it does not appear that

measures exist that would be in line with this recommendation. Future research may also benefit

from utilizing different theoretical constructs in explaining the variance in scores of child welfare

investigation outcome measures.

Future research in this area may also benefit from more effective recruitment methods

aimed at accessing a larger sample of graduate students of social work and current child welfare

caseworkers. There appears to be a different pattern of results among these groups based upon

initial eyeballing of the appraisal rating means from these groups. Higher numbers of

participants could also increase the power of future studies to find significant effects.

A final suggestion for future research is a direct result of feedback from participants of this

study and other professionals who were consulted while the study was active. It is suggested that

racial biases and multicultural competence may play a larger role in the treatment phase of the

child welfare process. Some professionals noted that abuse/neglect investigation decisions are

driven more by state and federal laws than by individual judgments and decisions. Perhaps

individual judgments and decisions, and therefore racial bias and multicultural competence, have

a greater impact when it comes to treatment and resource allocation decisions than substantiation

and risk decisions. This idea was partially supported by the results of this study in considering









that significant impacts of multicultural competence and race were seen more when it came to

appraisals of recommended intervention than substantiation and risk appraisals. Future research

might benefit from a focus on the treatment and resource allocation phase of the child welfare

service process rather than the investigative and substantiation phase of the process.

Implications


Due to the exploratory nature of this study and the first step this study has taken to

investigate the relationships between race of the child, multicultural competence of the worker,

and decisions made in the context of child welfare service provision, any implications discussed

are tentative and should be further explored with future research.

One implication of the findings of this study is that racial bias in making decisions about

child abuse/neglect appears to be a robust effect. When case vignettes presented on a computer

screen can elicit differential decisions based upon the race of the child presented in the vignette,

it may be reasonable to think that real-world situations in which a worker is confronted with

racial differences may have a similar, if not greater impact on his/her perceptions, thereby

leading to bias in the provision of services.

Regarding the use of the multicultural competence construct, the results of this study seem

to provide the impetus to apply this construct to future investigations of child welfare and other

social work processes. Newly created measures incorporating specific competencies relevant to

social work are missing from the current literature and the results of this study may provide

additional support for the viability and usefulness of creating such measures.










APPENDIX
FORMS AND MEASURES

Informed Consent


Protocol Title: Child Protective Decisions Based on Limited Information

Please read this consent document carefully before you decide to participate in this study.


Purpose of the research study:

The purpose of this study is to examine decisions made in the child protective services context
based on limited amounts of information.

What you will be asked to do in this study:

You will be asked to read a series of twenty statements from separate case report summaries and
make three separate ratings based on the information provided in each statement. Once you have
read all twenty statements and provided ratings for each, you will be asked to respond to a
number of items based on the extent to which you believe the item is a reflection of you. Some
of the items ask you to respond with either "true" or "false" and some of the items ask you to
respond by rating how accurate the item is as it relates to you as a professional on a 4-point scale.

Time required:

Approximately 45 minutes

Risks and Benefits:

There are no anticipated risks to participating in this study. Anticipated benefits are minimal,
but you may gain an increased awareness of your own decision making style in instances of
limited information.

Compensation:

You will not be compensated for your participation in this study.





Confidentiality:

All information provided in this study will be completely anonymous. You will not be asked to
provide any identifying information, nor will your location or internet address be tracked while
you are participating in the study. No attempts will be made to identify you in any way.











Voluntary participation:


Participation in this study is completely voluntary. There is no penalty for not participating.

Right to withdrawal from the study:

You have the right to discontinue participation in the study at any time, for any reason, without
consequence.

Whom to contact if you have questions about the study:

Michael N. Ghali, Graduate Student, Department of Psychology (Counseling) University of
Florida, Gainesville, FL 32611

Whom to contact about your rights as a research participant in the study:

UFIRB Office, Box 112250, University of Florida, Gainesville, FL 32611-2250;

ph 392-0433.



Agreement:

By checking the box below, you agree that you have read the procedure described above and
voluntarily agree to participate in the study. If you do not agree, please select the "Quit" button
below and this window in your web browser will be closed. If you do agree to participate, please
check the box marked "I agree" below, print this page for your records, and select the
"Continue" button.











Vignettes: Form A


The following statements were taken from child abuse or neglect report summaries.
Following each statement, you will be asked to respond to three items regarding your initial
appraisal of the situation. Please read each statement carefully and respond to each item to
the best of your professional ability based on the information provided.

