<%BANNER%>

Effects of Reporter and Child Characteristics on the Identification and Willingness to Report Child Maltreatment


PAGE 1

1 EFFECTS OF REPORTER AND CHILD CHAR ACTERISTICS ON THE IDENTIFICATION AND WILLINGNESS TO REPORT CHILD MALTREATMENT By DANIELLE D. MADERA A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORI DA IN PRACTICAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2007

PAGE 2

2 2007 Danielle D. Madera

PAGE 3

3 ACKNOWLEDGEMENTS I want to extend the highest measure of gratitude to a small few of the many family, friends, and colleagues who have helped me comple te this great journey in scholarship. Let me begin by thanking my mother, Christine Madera, a nd father, Demetrius Madera, who instilled in me the drive, motivation, and pr ide necessary to successfully comp lete ones doctorate. I would like also to thank Emily Hunter for her invalu able statistical assistance and collaboration throughout this process. On this same front, a special thanks goe s out to Susan Eichler of the Hernando County Department of Children and Fa milies for facilitating the use of her Child Protective Investigators and assisting in the da ta gathering process. A special thanks also goes out to each of th e four members of my dissertation committee, Dr. Tina Smith-Bonahue, Dr. Diana Joyce, Dr. Jaime Algina, and Dr. William Conwill, who made unique and positive contributions to the pres ent project. I appreciate you all more than I can say and hope to repay some small measure of what you have given me by always using my degree to help others as you have helped me.

PAGE 4

4 TABLE OF CONTENTS page ACKNOWLEDGEMENTS.............................................................................................................3 LIST OF TABLES................................................................................................................. ..........6 ABSTRACT....................................................................................................................... ..............7 CHAPTER 1 LITERATURE REVIEW.........................................................................................................9 Ecological Model of Child Maltreatment...............................................................................10 Ontogenic Development..................................................................................................10 Microsystem....................................................................................................................11 Exosystem...................................................................................................................... ..11 Macrosystem....................................................................................................................14 Culture, Race, and Ethnicity...................................................................................................15 Culture and Child Maltreatment.............................................................................................15 Focus in the Literature........................................................................................................ ....16 Effects of Culture on Treatment.............................................................................................18 Effects of Culture on Child Maltreatment..............................................................................20 Defining Child Maltreatment..................................................................................................20 Parenting Practices............................................................................................................ ......24 Reporting Child Maltreatment................................................................................................28 Disproportionality in Child Maltreatment..............................................................................34 Models of Disproportionality.................................................................................................38 2 METHOD......................................................................................................................... ......41 Participants................................................................................................................... ..........41 Measures....................................................................................................................... ..........42 Demographic questionnaire.............................................................................................42 Vignettes...................................................................................................................... ....42 Procedure...................................................................................................................... ..........43 Analysis....................................................................................................................... ...........44 3 RESULTS........................................................................................................................ .......47 Descriptive Statistics......................................................................................................... .....47 Correlations................................................................................................................... ..........47 Research Question 1............................................................................................................ ...47 Research Question 2............................................................................................................ ...48 Research Question 3............................................................................................................ ...49

PAGE 5

5 4 DISCUSSION..................................................................................................................... ....56 Limitations.................................................................................................................... ..........59 Conclusion..................................................................................................................... .........61 APPENDIX A DEMOGRAPHIC QUESTIONNAIRE..................................................................................63 B VIGNETTES...................................................................................................................... ....66 C INFORMED CONSENT........................................................................................................69 REFERENCE LIST................................................................................................................. ......71 BIOGRAPHICAL SKETCH.........................................................................................................75

PAGE 6

6 LIST OF TABLES Table page 2-1 Descriptive statistics for participants by form (n=150)................................................... 46 3-1 Descriptive statis tics for vignettes by form..................................................................... 51 3-2 Correlation matrix..................................................................................................... ......52 3-3 ANOVA an alyses for severity rating by form.................................................................53 3-4 Regression analys es for severity rating........................................................................... 54 3-5 Logistic regression an alyses for willingness to report.................................................... 55

PAGE 7

7 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy EFFECTS OF REPORTER AND CHILD CHARACTERISTICS ON THE IDENTIFICATION AND WILLINGNESS TO REPORT CHILD MALTREATMENT By Danielle D. Madera May 2007 Chair: Tina Smith-Bonahue Major: School Psychology This study examined the effects of reporter a nd child characteristic s on the identification and willingness to report child maltreatmen t among currently employed Child Protective Investigators. The subjects consisted of 150 Child Protective Investigators employed in seventeen different counties in North Central Flor ida. All participants completed a demographic questionnaire along with eight di fferent vignettes depicting problem atic parental behavior. The vignettes were presented in three different form s, Form A depicted the child victim as black, Form B as white, and Form C did not provide a description of the ch ilds ethnicity. The participants were then asked to read each vignette and rate the information on a Likert scale of seriousness from 1 to 7, with being not serious and being ve ry serious. After the respondents rated the seriousness of the scenario, they were then asked to indicate how they would tell the educator to respond by selecting one of three options: (1) no action, nothing needs to be done about the situation, (2 ) encourage the family to seek professional help, or (3) notify the child protection agency to investigate the si tuation. The research questions were analyzed using a one-way, between subjects analysis of variance, multiple regression, and logistic regression. Findings did not s upport that Child Protective Inve stigators rate scenarios of

PAGE 8

8 maltreatment differently based on the childs et hnicity. Partial suppor t was found for reporter characteristics affecting the overall severity ratin g and willingness to report child maltreatment. More specifically, years in soci al services of the reporter was found to significantly predict the overall severity ratings give n to several vignettes as well as willingness to report.

PAGE 9

9 CHAPTER 1 LITERATURE REVIEW Despite decades of attention and numerous prevention efforts, child abuse and neglect remains a serious social problem. An estimated 3 million children were alleged to have been abused or neglected and received investigations or assessments by state and local child protective service agencies during the Federal fiscal year 2004 (U.S. Department of Health and Human Services, 2006). Out of those 3 million children, an estimated 872,000 were determined to be victims of child abuse or neglect. More than 60% of child victims we re neglected by their parents or other caregivers. Approximately 18% were physically abused, 10% were sexually abused, and 7% were emotionally maltreated. Child maltreatment itself was not identified until 1962 when Dr. Henry C. Kempe first introduced what he termed battered ch ild syndrome (Kempe, Silverman, Steele, Droegemueller, & Silver, 1962). Kempes groundbr eaking work helped to raise awareness and legislation in the fight against ch ild abuse and neglect. Kempe coined the term battered child syndrome to describe a clinical condition in young children who received serious physical abuse, generally from a parent or foster parent . Furthermore, Kempe stated that the syndrome should be considered in any child exhibiting abus e or neglect in the presence of a discrepancy between the explanation and clinical findings Kempe also suggested that alcoholism, promiscuity, unstable marriages, and criminal activity are factors that may increase the risk for battered child syndrome. Since that time, many researchers have attempted to create a model complex enough to capture the multifaceted nature of child maltreatment. This paper reviews the literature in order to provide a theoretical model for conceptualizing child maltreatment, as well as the specific factors related to the comple x social, environmental, and cultural variables that underlie child abuse and neglect.