(Bold items will be used as between subj ects items)

1. An 8 year old European American girl was brought to the hospital three times for
being underweight. Each time the child gained weight during the hospital stay.
Upon return home, the child lost weight again.

2. A 12 year old Hispanic American boy was struck by his parent with a wooden stick. The
child suffered a concussion.

3. The parents of an 8 year old European American boy ignored their child's complaint of
an earache and chronic ear drainage. The child was found to have a serious infection and
damage to the inner ear.

4. A 12 year old African American girl was regularly left home alone all night.

5. The parents of an Asian American child regularly fail to feed their 8 year old for periods
of at least 24 hours. The young girl was hospitalized for six weeks for being seriously
underweight.

6. The parents of a 12 year old European American girl have repeatedly failed to keep
medical appointments for their child. The child has a congenital heart defect.

7. The parents of a 12 year old Asian American boy regularly leave their child with their
neighbors, without knowing who would assume responsibility and be in charge. On one
occasion, the child was seen at a store four blocks away.

8. The parents of a 12 year old Hispanic American girl have not given their child
medication prescribed by a physician. The child has a throat infection.

9. The parents of an 8 year old Hispanic American boy regularly leave the child alone inside
the house after dark. Often, they do not return until midnight. On one occasion, the child
started a small fire.

10. A 12 year old European American child is not given any healthcare by her parents. The
child complains of physical ailments.

11. The parent of a 12 year old European American girl hit the child in the face,
striking her with the fist. The child suffered a black eye and a cut lip.











12. The parent of a 12 year old African American boy banged the child against the wall while
shaking him by the shoulders. The child suffered large bruises.

13. The parents of an 8 year old Hispanic American child fail to prepare regular meals for
her. The child often has to fix her own supper.

14. The parents usually punish their 12 year old European American boy by spanking him
with a leather strap, leaving red marks on the child's skin.

15. On several occasions, an 8 year old African American boy was left with his grandmother
for periods of time up to ten days without providing any means of contacting the parents.
The child became ill during the most recent instance.

16. An 8 year old European American boy was burned on the chest and buttocks with a
cigarette by a parent. The child has second degree burns.

17. A 12 year old Asian American boy is regularly left alone outside during the day until
almost dark. Neighbors have spotted the child wandering five blocks from home.

18. The parents of an 8 year old European American girl feed only milk to the child. The
child has an iron deficiency.

19. An 8 year old Asian American girl was immersed in a tub of hot water. The child
suffered second-degree burns.

20. The parents of an 8 year old African American boy regularly leave the child alone inside
the house during the day. Often, they did not return home until almost dark. On one
occasion, the parents returned home to find broken glass on the floor.










Vignettes: Form B


The following statements were taken from child abuse or neglect report summaries.
Following each statement, you will be asked to respond to three items regarding your initial
appraisal of the situation. Please read each statement carefully and respond to each item to
the best of your professional ability based on the information provided.

(Bold items will be used as between subj ects items)

1. An 8 year old African American girl was brought to the hospital three times for
being underweight. Each time the child gained weight during the hospital stay.
Upon return home, the child lost weight again.

2. A 12 year old Hispanic American boy was struck by his parent with a wooden stick. The
child suffered a concussion.

3. The parents of an 8 year old European American boy ignored their child's complaint of
an earache and chronic ear drainage. The child was found to have a serious infection and
damage to the inner ear.

4. A 12 year old African American girl was regularly left home alone all night.

5. The parents of an Asian American child regularly fail to feed their 8 year old for periods
of at least 24 hours. The young girl was hospitalized for six weeks for being seriously
underweight.

6. The parents of a 12 year old African American girl have repeatedly failed to keep
medical appointments for their child. The child has a congenital heart defect.

7. The parents of a 12 year old Asian American boy regularly leave their child with their
neighbors, without knowing who would assume responsibility and be in charge. On one
occasion, the child was seen at a store four blocks away.

8. The parents of a 12 year old Hispanic American girl have not given their child
medication prescribed by a physician. The child has a throat infection.

9. The parents of an 8 year old Hispanic American boy regularly leave the child alone inside
the house after dark. Often, they do not return until midnight. On one occasion, the child
started a small fire.

10. A 12 year old European American child is not given any healthcare by her parents. The
child complains of physical ailments.