PAGE 10

10 Ecological Model of Child Maltreatment The most comprehensive and widely-accepted model in explaining the etiology of child maltreatment is that proposed by Belsky (1980) a hybrid of the mode l first introduced by Bronfenbrenner (1979), to explain the ecological perspective of human development. Belskys addition of a fourth level, the ontogenic level, in describing the context of child maltreatment differentiates it from Bronfenbr enners earlier work. Bels kys model describes four interconnected levels: (1) ontongenic, (2) micros ystem, (3) exosystem, and (4) macrosystem. Within this framework, understanding child ma ltreatment is contingent on examining the relationship and interaction of a ll four levels; examining any one level in isolation yields an inconclusive explanation. Moreover, each le vel is influenced by the familys culture (Scannapieco & Connell-Carrick, 20 05) suggesting that exploration of the cultural context of the family is necessary to understand any child or parent behavior. Ontogenic Development In Belskys (1980) description of ontogenic development, the i nvestigation of the childhood history of an abusive parent is ne cessary for developing a more comprehensive understanding of how their parenting behaviors may have been shaped and reinforced. Previous research has shown that a chil dhood history of abuse alone does not determine whether victims will ultimately abuse their own children. While approximately 70% of parents who abuse children were neglected or abus ed in their childhood (Egeland, 1993), other factors have been identified to contribute to this high statistic. Parents with a history of childhood abuse, reported feeling depressed and angry over their childhood, having substance abuse problems, as well as having limited parenting skills. Thus, a chil dhood history of abuse alone will not result in becoming an abusive parent, but is a crucial factor in understanding the etiology of child

PAGE 11

11 maltreatment and how it may predispose parents to react to certain stimuli in the microsystem and exosystem (Scannapieco & Connell-Carrick, 2005). Microsystem Consistent with Bronfenbrenners model, in Belskys model, the Microsystem is composed of the family unit and the immediate context in which the abuse occurs This level of the ecological model of abuse inco rporates the family, character istics of the parent, and characteristics of the child (Belsky, 1980). Many researchers have investigated parent and child characteristics that are related to an increased risk of abuse and neglect. For example, psychological characteristics such as low self-esteem and poor impul se control have been shown to exist among parent perpetrators (Dubowitz, 1999; Milner, 1998). Additionally, parenting skills have been a noted deficien cy in many offending parents. Abusive parents are more likely to utilize punishment, threat s, and coercion and less likely to use reasoning and positive reinforcement (Milner). While the responsibility for abuse should neve r fall on a child, research has shown that some child characteristics incr ease the likelihood that a child will suffer maltreatment at the hands of their caregiver. Child characteristic s such as difficult temperament and disabilities (e.g., learning disability, speech or language delay, emotional/behavioral problems) have been shown to leave children at an increased risk of ex periencing maltreatment (Sullivan & Knutson, 2000). Thus, one could imagine how the combinati on of a child with spec ial needs and a parent with poor impulse control or other limitations c ould create an interactive relationship with an increased risk of abusive parental responses. Exosystem Belsky (1980) describes the exosystem as the la rger social structures that encompass the family unit incorporated with the mesosystem de scribed in Bronfenbrenne rs original model.

PAGE 12

12 Research has indicated many family characterist ics strongly correlate wi th the occurrence of child maltreatment. While child maltreatment transcends all socioeconomic levels, poverty has been the most noted risk factor. Sedlack and Broadhurst (1996) found that family income was the strongest correlate of incidence among all cat egories of maltreatment Additionally, children from lower income families had the highest likel ihood of experiencing neglect and serious injury or death. Poverty may be related to such factor s as unemployment, lack of education, substance abuse, and social isolation, all of which contribute to the overall risk of maltreatment (Horton & Cruise, 2001). One of the seminal studies in the fiel d conducted by Gelles (1992) examined the relationship between poverty and violence towards children. The da ta utilized in the study were collected in two national surveys. The first su rvey, The First National Family Violence Survey, was conducted in 1976 while the second survey, Th e Second National Family Violence Survey, was conducted in 1985. The concept of viole nce was operationalized using the Conflict Tactics Scales (CTS). This measure contains 18 items that ask the respondent how often they use each technique when angry at a family member. The First National Family Violence Survey sampled 2,143 households in which there was at l east one child at home between the ages of 3 and 17 years. The 1985 National Family Violence Survey utilized a sample of 3,233 households in which there was at least one child under the ag e of 18 years. The results for the 1976 survey found that the rate of severe violence was 62% higher in poverty-income families compared to non-poverty income families. Moreover, the ra te of severe violence was 250% higher in poverty-income families. The 1985 survey produced similar results. Severe violence was found to be 46% higher in poverty-income families a nd very severe violence was 100% higher in

PAGE 13

13 poverty-income families. This study was one of th e first in the field to provide support for the claim that violence is more likely to occur in low income families. In a similar study, Drake and Pandey (1996) examined the relationship between neighborhood poverty and rates of re ported and substantiated cases by Child Protective Services. The data utilized in this st udy were from the 1992 Child Abuse a nd Neglect database in Missouri along with the 1990 census data to assist in categorizing neighborhood poverty levels. Families were divided into three categories based on th e location of their homes : least poverty subgroup, median poverty subgroup, and highest pove rty subgroup. The study examined 481,722 total families, across all three groups and found that fo r all types of abuse, the least poverty subgroup had the lowest number of reported and substantia ted cases of child maltreatment. The reporting rates for physical abuse grew significantly acro ss the three poverty leve ls starting from the lowest to highest. The substantiation rates for physical abuse ranged from 8% to 23% across the three poverty levels moving from lowest to highest. Similar results were found for child neglect. Both reporting and substantiation rates increased along with the poverty level. Consistent with the results reported in Gelles (1992), this study provides more support for poverty as a risk factor for all types of child maltreatment. Understanding how the social context in whic h a family lives may affect child rearing practices. Many researches have investigated how residing in certain neighborhoods may affect the occurrence of child maltreat ment. When socioeconomic status of neighborhoods was held constant, differences in maltreatment rates we re still found between the neighborhoods (Garbino & Sherman, 1980), suggesting the importance of factors beyond basic resource levels. Upon further investigation, formal and informal social supports were related to neighborhoods with a lower rate of maltreatment.

PAGE 14

14 Macrosystem The Macrosystem is the larger cultural framew ork that encompasses th e individual, family, and community (Belsky, 1980). For example, phys ical punishment of children has been shown to have differential support acro ss various cultural groups Researchers have consistently shown that harsh physical punishment was more prev alent among African-Americ an families than in European families (Baumrind, 1995; Deater-D eckard, Dodge, Bates, & Pettit, 1996). Differences have also been reported between Hispanic, African-American, and Caucasian parents use of parental nurtur ing behaviors with their children, with African-American parents displaying more nurturing behaviors than thei r Hispanic counterpart s (Ferrari, 2002). Racism is another component described as part of the macrosystem (Scannapieco & Connell-Carrick, 2005). As institutional racism continues to block opport unities for increased educational and employment prospects for some, the frustration and stress among these cultural groups may become heightened creating an increase d risk of child maltreatment. Recently, more attention has been given to the ove rrepresentation of certain minori ties in both reporting statistics and substantiation of child maltreatment cases, le nding support for the existence of institutional racism (Fluke, Hedderson, & Curtis, 2003). In subscribing to the ecological model of abuse, understanding the role of culture is critical to understanding each level of the model. The characteristics of the parent, child, and family all are contingent upon the cu lture of the family, neighborhood, and so ciety as a whole. Despite the models difficulty to be tested empirically, the ecological model of maltreatment has been noted as the best theoretical model for explaining ch ild maltreatment (Garbari no, 1977). Instead of trying to focus on a single process or explanati on, it provides a model for the comprehensive interaction of all the factors involved in increasing the risk of child maltreatment.

PAGE 15

15 Culture, Race, and Ethnicity Throughout the literature, definitio ns of the terms race, culture and ethnicity are discussed extensively, with a number of author s highlighting the lack of clarity in the use of these terms. The American Psychological Association (2002) de fined culture as the belief systems and value orientations that influence custom s, norms, practices, and social institutions (p.2). Furthermore, all individuals are cultural beings and have a cultural, ethnic, a nd racial heritage (p.3). The APA argues that the term race is a socially constructed categ ory by which society judges and stereotypes individuals. On the other hand, ethnicity is defined as the acceptance of the group mores and practices of ones culture of orig in (p.3). These terms have no agreed upon definitions among researchers and thus are used interchangeably throughout the literature base. Therefore, in the present literature review the te rms race, ethnicity, and culture may be seen as interchangeable, as the terms were used intercha ngeably during the search of the literature. Moreover, the term cultural competence will re fer to an awareness of the individuals race, ethnicity, and/or cultural and how those belief systems may influence their customs, norms, practices, and social institutions. Culture and Child Maltreatment As our society becomes increasingly culturall y diverse, the growing need for cultural competence and cultural sensitivity has come to the forefront as a national c oncern. In no area is the need more evident than child abuse and child protection. Culture provid es the context for all interactions and events affecting children within th eir family structure. It also provides the basis for defining and responding to child maltreatme nt. As current demographic trends have predicted, the changing ethnic composition of this c ountry is a growing rea lity that further calls for a high standard of cultura l competence (Korbin, 2002).

PAGE 16

16 Cultural competence was first introduced to child protection by anthropologist James Green (1982) as the term ethnic competence, which referred to the growing need for human service workers to be trained in cultural se nsitivity (Dubowitz, 1999). According to Green, cultural competence is a repertoire of skills and knowledge that enables one not only to acquire sensitivity to cultural diversity but also to transcend cultural boundaries In other words, culturally competent practices prioritize childrens well-being and protection while understanding the cultural context of the child (Dubowitz, 1999). Thus th e culturally competent Child Protection Investigator would have the abi lity to differentiate between cultural practices and cultural practices which are harmful to the child. Within the field of child pr otection, increased attention ha s also surrounded the use of cultural evidence to defend child rearing practices that differ from the mainstream definitions of child maltreatment. Cultural evidence is defined as information offered by immigrant or nonmainstream parents to influence the legal rami fications following allegations of child abuse and neglect (Levesque, 2000). This creates anot her layer of complication for professionals working in the field. Child protection workers mu st be informed and educated on how the laws differ from state to state and what may be permi ssible as cultural evidence. Most states have identified exceptions for cultural practices in such areas as religious beli efs in seeking medical attention, corporal pun ishment, and poverty (Terao, Borrego, & Urquiza, 1999). Thus, assuming variability in clinicians unders tanding of cultural issues, vast differences may exist in the decision-making process surrounding identify ing and reporting child maltreatment. Focus in the Literature With such an increased focus on cultural competence and sensitivity, Nybell and Gray (2004) conducted a study investigating the proce ss of culturally competent organizational change among social service workers, supervisors, and managers from three nonprofit child and family

PAGE 17

17 agencies. Interviews and focus groups were performed with each agency to gain an understanding of how employees viewed cultural competence and its development within their agency. The results were also compared by r ace, agency position, and pr ogram affiliation of the interviewee. The case studies highlighted disparate definitions of cultural competence between individuals as well as an overw helming need for more training on cultural differences. More specifically, staff discussed the need for increa sed diversity in the hi ring process within the agency, along with the cultural clash between pred ominantly white workers and minority clients. Workers indicated the desire for more open dialo gue in the workplace pertaining to racial and cultural differences. Also mentioned was the need for minority speakers to assist in training the staff towards becoming more culturally competen t. While this study is limited in that the findings cannot be generalized to other agenci es, it suggests the need for additional training opportunities within and between agencies. Researchers have also examined the extent to which child maltreatment literature has attended to the issue of ethnic ity in the design and sampling of research studies. In their 2001 study, Behl, Crouch, May, Valente, and Conyngham provi ded a content analysis of 1,133 articles published between 1977 and 1998 in the child maltreat ment literature. Th e authors sought to identify which studies provided a focus on ethnici ty in their sampling, pa rticipants, and design. The articles were taken from three sp ecialty journals in child maltreatment, Child Abuse & Neglect Child Maltreatment and Journal of Child Sexual Abuse and randomly assigned to three coders. Coders independently reviewed each article for the presence of a focus on ethnicity, analysis by ethnicity, or ethnicity held constant. Reliability was calculated across 20% of the articles resulting in interrater agreement ranging from .77 to .95.

PAGE 18

18 The researchers found that of the 1,133 arti cles analyzed, 6.7% included a focus on ethnicity. The percentage of articles that reported data on the ethnic comp osition of the participants increased significan tly across the years examined. Surprisingly, only 50% of the most recent articles examined (1995) included any information on ethnicity of the subjects. In addition, approximately 12% of th e articles used ethnicity as a c ontrol variable in the analyses. Those studies generally held ethnicity constant in order to investigate the relationship among the variables of interest. Surprisingly, of the most recent articles (1995), approximately 75% did not include ethnicity as a vari able in the design or analyses. Implications of this study would be the increased need for researchers to include ethnicity at a minimum as a descriptive variable if not a variable of interest. Limitations of this study include th e focus on only three of the most prominent journals in the area of child maltreatment to the exclusion of lite rature in other fields that may be of relevance. Also, the data examin ed provide trends in th e research available only to the year of 1998, excluding the last six years of research. Effects of Culture on Treatment Researchers have also examined the importance of culture in treating abused and neglected children. Cohen, Deblinger, Mannarino, and Are llano (2001) reviewed the extent to which cultural issues have b een addressed in the empirical trea tment outcome literature for child maltreatment, the measurement limitations, and po ssible explanations fo r the results found. Some evidence in the literature suggests that ethnicity may play a role in the severity of symptoms exhibited by child victims of maltreatme nt. Several researcher s report that minority children suffer more lasting detrimental effects following child maltreatment as compared with Caucasian children (Cohen et al.). For example, Morrow and Sorrell (1989) found that in children having suffered child sexual abuse, Hisp anic and African-American children displayed significantly more behavioral problems and depressive symptoms than their Caucasian

PAGE 19

19 counterparts. Rao, DiClemente, and Ponton (199 2) found that Asian children who had suffered sexual abuse were significantly less likely to di splay anger and sexualized behavior but more likely to display suicidal ideation versus Afri can-American, Hispanic, and Caucasian children. However, the lack of studies that control for treatment received among the children is a confounding variable requiring much more research before interpreting thes e results too broadly. Cohen et al. (2001) also reviewed the lit erature surrounding minority use of treatment services. Several researchers cite the growing evidence that Asian American and Hispanic children are less likely to receive mental health treatment re gardless of need and socioeconomic status (Bui & Tekeuchi 1992; McCabe et al., 1999). Surp risingly, African-American children covered by Medicaid were shown to receive a disproportionately large amount of mental health treatment, while those that were not covered by any insurance received the least amount of services (McCabe et al.). In understanding the relationship between culture and child maltreatment, several methodological designs have been used. Me thods have included ethnographic approaches, vignettes, focus groups, clinical studies, and population survey s (Korbin, 2002). Most often, ethnographic portraits are used in studying cult ure to allow for a comprehensive view of a specific culture. This type of methodology often provides inform ation that cannot be gained through a statistical analysis. Conversely, th ese types of qualitative methods, rich with descriptive information, often lack the gene ralizability that a s ound empirical study would produce. Many of the ethnographic studies involve much smaller sample sizes and labor intensive analyses that make replication vi rtually impossible. Other types of studies such as survey work or aggregate analyses provide much larger sample sizes, more powerful results, and more

PAGE 20

20 generalizable findings. However, these types of studies also lack the information on the cultural context and processes that the individual familie s endure (Korbin, 2002). Thus, Korbin calls for researchers to provide a more comprehensive desi gn when initiating research that allows for both qualitative and quantitative methodology within each study. In addition, the importance for researchers to include ethnicity as a must hav e variable in all research on child maltreatment from this point forward is discussed (Korbin). Effects of Culture on Child Maltreatment Understanding the importance of culture in th e etiological model of abuse, researchers have investigated the effect culture has on abus e and its reporting. Cultu re has been shown to affect the way in which lay people define child maltreatment, the reporti ng characteristics of both lay people and professionals deemed as manda tory reporters of child maltreatment, and the reporting and substantiation rate s of actual cases of child maltreatment. As mentioned previously, ethnicity of victims of maltreatment as a variable of interest examined in the literature is relatively new with only 6.7% of th e articles published before 1998 including it as a focus of research (Behl, Crouch, May, Valente, & Conyngham, 2001). The few studies that have examined culture and ethnicity since then have yi elded differing results, small sample sizes, and studies that are difficult or im possible to replicate. Thus, th e current research leaves many unanswered questions concerning the effect of culture on all aspects of child maltreatment. In the following literature review, the areas of de fining child maltreatment, parenting practices, and reporting child maltreatment are discussed. Defining Child Maltreatment Previous research focusing on the etiology of child maltreatment has suggested that cultural differences in child rearing beliefs may influence specific definitions of child abuse and neglect. Researchers using vi gnette studies have indicated th at minority groups may provide