11. The parent of a 12 year old African American girl hit the child in the face, striking
her with the fist. The child suffered a black eye and a cut lip.









12. The parent of a 12 year old African American boy banged the child against the wall while
shaking him by the shoulders. The child suffered large bruises.


13. The parents of an 8 year old Hispanic American child fail to prepare regular meals for
her. The child often has to fix her own supper.

14. The parents usually punish their 12 year old European American boy by spanking him
with a leather strap, leaving red marks on the child's skin.

15. On several occasions, an 8 year old African American boy was left with his grandmother
for periods of time up to ten days without providing any means of contacting the parents.
The child became ill during the most recent instance.

16. An 8 year old African American boy was burned on the chest and buttocks with a
cigarette by a parent. The child has second degree burns.

17. A 12 year old Asian American boy is regularly left alone outside during the day until
almost dark. Neighbors have spotted the child wandering five blocks from home.

18. The parents of an 8 year old European American girl feed only milk to the child. The
child has an iron deficiency.

19. An 8 year old Asian American girl was immersed in a tub of hot water. The child
suffered second-degree burns.

20. The parents of an 8 year old African American boy regularly leave the child alone inside
the house during the day. Often, they did not return home until almost dark. On one
occasion, the parents returned home to find broken glass on the floor.










Appraisal Rating Questions


To what extent to do agree that the information provided in the preceding case summary is
evidence of actual child abuse or neglect?

1. Strongly disagree

2. Disagree

3. Agree

4. Strongly agree





In your best professional opinion, based on information given in the preceding case summary,
please rate the degree to which you believe this child to be at risk:

1. No risk

2. Low risk

3. Some risk

4. Intermediate level of risk

5. High risk

6. Very high risk

7. Extreme risk





In your best professional opinion, based on the information in the preceding case summary,
please choose ONE of the following options for intervention below:

1. No further intervention.

2. Indirect intervention through professionals already in contact with the family.

3. Direct intervention with this family without removing the child from the home.

4. Direct intervention with the family, place the child in "day placement" until the evening.











5. Removal of the child from the home for an extended period of time, with the parents'
agreement, while continuing to work with the family.

6. Removal of the child from the home for an extended period of time, even without the
parents' agreement (with court order, if necessary), while continuing to work with the family.










Multicultural Counseling Inventory (MCI)


The following statements cover counselor practices in multicultural counseling. Indicate how
accurately each statement describes you as a counselor, psychologist, or student in a mental
health training program when working in a multicultural counseling situation. Give ratings that
you actually believe to be true rather than those you wish were true.

The scale ranges from 1 (very inaccurate) to 4 (very accurate). The scale indicates the following:

1 very inaccurate
2 somewhat inaccurate
3 somewhat accurate
4 very accurate





When working, with minority clients...

1. I perceive that my race causes the clients to mistrust me.

2. I have feelings of overcompensation, oversolicitation, and guilt that I do not have when
working with maj ority clients.

3. I am confident that my conceptualization of client problems does not consist of
stereotypes and value-oriented biases.

4. I find that differences between my worldviews and those of the clients impede the
counseling process.

5. I have difficulties communicating with clients who use a perceptual, reasoning, or
decision-making style that is different from mine.

6. I include the facts of age, gender roles, and socioeconomic status in my understanding of
different minority cultures.

7. I use innovating concepts and treatment methods.

8. I manifest an outlook on life that is best described as "world-minded" or pluralistic.

9. I examine my own cultural biases.

10. I tend to compare client behaviors with those of majority group members.

11. I keep in mind research findings about minority clients' preferences in counseling.










12. I know what are the changing practices, views, and interests of people at the present time.

13. I consider the range of behaviors, values, and individual differences within a minority
group.

14. I make referrals or seek consultations based on the clients' minority identity
development.

15. I feel my confidence is shaken by the self-examination of my personal limitations.

16. I monitor and correct my defensiveness (e. g., anxiety, denial, anger, fear, minimizing,
overconfi dence).

17. I apply the sociopolitical history of the clients' respective minority groups to understand
them better.

18. I am successful at seeing 50% of the clients more than once, not including intake.

19. I experience discomfort because of the clients' different physical appearance, color,
dress, or socioeconomic status.

20. I am able to quickly recognize and recover from cultural mistakes or misunderstandings.

21. I use several methods of assessment (including free response questions, observations, and
varied sources of information and excluding standardized tests).