PAGE 21

21 different definitions of child maltreatment, yet do not regard scripted scenarios with any less significance than the mainstream culture (Kor bin, Coulton, Lindstrom-Ufuti, and Spilsbury, 2000). Korbin et al. investigated the link be tween definitions of child maltreatment among parents residing in different nei ghborhoods with differing levels of risk for child maltreatment. The sample included 400 parents taken from 20 different neighborhoods in Cleveland, Ohio, representative of all levels of the US census. Households were eligible for participation in the study if there was at least one child under the age of 18 living in the home and one parent or guardian. The sample was mostly female (81.8%) with the following racial diversity: AfricanAmerican (54.8%), Caucasian ( 36.3%), Hispanic (6.8%), Asian and Native American (2.1%). Individual variables such as age, race, income, education, and marital stat us were collected as well. In an attempt to investigate the pr esence of neighborhood definitions of child maltreatment, the respondents were first asked to iden tify three things they considered to be child abuse and neglect. Participants were then as ked to rate 13 factors drawn from the current literature on the etiology of ch ild maltreatment on a scale from 1 (contributes nothing) to 10 (contributes a lot) as to how much they thought the factor may contribute to child maltreatment. Two measures were utilized for analyses in th is study, a census-based measure of the level of impoverishment in the neighborhood and a responde nt based measure of perceptions of the neighborhood conditions. While the researchers found an overall agreemen t on the definitions of child maltreatment, several significant differences we re also discovered. African-Ame ricans were more likely to list behaviors of neglect while Caucasians were mo re likely to list behaviors of physical abuse during the free response exercise. Women were also more likely than men to include lack of supervision as an example of child maltr eatment. Neighborhood pe rceptions of factors

PAGE 22

22 contributing to the etiology of child abuse and neglect also yielded significant findings. Respondents living in poorer, more chaotic, high crime neighborhoods were more likely to attribute the etiology to povert y and family structure. Conve rsely, respondents living in less poor, less chaotic neighborhoods were more likely to attribute the etiology to a lack of family values. Limitations of this study include the overwhelming repr esentation of women in the sample as well as selection bias based on recruiting of subjects. Another related study complete d by Fontes, Cruz, and Tabac hnick (2001) explored the qualitative differences between African-American and Latino views of child sexual abuse. The authors utilized focus group interviews to addres s any differences in definitions of child sexual abuse, signs that it is occurring, whether it occu rs in their communities, and if men and women hold different views. The participants in cluded 34 men and 24 wome n divided into eight separate focus groups defined by older a nd younger African-American women, older and younger African-American men, Latino English-sp eaking men, Latina English-speaking women, Latino Spanish-speaking men, a nd Latina Spanish-speaking women. The only demographic data obtained was age, ethnicity, and th at generally participants were from working class or poor and have low levels of academic education. All the information available on r ecruiting describes that all participants who volunteered and met the criteria of being over 18 and of African-American or Latino descent were accepted. The Latino gro up was comprised of mostly Puerto Ricans with a few people from the Caribbean and Central Amer ica. The focus groups lasted between 45 to 90 minutes and were videotaped for later analysis. The results showed several differences between the responses of the focus groups. Both Latino men and women spoke more about sexual abuse as an issue within their own family or community while in African-American groups almo st no personal stories were shared. In the

PAGE 23

23 rare occurrence that they were shared, the stor y was ignored by the groups. Male participants tended to be more detailed in de scriptions of child sexual abuse while females were hesitant to name specific acts. Only Latinos, both men and women, described changes in the family, culture, and community as factors contributing to child sexual abus e. In English-speaking Latina groups, participants described so cial and economic pressure to keep a man around the house as a risk factor for child sexual abuse. African-A merican groups, both men and women, referred to drug use and abuse as a risk factor for child sexual abuse (Fontes et al., 2001). While the results obtained provide an in dept h understanding of the participants beliefs, generalizability of the re sults is limited. First, the participants were not randomly selected and the sample size was relatively small. Additiona lly, only limited demographic data was collected. The duration of the focus group was short, allowi ng limited time for the participants to feel comfortable sharing such intimat e information. Finally, the fo cus groups were all led by men, which the authors state could have affected so me of the participants willingness to share personal information. In summary, results of studies examining how culture affects definitions of child maltreatment suggest that there ma y be differences in the type of abuse certain cultures are willing to discuss or admit occur within their communities. Findi ngs have also suggested that the neighborhood in which one resides may have an affect on how one conceptualizes the etiology of child maltreatment. Additionally, gend er differences have been reported to affect both of the aforementioned findings, however; these results are extracted from studies with an overwhelmingly female sample. Thus, completing mo re studies incorporating a larger sample of both men and women along with a mo re diversified group of particip ants are needed to further the research in this area.

PAGE 24

24 Parenting Practices Researchers have also focused on cultural differences and pred ictors of parenting attitudes and behaviors, hypothesizing their potential impa ct on child maltreatment. With such a high percentage of U.S. parents utilizing physical punishment, researchers have examined the longterm effects of using such disc ipline practices. Some researchers maintain that simply using physical punishment may result in later dysfuncti on while others suggest the detrimental effects depend upon the context in which the punishment is used. Wissow (2001) analyzed data from the Commonwealth Fund Survey of Parents with Young Children, a na tional sample of 2,017 parents with at least one child younger than three living in the ho me. The sample consisted of 1,320 mothers/female guardians and 687 fathers/male guardians. The parents were interviewed to assess their use of six types of disciplin ary practices, 3 negative (yelling, spanking, and hitting) and 3 positive (time out, taking away t oy, and explaining). Only the parents of 18month-olds were asked about ne utral/positive practices. The pa rents were further questioned about how often in the last week they engaged in specific nurt uring activities (e.g., reading to child). Finally, parents were gi ven a self-report survey that measured depressive symptoms. The study found several significant results. First, 40% of parents report spanking sometimes or often, while another 40% report spa nking at least once. The most common form of neutral discipline was time out, with 62% of parents with a child between the ages of 18 months reporting use, while 73 % of parents of 24 month olds reporting use. Boys were significantly more likely to be ye lled at or spanked than girls. Based on the amount of reported spanking, the spankers were grouped into averag e, above average, and highest levels of spanking. The average spankers were found to do so in context with a larger use of nurturing interactions as well as both negative and neut ral disciplinary interacti ons. The above average spankers report less reading, listening to musi c, playing, and hugging their children and use less

PAGE 25

25 of other disciplinary strategies. The highest levels of spanki ng were found at the two extreme ends of the participant groups. The first group was comprised of the greatest proportion of lowincome parents and largest number of Hispanic and African-American parents. The second group with the highest levels of spanking was comprised of the smallest proportion of lowincome parents and was nearly 80% Caucasian. These results indicate that negative disciplinar y practices are not just defined by ethnicity or income level but a more complex constellation of factors. Use of spanking was found to be strongly associated with parental reports of depressive sympto ms. Consequently, groups of parents with higher rates of physi cal punishment seem to be more defined by parental mood than by race, ethnicity or socioeconomic status. A lim itation of the study was the lack of information on how the term spanking was defined and expl ained to the participants. Also limiting the results found in the study is the lack of supplemental informati on on the family environment and context within which the discipline occurs. Fi nally, only one parent wa s randomly chosen from two parent homes to participate without question as to the consistency of disciplinary techniques across caregivers. Ferrari (2002) examined the relationship between culture and pa renting behaviors and attitudes. This study also examined the imp act a childhood history of abuse would have on the parenting beliefs. This study hypot hesized that cultural variables would be more predictive of parenting practices than ethnic ity as a demographic variable. The study included 75 fathers and 75 mothers who were recruited as nontraditional students attending night classes at a community college. The sample was comprised of Afri can-Americans (33%), Hispanics (33%), and Caucasians (33%). The participants were give n a booklet containing seve n scales or checklists in a randomized order: The Machismo Scale, Familism Scale, Valuing Children Scale, Conflict

PAGE 26

26 Tactics Scale, Nurturance Scale, vignettes of child maltreatment, and the Childhood Trauma Questionnaire. The study found that regardless of ethnic status fathers who held familism in low regard were more likely to use physical punishment to di scipline their children than fathers who valued familism more. Machismo predicted more use of physical punishment from fathers, but not for mothers when controlling for ethni city. For all parents, valui ng children more was associated with lower tolerance for mistreating behaviors. A history of childhood abuse or neglect was predictive of the mothers use of physical and ve rbal punishment. Interestingly, the fathers who reported a history of child abuse were less likel y to use physical punishment. The data suggests that the intergenerational transmission of abuse may hold true for mother but not for fathers, at least within this population. No ethnic differences were found in the seriousness ratings of child abuse and neglect. African-American parents were found to report participating in more nurturing behaviors than Hispanics, while also using more physical discipline than both Hispanic and Caucasian participants. African-American mothers were found to use significantly more verbal and physical abuse than African-American fathers. The findings indicate that both ethnicity and gender are complex factors which are comprised of social roles, gender roles, and norms, all of which may impact parenting styles and definitions of ch ild maltreatment. This study is limited by the use of only college students with no measure of socioeconomic status or acculturation. Investigating similar construc ts, Coohey (2001) concentrated on different dimensions of familism and the relationship to child maltreatm ent in Latino and Anglo families. Familism is the most important culture-specific set of att itudes and beliefs held by Mexican, Puerto-Rican, Cuban, Central and South Americans. These cultu re specific attitudes and beliefs emphasize the

PAGE 27

27 importance of maintaining a supportive extended family network. Coohey hypothesizes that less familistic attitudes would thus be linked to child maltreatment among Latinos. The study sought to investigate if varying degrees of familis m existed between and within Latino and Anglo families and if these differences were in fact related to child maltreatment. The participants included Anglo and Latina mothers taken from Ch ild Protection Services records and matched with public school parents from Chicago on mothers ethnicity, c ountry of origin and income level. The sample included 51 nonabusive Anglos, 51 abusive Anglos, 35 nonabusive Latinas, and 35 abusive Latinas. The study utilized a netw ork circumscription approach in capturing the majority of mothers informal supports. The moth ers were asked to name formal and informal supports and evaluate each person on a variety of characteristics measuring familism. The three dimensions of familism examined were attitudina l familism, behavioral familism, and structural familism. Attitudinal familism is the idea that relatives possess more positive characteristics than nonrelatives and a preference for exchan ging valued resources with relatives over nonrelatives. Behavioral familism is the actual receipt of support from relatives as opposed to attitudes or beliefs about what others should give or get. Mo reover, it focuses on support Latinos receive from extended kin and not everyone in their informal suppor t network. Finally, structural familism is the actual size and proximity of kinship networks. Several differences were found across all dimensions of familism. Anglos, both abusive and non-abusive, were found to be more likely to report at least one of their kin members were critical of them as well as having at least one unreliable relative. A significant interaction was found for the abusive Anglo mothers who reporte d they had at least one kin member who couldnt control his or her anger. Nonabusive mo thers, regardless of et hnicity, report more kin who were warm and caring. Main effects for bot h abuse status and ethnicity were found on the

PAGE 28

28 level of interest in getting emotional support from friends. Abusive mothers and non-abusive Anglo mothers were more inte rested in obtaining emoti onal support from friends. Significant results were also found across the dimensions of behavioral and structural familism. A significant main effect for abuse status on the number of kin members was found. Nonabusive mothers had significantly more kin in th eir social networks than abusive mothers. Abusive mothers had fewer contacts with their kin than nonabusive mothers. Abusive mothers also have fewer friends than nonabusive mothers. While the authors hypotheses were partially supported by this study, several implications for culturally relevant practices were found. Based on the results, practitioners s hould inquire as to the supportive or unsupportive nature of the mothers kin when either ascertaining level of support or investigati ng possible out of home placements for children. Also, consistent with previous research, a pattern of disconnectedness from kin could put poor Anglo mothers at risk fo r maltreatment. Practitio ners should investigate both kin and friends level of support, proximity, and availability for Anglo mothers. A pattern of receiving more emotional support from friends versus kin may characterize a family cultural pattern that negatively influences parenting for Latinas in this study. Some limitations of this study include the participation of only women in the sample. Also, during the interview process questions were translated for non English-speak ing participants, possibly losing some meaning during the questionnaire process. Reporting Child Maltreatment Identification and willingness to report child ma ltreatment have gained a large focus in the child maltreatment literature. With such a diverse number of professionals designated as mandatory reporters of child maltreatment unde r state statutes, many different groups have subsequently been studied in the literature. Without the ability to control and deviate case variables such as the childs race, SES, or age within real life child maltreatment cases,

PAGE 29

29 researchers have overwhelming adopted the use of simulated case vignettes to access the effect of reporter characteristics and the characteristics of the scenario on the reported severity of abuse and willingness to report. The process of identifying and reporting child maltreatment is a comprehensive procedure involving three distinct phases (Ashton, 2004). The reporter will first gather data about a given situation, assess the behavior, and respond in a specific manner (Ashton). Personal characteristics of the reporter, such as beli efs, attitudes, and opinions developed through experiences may play a role in each step of th e above process. Ashton examined the relationship between personal characteristics of 276 entry-le vel workers and their decision to report child maltreatment. The participants in this study were a group of cu lturally diverse college seniors from an urban area all majoring in the social se rvice field. The study provided the participants with a self-administered questionnaire comprise d of case vignettes depicting different child maltreatment scenarios. Eight i ndependent variables were assesse d for each participant: age, gender, parenthood, mother and fath ers education, college major, ethnicity, immigration status, and whether or not the respondent was born in th e United States. The sample consisted of 30% African Caribbean, 24% African-A merican, 25% Hispanic, 9% White 6% Asian, and 6% other. The participants were presented with 12 vignettes and asked to decide between three outcomes: nothing needed to be done, the family needs to s eek mental heath assistance, or the family needs to be reported to child protection. The respondents also needed to rate each situation on a Likert scale from 1 to 7 for the seriousness of the situation. Ashtons study found that out of all eight char acteristics, only ethni city and immigration status were related to the likelihood of the partic ipants to report the abuse. Caucasians were found to be significantly more likel y to report abuse, while Asians were the least li kely to report