22. I have experience at solving problems in unfamiliar settings.

23. I learn about clients' different ways of acculturation to the dominant society to
understand the clients better.

24. I understand my own philosophical preferences.

25. I have a working understanding of certain cultures (including African American, Native
American, Hispanic, Asian American, new Third World immigrants, and international
students.

26. I am able to distinguish between those who need brief, problem-solving, structured
therapy and those who need long-term, process-oriented, unstructured therapy.

27. When working with international students or immigrants, I understand the importance of
the legalities of visa, passport, green card, and naturalization.

Evaluate the degree to which the following, multicultural statements can be applied to you.

28. My professional or collegial interactions with minority individuals are extensive.











29. In the past year, I have had a 50% increase in my multicultural case load.

30. I enj oy multicultural interactions as much as interactions with people of my own color.

31. I am involved in advocacy efforts against institutional barriers in mental health services
for minority clients (e. g., lack of bilingual staff, multiculturally skilled counselors,
minority professional leadership, and outpatient counseling facilities).

32. I am familiar with nonstandard English.

33. My life experiences with minority individuals are extensive (e. g., via ethnically
integrated neighborhoods, marriage, and friendship).

34. In order to be able to work with minority clients, I frequently seek consultation with
multicultural experts and attend multicultural workshops or training sessions.


When working; with all clients,...

35. I am effective at crisis interventions (e. g., suicide attempt, tragedy, broken relationship).

36. I use varied counseling techniques and skills.

37. I am able to be concise and to the point when reflecting, clarifying, and probing.

38. I am comfortable with exploring sexual issues.

39. I am skilled at getting a client to be specific in defining and clarifying problems.

40. I make my nonverbal and verbal responses congruent.









The Marlowe-Crowne Social Desirability Scale

Personal Reaction Inventory

Listed below are a number of statements cancerning personal attitudes and traits. Read each item
and decide whether the statement is True or False as it pertains to you personally.

1. Before voting I thoroughly investigate the qualifications of all the candidates. (T)

2. I never hesitate to go out of my way to help someone in trouble. (T)

3. It is sometimes hard for me to go on with my work, if I am not encouraged. (F)

4. I have never intensely disliked anyone. (T)

5. On occasion I have had doubts about my ability to succeed in life. (F)

6. I sometimes feel resentful when I don't get my way. (F)

7. I am always careful about my manner of dress. (T)

8. My table manners at home are as good as when I eat out in a restaurant. (T)

9. If I could get into a movie without paying and be sure I was not seen, I would probably do it.
(F)

10. On a few occasions, I have given up doing something because I thought too little of my
ability. (F)

11. I like to gossip at times. (F)

12. There have been times when I felt like rebelling against people in authority even though I
knew they were right. (F)

13. No matter who I'm talking to, I'm always a good listener. (T)

14. I can remember "playing sick" to get out of something. (F)

15. There have been occasions when I took advantage of someone. (F)

16. I'm always willing to admit it when I make a mistake. (T)

17. I always try to practice what I preach. (T)


18. I don't find it particularly difficult to get along with loud-mouthed, obnoxious people. (T)











19. I sometimes try to get even rather than forgive and forget. (F)

20. When I don't know something I don't at all mind admitting it. (T)

21. I am always courteous, even to people who are disagreeable. (T)

22. At times I have really insisted on having things my own way. (F)

23. There have been occasions when I felt like smashing things. (F)

24. I would never think of letting someone else be punished for my wrongdoings. (T)

25. I never resent being asked to return a favor. (T)

26. I have never been irked when people expressed ideas very different from my own. (T)

27. I never make a long trip without checking the safety of my car. (T)

28. There have been times when I was quite jealous of the good fortune of others. (F)

29. I have almost never felt the urge to tell someone off. (T)

30. I am sometimes irritated by people who ask favors of me. (F)

31. I have never felt that I was punished without cause. (T)

32. I sometimes think when people have a mistortune they only got what they deserved. (F)

33. I have never deliberately said something that hurt someone's feelings. (T)













By now, you may be curious about the questions being addressed in this study. This page
will serve to disclose to you the specific aims of this study, ask you to acknowledge that you
have been made aware of and understand the aims of this research, and direct you to a page of
links that provide additional information about the topic on which this research is focused.

The aim of this study was to examine the relationship between multicultural competence
and decision making in the context of child protective services. As populations of countries
across the globe become more racially/ethnically diverse, multicultural competence is becoming
increasingly important in all areas of human service interaction. In the present study, we were
interested in examining the relationship between multicultural competence and the assessment
and intervention recommendations made by individuals who work with children and families
from diverse racial and ethnic backgrounds. We hope to draw on this research to help develop
and improve training programs aimed at enhancing multicultural competence among
caseworkers and others who work with children and families who may become involved in the
child welfare system.

Fulfilling the aims of this research requires that the exact nature of the research aims, and
focus of the study, not be disclosed until after participants have completed the survey. Your
answers to all survey items remain completely anonymous and individual responses cannot be
linked to you in any way.

In a moment, you will be directed to a page that includes various links to websites which
can provide information, training, and further links regarding multicultural competence, as well
as a short questionnaire regarding your experience with participating in this study. Utilization
(or non-utilization) of the links provided and feedback questions cannot be linked to answers to
previous items in this study and remain anonymous.

Please check the box or sign below to signify that you have read this form and
understand the aims of this study and the need for this information to be provided you after you
have completed the study.


Whom to contact if you have questions about the study:

The principal investigator for this study can be contacted using the following
information:
Michael Ghali M.A.
Doctoral Student in Counseling Psychology
University of Florida
Department of Psychology
Room 114, P.O. Box 112250
Gainesville, FL 32601
Phone: (352) 392-0601
Fax: (352) 392-7985


Debriefing









Email: ghali@ufl.edu


Additional questions can be directed to:


UFIRB Office, Box 112250, University of Florida, Gainesville, FL 32611-2250; ph 392-


0433.


I have read the above debriefing form and have been made aware of the full purpose of this
research. Any attempts at deception have been disclosed and explained to me and I am aware of
whom to contact should I have any questions or concerns regarding the research in which I have
participated.


Signature of participant


Date


Please initial if you would like your data to NOT be used in this study:










Feedback Form


Please share your feedback regarding your experience participating in this research study.
Your answers are complete anonymous, so please be honest and forward with your responses.
Your feedback can help in the design and administration of improved research studies in the
future. Completion of this feedback form is optional and will in no way effect your responses to
previous items in this study.


1. How would your rate your overall experience as a participant in this research study?
a. Excellent
b. Good
c. Okay
d. Poor

2. The time required to participate in this study was:
a. Much too long
b. Somewhat long
c. Adequate
d. Too short

3. Despite the original description of the study, the true aims of the study were:
a. Absolutely obvious
b. Not too hard to figure out
c. Not clear
d. I had no idea

4. The instructions for participating in the study were:
a. Very clear and easy to follow
b. Somewhat clear and easy to follow
c. At times hard to follow
d. Not at all clear or easy to follow

5. Disclosing all of the true aims of the study after you completed your participation was:
a. Completely justified
b. Probably unnecessary, but it didn't bother me
c. Probably necessary, but it bothered me that I was initially deceived
d. Completely unjustified

6. Please share any other comments you may have about your experience in participating in
this study:









REFERENCES


Bellini, J. (2003). Counselors' multicultural competencies and vocational rehabilitation
outcomes in the context of counselor-client racial similarity and difference.
Rehabilitation Counseling Bulletin, 46(3), 164-173.

Bernatovicz, F. (n.d.). Retention of child welfare caseworkers: A report National Child Welfare
Resource Center for Organizational Improvement. Accessed 7/19/06 at:
www. muski e.usm. maine. edu/helpki ds/pub stext/retenti on.html .

Carter, R. T., Helms, J. E., & Juby, H.L. (2004). The relationship between racism and racial
identity for white americans: A profile analysis. Multicultural Counseling and
Development, 32, 2-17.

Chibnall, Dutch, Jones-Harden, Brown, & Gourdine (2003). Children of color in the child
welfare system: Perspectives from the child welfare community. National
Clearninghouse of ChildAbuse andNeglectlnfornzation. Accessed 11/14/05 at"
http ://nccanch.acf.hhs.gov/pubs/otherpubs/chilrnidxcfm.

Child Welfare League of America. (1999). Minimum education required by state child welfare
agencies, percent, by degree type, 1998. State Child Welfare Agency Survey.

Cohen, E. (2003). Framework for culturally competent decision-making in child welfare. Child
Welfare. 82(2), 143-155.