PAGE 30

30 the incident to child protecti on. The only other significant resu lt found was that respondents who were born in the United States were significantly more likely to report versus immigrants who were born elsewhere. This study was limited in that it was a self-repo rt measure which only required college seniors to hypothesize if they would report, not if they actually did in a real life situation. Furthermore, college seniors are not actual mandated reporters of abuse thus limiting the generalizability of the re sults reported in this study. In a similar study completed by Maynar d and Wiederman (1997), a group of 413 undergraduate students recruited from an introductory psychology course were presented with one of eight vignettes de picted child sexual abuse and asked to rate the severity of the scenario, and respond to questions examini ng their perceptions of the situ ation and any attributions of responsibility or blame. The sample consis ted of 208 men (50.4%) and 205 women (49.6%). Furthermore, 87.2% of the respondents were Wh ite, 6.5% were African-American, 2.4% were Asian, and 1.9% were Hispanic. The vignettes de picting a 15-year-old child were rated as less severe than those situations depicting a 7-year-o ld victim. Additionally, ratings of severity of abuse and attributions of blame were significantl y lower when the child and the victim were the same sex. Gender-role attitudes were not signifi cantly related to ratings of abusiveness or attributions of blame. Limiti ng this study was the inclusion of undergraduate students instead of mandated reporters of abuse. Moreover, the sample was overwhelmingly Caucasian further highlighting the need for more culturally divers e samples in the child maltreatment literature. In addition to research focused on undergradu ate responders, graduate students in the area of social work were sampled in a study comp leted by Ashton (1999) using 12 vignettes depicting problematic parental behavior. Th e participants were then asked to rate the seriousness of the situation and whether or not they would report the incident. Th e sample consisted of 86 first-

PAGE 31

31 year graduate students majoring in social wo rk. The sample was overwhelmingly female (83.6%) and predominately Caucasian. Vignettes de picting physical violence, imminent harm, or young children were rated more serious as well as being more likely to be reported. This study did not find any relationship between respondent data, such as age, gender, or ethnicity, and the ratings of severity and willingness to repo rt. The author provides the reasoning that the sample was overwhelmingly homogeneous, thus again calling for a more culturally diverse sample. While sampling first year graduate st udents in a social work program increases the generalizability versus undergraduate samples, it is still not representative of actual mandated reporters. Additionally, previous research has shown that graduate pr ograms in clinical, counseling, and school psychology did not meet minimum sta ndards imposed by the American Psychological Association (APA) in the area of child maltreatment tr aining (Champion, Shipman, Bonner, Hensley, & Howe, 2003). A survey assessed coursework, practica, and research experience in the area of child maltreatment and found that a ll programs fall short of minimal levels of competence, with no increased attention or focu s over the last ten years. For most of the programs, child maltreatment was briefly covered as a subject in law and ethics seminars offered during a semester long course. Thus, research sampling college students may not provide as much relevant information for understanding th e reporting characterist ics of professional mandatory reporters. While detecting and reporting child maltreat ment is not a mandatory requirement of students, it is a legal re sponsibility placed on school personnel. Thus, research has also focused on the reporting characteristics of educators and other sc hool personnel. While this responsibility is of crucial impor tance, studies show that teachers feel unprepared, unsupported,

PAGE 32

32 and ill-informed when it comes to reacting in s ituations that call for c linical and professional judgment. Kenny (2001) surveyed 197 teachers on their knowl edge of child abuse laws, procedures for reporting, and ability to dete ct child maltreatment in two vigne ttes. Of the sample surveyed, 11% of the teachers stated that th ey failed to report incidents of abuse at their school. When asked what impacted their decision not to repor t, 39% stated that they feared making an inaccurate report. Thirty-four percent of both special education and regular education teachers report their pre-serviced training to properly handle child maltreatme nt as being inadequate. The respondents identified a lack of preparation and trai ning in their course work, as well as a lack of knowledge of specific legal requirements. The teachers responded to the vignettes, which both described obvious scenarios of child maltreatment as reporting only 26% for the first case and 11% for the second case. Kenny (2001) also em phasizes the ethnic differences observed among teacher respondents. None of the African-Ameri can or the African Caribbean teachers reported the need to alert child protect ion in the scenarios presented. The results reported in Kenny (2001) are echoed throughout the child maltreatment literature. In a similar study conducted by Hi nson and Fossey (2000), elementary classroom teachers were administered a survey on their abil ities to recognize abus e, perceptions of the reporting process, and their understanding of liabili ty. The results showed many of the teachers were unable to recognize the signs of abuse a nd were unaware of the reporting procedures at their schools. Further, most of the respondents were unaware of immunity rights in the law, indicating that they felt fear and anxiety over possible retributi on from the abuser. In addition, educators reported considering ch ild protective services ineffective and unable to help the children (Hinson & Fossey, 2000).

PAGE 33

33 In a study completed by Webster, OToole, O Toole, and Lucal (2005), a sample of 480 teachers responded to vignettes in which the case characteristics, including socioeconomic status and ethnicity, were systematically manipulate d. The teachers responded to whether they believed the situation to be child abuse and if they felt they would report the incident. The sample was overwhelmingly Caucasian (93.1%) as well as largely female (74.2%). The results showed that in the sample of teachers, overre porting was rare, occurring in only 4.2% of the cases, with underreporting occurrin g in one third of the cases. A difference was found between teachers working in rural schools versus suburban schools, with those in rural schools showing an increased likelihood of underrepor ting the abusive situations. Th is study provided a relatively untapped characteristic of child maltreatment, the difference betw een professionals employed in rural and suburban settings and the effect that may have on re porting rates. Again, a more culturally diverse sample needs to be examined as well as one incorporating an equal amount of men as women. Along with professional educators, other profes sionals have also b een studied throughout the literature. In a study completed by Warner-Rogers, Hanse n, and Spieth (1996) a sample of sixty 3rd and 4th year medical students were provided a set of sixteen vignettes depicting scenarios of child maltreatment that systemati cally manipulated the in jury severity, injury explanation, and delay in seeki ng medical attention while holdi ng SES and ethnicity constant. The sample consisted of 40 males (66.7%) and 20 females (34.6%) with no information on the respondents ethnicities. The study found a signi ficant three-way intera ction between injury severity, explanation, and delay in seeking treat ment. Reporter gender or amount of clinical experience was not shown to affect identification of abuse. Limiting the findings of this study is the use of medical students. While the students in this study were in their 3rd or 4th year of

PAGE 34

34 medical school, the ability to generalize to licen sed physicians and other mandated reporters is limited. Licensed psychologists and certif ied Masters social workers were surveyed by Hansen, Bomby, Lundquist, Chandler, Le, an d Futa (1997) in an effort to identify case and professional variable on the identification and reporting of child maltreatment. The sample consisted of 125 licensed psychologists, 69 (55.2%) female a nd 56 (44.7%) male. The sample was also predominantly Caucasian (95.2%). Also survey ed were 85 Certified Ma ster Social Workers (CMSWs), with 45 being female (52.9%), 40 being male (47.1%). The majority of the CMSWs were also Caucasian (91.8%). The respondents were presented with fi ve case vignettes that systematically manipulated three variables, the childs age, race, and SES. Consistent with previous research, the vignettes depicting a younger child received a higher severity of abuse rating. An effect of socioeconomic status was found, but only relate d to the sexual abuse vignettes. The scenarios depicting families of lower socioeconomic status were associated with higher ratings of abuse. The factor that produced the strongest effect on ratings of abuse was the ethnicity of the parent and child described. The vignettes depicting African-American families were rated as being less severe and less likely to be reported than the same identical scenario depicting a Caucasian family. No significant differences for gender were found in the study. Replicating this study with a more culturally dive rse sample would increase the understanding of the complex issue of race when investigating child maltreatment. Disproportionality in Child Maltreatment The occurrence of disproportiona te representation of race and ethnicity in actual reporting statistics has been examined in the child ma ltreatment literature. Fluke, Hedderson, and Curtis (2003) investigated the racial and ethnic representation of ch ild maltreatment reports using National Child Abuse and Neglect Data System (NCANDS) data for more than 700,000 children

PAGE 35

35 across five states. The data we re collected in year 2000, from one state in the mid-Atlantic region, one in the Midwest, two from the South, a nd one from the Southwest. The states were specifically chosen based on their overall populat ion and diverse racial populations. For each state, a disproportionality representation index (DRI) was constructed for both investigation and victim data. The Investigation DRI is defined as the ratio calculated by dividing the percentage of children in a specific racial group who were subject of an inve stigation by the percentage of children of the same racial group in the population in the state. The Victim DRI uses the same equation to calculate the number of children of a specific raci al group that were found to be victims of maltreatment by the percentage who we re subject of an investigation. The results showed that across all five stat es, the Investigation DRI for African-American children who were subject of an investigation is gr eater than 1. In add ition, across all five stat es, the Victim DRI for African-American children was greater than 1. The Victim DRI was substantially less than the Investigation DRI for African-Ame rican children suggesting a bias in reporting procedures. Great variation among both the I nvestigation and Victim DRI fo r all other ethnic groups was found (Asian/Pacific Islander, White, Hispanic, and Native American), with no consistent results greater than one. In a similar study, researchers investigated th e records of 3,936 childre n receiving care in out-of-home placement for suspected child maltreatment (Lu, Landsverk, Ellis-Macleod, Newton, Ganger, & Johnson, 2004). The study examined the relationship between child characteristics such as age, gender, race, and reason for referral with case outcome decisions. The analysis revealed significant difference s among case outcomes for children of different racial backgrounds. African-American children, wh en compared to relevant census data, were over-represented in the child protective system as well as out-of-home care. Furthermore,

PAGE 36

36 African-American children were significantly le ss likely to reach the goal of reunification with their family than all other ethn icities examined. The study also investigated the initial child abuse reporting rate, uncovering another significan t effect of race. Af rican-American children were found to be over-reported while Asian and Hispanic children were found to be underrepresented. The above study is consistent with results found by Barth (1997). Barth utilized a database that included all children who were placed in out-of-home care in California since the year 1988. The sample studied was a cohort of 3,873 children who were selected because it was the first time they entered out-of-home care, were less th an six years old at placement, were AfricanAmerican, Caucasian, or Latino, a nd were placed in a nonkinship fo ster family. The analysis demonstrated that African-American children we re less likely than Caucasian children to be reunited with their families (41% v. 58%). Fu rthermore, African-American children are more likely to remain in long-term care while Caucasia n children are more likely to be adopted than stay in long-term care. Latino ch ildren were found to be equally as likely to be placed in either setting. Barth expounds as to possible c ontributing factors in this overwhelming disproportionality in ch ild welfare outcomes for African-Ame rican children. He states that hiring additional ethnically competent social wo rkers (p.6) may increase adoption rates for African-American children. McCabe et al. (1999) investig ated representation of differe nt ethnic/racial groups across five public youth service sectors. The sample analyzed in the study was drawn from the Patterns of Youth Mental Health Care in Public Services Systems Project (POC) which includes children (birth to 18 years) who were active in one or mo re of five areas of care (drug treatment, child welfare, juvenile justice, ment al health, and SED services in public schools). The sample

PAGE 37

37 consisted of a cohort of 11,515 children who resided in San Diego during the 1996 year. African-Americans were found to be overrepresente d in the child welfare system, mental health services, juvenile justice, and SED. The onl y sector that they were not shown to be overrepresented was in drug treatment. Caucasia n children were found to be underrepresented in child welfare, mental health, and juvenile ju stice. Furthermore, Caucasian children were represented at expected rates in drug treatmen t and were overrepresented in SED within the public schools. Latinos were unde rrepresented across all five public sectors. With AfricanAmerican children being present in the child welf are system at 264% of the expected rate, the authors suggest that overidentification ma y be to blame. Investigating the same problem, with a slig htly different focus, Garland and Besinger (1997) examined court referrals for mental h ealth services between children of different ethnicities. The sample was ta ken from a longitudinal study, the Foster Care Mental Health (FCMH) research project, and in cluded 184 randomly selected cas es. The authors analyzed the childrens child welfare case file s and found that Caucasian childre n were most likely to have received counseling before entering the system. Moreover, they were most likely to receive counseling within the first eight m onths of out-of-home care as comp ared to the other ethnicities. Upon further analysis, African-American and Hi spanic youth were less likely to use therapy services when controlling for age, type of maltreatment, and prev ious use. Garland and Besinger discuss possible causes of this disparity in th e system. They hypothesize that in addition to family cultural differences, service system factor s (p.663) may play a role in these findings. More specifically, they state th at service systems may not make appropriate services available and accessible to minority families (p.663).

PAGE 38

38 The aforementioned studies indicate that ethnic ity plays a significant role in the decision making process at all levels th roughout the child welf are system. It suggests the need to reevaluate the current model of child protective servic es in an effort to understand exactly how and why the needs of minority clients are not be ing met. More specif ically, the workers, agencies and policies that hold the responsibility for decision making need to be addressed more directly within the relevant literature. In a recent review of the rese arch available on minority children in the child welfare system, Hines, Le mon, Wyatt, and Merdinge r (2004) conclude that ethnicity and/or SES is clearly a factor contributing to the di sproportionate representation of minority children in the child welfare system. Mo re specifically, the authors state, research in the area of bias in reporting remains scant and inconclusive (pg. 522). Models of Disproportionality While overrepresentation of certa in ethnicities throughout the di fferent stages of the child welfare system has been researched and di scussed for years, large scale studies of disproportionality in th e reporting and substantiation of child maltreatment cases are few and far between (Barth, 2005). Barth propo ses several possible explanations for the overrepresentation of African-American children and families in the child welfare system. In his model entitled the Child Welfare Services Decisionmaking Model, Barth describes the disproportionality in the population being primarily attributed to agency char acteristics that allow th e race of the child to play a role in every decision making point throughout the child we lfare system. More specifically, factors such as staffing patterns, culturally incompetent staff, institutionalized racism, and inadequate duration or configurati on of services (p.36) ar e provided as possible contributing characteristics. Barth also cautions that this model is only in the infancy of development due to the difficulty of teasing out race and ethnicity and being able to control and manipulate such variables within the research. Furthermore, Ards et al. (1998) reports that

PAGE 39

39 cultural differences in child-rea ring beliefs between the reporter and the perpetrator along with an unwillingness to report perpetrators of ones own socioeconomic status as a possible reason for differing reporting rates. This model provides a framework for unde rstanding bias in the reporting and substantiation rates for child maltreatment by que stioning the cultural competence and training of child protection workers. However, as shown in the preceding review of the literature, few researchers have actually surveyed the individuals responsible for substa ntiating the allegations of child maltreatment. Regardless of this fact, th e researchers continue to extrapolate the results and hypothesize that bias in the system may be a contributing factor. In summary, this literature review has desc ribed the effects that reporter characteristics have on the seriousness that repo rters rate vignettes depicting child maltreatment as well as the willingness to report these cases to the Child Pr otective Investigators. Such characteristics include gender, ethnicity, profession, and count y in which one works (Ferrari, 2002; Fontes, Cruz, & Tabachnick, 2001; Korbin, Coulton, Lindstrom-Ufuti, & Spilsbury, 2000). In addition to reporter characteristics, the characteristics of the child have been shown to play a significant role. The childs age, gender, ethnicity, and soci o-economic status have all been shown to have an effect on reporter responses (Maynard & Wiederman, 1997 ). Along with this research, a growing awareness of the dispr oportionality in the ethnicity of children being reported and substantiated as victims of ch ild maltreatment has increased the importance of such studies (Fluke, Hedderson, & Curtis, 2003; Lu, Landsve rk, Ellis-Macleod, Newton, Ganger, & Johnson, 2004). Previous researchers have examined the impact of both reporter and victim characteristics among professionals such as doc tors, nurses, medical students, undergraduates and teachers (Hansen et al., 1997; Warner-Rogers, Hansen, and Spieth, 1996; Webster, OToole,

PAGE 40

40 OToole, and Lucal, 2005). The results of such research have been used in hypothesizing the reason for the growing disparity in the child maltreatment statistics. Researchers have suggested that Child Protective Investigators lack of cultura l sensitivity training alo ng with reporter biases have been contributing factors (B arth, 2005). In order to isolate ethnicity as the variable of interest, gender of the child will be held constant in the following study.