Constantine, M. G., Gloria, A., & Ladany, N. (2002). The factor structure underlying three self-
report multicultural counseling competence scales. CulturalDiversity and Ethnic
Minority Psychology, 8(4), 334-345.

Constantine, M.G., & Ladany, N. (2001). New visions for defining and assessing multicultural
counseling competence. Handbook ofnaulticultural counseling (2nd Ed.), (pp.482-498).
Sage Publications: Thousand Oaks, CA.

Constantine, M. G., & Ladany, N. (2000). Self-report multicultural competence scales: Their
relation to social desirability attitudes and multicultural case conceputalization ability.
Journal of Counseling Psychology, 47(2), 155-164.

Crowne, D., & Marlowe, D. (1960). A new scale of social desirability independent of
psychopathology. Journal of Consulting Psychology, 24, 349-354.

D'Andrea, M., Daniels, J., & Heck, R. (1991). Evaluating the impact of multicultural
counseling training. Journal of Counseling and Development, 70, 143-150.

Davidson-Arad, B. (2001). Predicted changes in children's quality of life in decisions regarding
the removal of children at risk from their homes. Children and Youth Services Review,
23(2), 127-143.










Eckenrode, J., Powers, J., Doris, J., Munsch, J., & Bolger, N. (1988). Substantiation of child
abuse and neglect reports. Journal of Consulting and Clinical Psychology, 56(1, 9- 16.

Field, A. (2000). Discovering statistics using SPSS for Windows. London: Sage Publications.

Giovannoni, J., & Becerra, R. (1979). Defining child abuse. New York: The Free Press.

Harran, E. (2002). Barriers to effective child protection in a multicultural society. Child Abuse
Review, 11, 411-414.

Hong, G., & Hong, L. (1991). Comparative perspectives on child abuse and neglect: Chinese
versus hispanics and whites. Child Welfare, 70(4), 463-475.

LaFromboise, T., Coleman, H., & Hernandez, A. (1991). Development and factor structure of
the cross-cultural counseling inventory--revised. Professional Psychology: Research
and Practice. 22(5), 380-388.

Lamb, K. L. (1979). Outcomes of child abuse complaints: Caseworkers as predictors. (Doctoral
Dissertation, University of Arizona, 1979). Dissertation Abstracts hIternational, 40 (2-
A), pp. 1109-1110.

McPhatter, A. (1997). Cultural competence in child welfare: What is it? How do we achieve it?
What happens without it? Child Welfare, 76(1, 255-288.

Paternoster, R., Brame, R., Mazerolle, P., & Piquero, A. (1998). Using the correct statistical test
for the equality of regression coefficients. Criminology, 36(4), 857-866.

Pierce, R. L. & Pierce, L. H., (1996). Moving toward cultural competence in the child welfare
sy stem. Children and Youth Services, 18(8), 713-73 1.

Ponterotto, J., & Casas, M. (1991). Handbook ofracial ethnic minority counseling research.
Springfield, IL: Charles C. Thomas.

Ponterotto, J., Gretchen, D., Utsey, S., Rieger, B., & Austin, R. (2002). A revision of the
multicultural counseling awareness scale. Multicultural Counseling and Development,
30, 153-180.

Pope-Davis, D., & Dings, J. (1995). The assessment of multicultural counseling competencies.
In Joseph G. Ponterotto & Manuel J. Casas (Eds.) Handbook ofnaulticultural counseling.
Thousand Oaks, CA: Sage Publishing.

Pope-Davis, D., & Dings, J. (1994). An empirical comparison of two self-report multicultural
counseling competency inventories. Measurement and Evaheation in Counseling and
Development, 27(2), 93-102.










Pope-Davis, D., Liu, W., & Nevitt, J. (2000). The development and initial validation of the
multicultural environmental inventory: A preliminary investigation. Cultural Diversity
and Ethnic M~inority Psychology, 6(1, 57-64.

Pope-Davis, D. B., Liu, W. M., Toporek, R. L., & Brittan-Powell, C. S. (2001). What's missing
from multicultural competency research: Review, introspection, and recommendations.
Cultural Diversity and Ethnic M~inority Psychology. 7(2), 121-13 8.

Roberts, D. (2002). Manciel~l~(lbonds: The color of child welfare. New York, NY: Basic Civitas
Books .

Shapira, M., & Benbenishty, R., (1993). Modeling judgments and decisions in cases of child
abuse and neglect. Social Work Research and Abstracts, 29(2), 14-19.

Sodowsky, G. R. (1996). The multicultural counseling inventory: Validity and applications in
multicultural training. In G. R. Sodowsky and J.C. Impara (Eds.), M~ulticultural
assessment in counseling and clinical psychology (pp. 283 324). Lincoln, NE: Buros
Institute of Mental Measurements.

Sodowsky, G, Kuo-Jackson, P., Richardson, M., & Corey, A. (1998). Correlates of self-
reported multicultural competencies: Counselor multicultural social desirability, race,
social inadequacy, locus of control racial ideology, and multicultural training. Journal of
Counseling Psychology. 45 (3), 256-264.

Sodowsky, G., Taffe, R., Gutkin, T., & Wise, S. (1994). Development of the multicultural
counseling inventory: A self-report measure of multicultural competencies. Journal of
Counseling Psychology, 41(2), 137-148.

Sue, D. W., Bernier, J. E., Durran, A., Feinberg, L., Pedersen, P., Smith, E. J., Vasquez-Nutall,
E. (1982). Position paper: Cross-cultural counseling competencies. The Counseling
Psychologist, 10(2), 45-52.

Sue, D. W., Arredondo, P, & McDavis, R. (1992). Multicultural counseling competencies and
standards: A Call to the Profession. Journal of2~ulticultural Counseling and
Development. 20(2), 64-68.

Taussig, H. (2002). Risk behaviors in maltreated youth placed in foster care: A longitudinal
study of protective and vulnerability factors. Child Abuse and Neglect. 26, 1179-1199.

Third National Incidence Study of Child Abuse and Neglect (1993). Retrieved 1 1-4-04 at:
http ://www.healthieryou.com/cabuse.html .

United States Census Bureau (2004). Population by race 2003. Retrieved 11-4-04 at:
http://www. census. gov/popest/states/asrh/tabl es/S C-E ST2003 -04.xl s .










Walker, S. (2002). Culturally competent protection of children's mental health. Child Abuse
Review, vol. 1, 380-393.

Walters, K.L., & Wheeler, D. P. (2002). Assessment of multicultural competence scales for
social work practitioners. Retrieved 10/14/05 at :
http: www. columbia. edu co csswp/resea~rch/desriptions/Waltershtm.

Wierzbicki, M., & Pekarik, G. (1993). A meta-analysis of psychotherapy dropout. Professional
Psychology: Research and Practice, 24(2), 190-195.

Worthington, R., Mobley, M., Franks, R., & Tan, J.A. (2000). Multicultural counseling
competencies: Verbal content, counselor attributions, and social desirability. Journal of
Counseling Psychology, 47(4), 460-468.

Zuravin, S., Orme, J., & Hegar, R. (1994). Predicting severity of child abuse with ordinal probit
regression. Social Work Research, 18(3), 131-138.









BIOGRAPHICAL SKETCH

Michael Ghali was born in Fort Thomas, KY, in 1975. The second of three children,

Michael graduated from Newark High School in Newark, OH, in 1993. During his high school

years, he was active on the soccer, swimming, and tennis teams as well as a singer in numerous

choruses. Upon acceptance to the University of Missouri--Columbia School of Journalism,

Michael moved to Missouri, but quickly realized his career passion for communication would be

better served in helping others, rather than simply reporting news.

Michael graduated from M~izzou with a Bachelor of Science degree in psychology in 1997

and spent the next few years working for the Family Counseling Center of Missouri, Inc. as a

case manager and program director. During this time, Michael met his future wife, Chris,

through some mutual friends, began playing guitar, and participated on numerous city league

softball teams.

Satisfied in his chosen Hield, Michael moved to Colorado with Chris to pursue graduate

studies at the University of Northern Colorado. Completion of a two-year program culminated

in the award of a Master of Arts degree in community counseling. Immediately following

completion of this program, Michael and Chris moved to Florida with the hopes of being closer

to family, starting a family of their own, and finding a location for Michael to continue his

education.

After providing individual, family, and group counseling services for adolescents and

adults, Michael was accepted into the doctoral program in counseling psychology at the

University of Florida. He currently resides in Gainesville, FL, with Chris and their 2-year-old

son, Jace. Upon completion of a Ph.D., Michael intends to teach undergraduate psychology at a

college and pursue private practice working with children and families.