PAGE 41

41 CHAPTER 2 METHOD The purpose of this study is to examin e how reporter characteristics and child characteristics affect reporter response among currently employed Ch ild Protective I nvestigators. To this end, the following questions will be addressed: 1. Do Child Protective Investigators rate severity of maltreatment differently based on victims race/ethnicity when gender is controlled? 2. Do Child Protective Investigators rate seve rity of maltreatment differently based on characteristics such as race/et hnicity, gender, amount of cultural sensitivity training of the reporter, and setting of employment (i.e., rural versus suburban)? 3. Is there a difference in willingness to report based on the childs race/ethnicity, reporter race/ethnicity, reporter gender, amount of cultura l sensitivity training of the reporter, or the populations of the county in which the reporter works? Participants A total of 162 participants completed the study. After examining the data for missing values, twelve participants were removed from analyses due to missing independent variables. Therefore, the final sample consisted of 150 Child Protective Investigators employed by the Department of Children and Families (DCF) who are currently working in seventeen different counties in North Central Florida. The particip ants were asked to identify themselves as African-American or Black (B), Caucasian or White (W), Hispanic or Latino (LA), or belonging to some other cultural group (OT). Table 2-1 illustrates the frequency distribution of gender, ethnicity, and county among the participants with in the three forms. The entire sample was comprised of 111 females (74%) and 39 males (26 %), ranging in age from 22 to 64 years. Thirty-two percent of the sample was African -American, 66% was Caucasian, 6% was Hispanic, and 5% identified themselves as Other. Twenty-seven percent of the sample worked in a rural county, while 72% reported working in an urba n/suburban county. The participants reported time on the job ranged from 1 month to 38 years of experience.

PAGE 42

42 The participants were recrui ted through their DCF supervisor s. Every Child Protective Investigator in the sampled 16 counties was aske d to participate in th e study. The first 150 participants who volunteered were included in the study. Each s upervisor was given a summary of the purpose of the study, type of subject s needed, confidentiality of information, and procedures. Additionally, a less detailed summary of the study was also distributed to the supervisors to be disseminated among the par ticipants. The supervisors were consulted regarding the least disruptive manner in which to conduct the study. Date s for data collection were scheduled with the supervisors who were in structed to disperse the summary of the study one day prior. Supervisors were informed that a copy of the results will be available for all participating DCF fac ilities. Informed consent was read aloud and obtained from each participant prior to the participan ts data being used for the study. Measures Demographic questionnaire The measures used in this included a de mographic questionnaire (see Appendix A), along with eight vignettes ad apted from Ashton (1999) (see A ppendix B). The demographic questionnaire was specifically deve loped for this project and reque sted participants provide the following information: gender, ethnicity, age, degree, marital status, religion, childhood history of abuse, number of children, population of count y in which they work, amount of training, and amount of time with current employer. Vignettes Following completion of the demographic data sh eet, participants then read and completed several questions pertaining to eight vignettes. Each vignette depict ed a fictitious description of problematic parental behaviors. The participants were then asked to imagine that as caseworkers they were approached by an educator who became aware of certain circumstances. The

PAGE 43

43 participants were then asked to read each vignette and rate the information on a Likert scale of seriousness from 1 to 7, with being not serious and being ve ry serious. After the respondents rated the seriousness of the scenario, they were then asked to indicate how they would tell the educator to respond by selecting one of three options: (1) no action, nothing needs to be done about the situation; (2 ) encourage the family to seek professional help; or (3) notify the child protection agency to investigate the si tuation. The vignettes were developed by Vicki Ashton (1999) who administered a total of 12 vignettes to first-ye ar graduate students. Ashton tested the vignettes for face cont ent validity of seriousness by presenting them to a group of social work professionals. They were asked to ra te the seriousness of th e vignettes on a scale of 0 to 6, according the amount that each vignette depicted a form of child maltreatment. Ashton also tested the reliabil ity of the vignettes by a 1-week test /retest on a sample of undergraduate social work majors. She reported the median co rrelation coefficient (Pearson r) for the vignettes as .61. Ashton cites the restricted range of responses as contributing to such a low estimate. All four vignettes were used from the low and moderate level of seve rity to construct the eight vignettes that will be utilized in this study. In order to isol ate and examine ethnicity as the variable of interest, the gende r of the child in the vignett es was controlled for. Some modification of the wording was necessary within each chosen vignette in order to change the victims gender. Furthermore, the ethnicity of the child was ma nipulated and three forms of the vignettes will be created. One form will contai n all black children, one form will contain all white children, and one form will provide no info rmation concerning the childs ethnicity. Each form will be randomly assigned among the participants. Procedure Each Department of Children and Families supervisor decided which day was optimal for data collection at their respective agency. On the day scheduled, the principal investigator

PAGE 44

44 handed out consent forms that were read aloud to the par ticipants. The partic ipants then had the opportunity to ask any questions th ey had. Participants were told that if they wished to participate in the study, they should sign the form to indicate that they understood what I read. The consent form instructed the participants not to put their names on any of the study materials in order to protect their privacy. Participants who did not wish to complete the study were free to leave. Each participant was given a packet cont aining a demographic que stionnaire along with eight vignettes depicting different fictitious cases of child maltreatment. Following each vignette were questions pertaining to the seriousness of the scenario and if the Child Protective Investigator would advise an e ducator to report the incident. Three forms of the measure were created to allow for changing the ethnicity of th e children in the scenarios. Each form was randomly assigned among the participants. Partic ipants were then asked to refrain from discussing the study with or around othe r Child Protective Investigators. Analysis The goal of this study was to determine if reporter characteristics and/or case characteristics was related to the reported leve l of severity and/or the decision to report the incident to Child Protective Services. A one-way, between-subjects ANOVA was utilized to investigate whether the reported rate of severity differed base d on the childs ethnicity. The independent variable was the form code (thr ee levels) and the dependent variable was the reported level of severity per vignette. Multiple regression analyses were utilized in analyzing the second research question for each of the eight vignettes. The dependent variable was the reported level of severity, with the independe nt variables being child ethnicity, reporter race/ethnicity, reporter gender, population of the county in which the reporter worked, years in social services, and the amount of cultural sensitivity training. Logistic regression was used to

PAGE 45

45 analyze the third research question. The depe ndent variable was the decision to report the vignette to Child Protective Services. The inde pendent variables were again child ethnicity, reporter race/ethnicity, reporter gender, population of the county in which the reporter works, years in social services, and the amount of cultural sensitivity training.

PAGE 46

46 Table 2-1. Descriptive statistics for participants by form (n=150) Form A Form B Form C Variable n= 52 n= 47 n= 51 Gender Male 14 11 14 Female 38 36 37 Ethnicity Black 12 7 13 White 35 34 31 Hispanic 3 1 6 Other 2 5 1 County Rural 16 12 13 Suburban 36 35 38

PAGE 47

47 CHAPTER 3 RESULTS The results of this study are presented in four sections. The first section presents the results of the descriptive statistics for all of th e variables investigated in this study. Included in this section are the results of correlation analyses between the variables investigated in this study. The following three sections present the results for each of the main hypotheses of the study. The implications of these fi ndings are discussed in Chapter 4. Descriptive Statistics Descriptive statistics illustrating the mean se verity rating of each vignette, as well as the percentage of workers who chose to report each vignette are presented in Table 3-1. Mean severity ratings, across each vignette, ranged from a low of 4.27 to a high of 6.74. Standard deviations for the 8 incidents ra nged from a low of .60 to a hi gh of 1.65. Internal consistency reliability was calculated for the 8-item overall severity scale using Cronbachs Coefficient Alpha and the following alphas were found: = .79 for all forms combined, = .84 for form 1, = .75 for form 2, and = .77 for form 3. Correlations The correlations between overall severity rating, gender, ethnicity, county, cultural sensitively training, and years in social services are given in Ta ble 3-2. The overall severity rating was found to significantly correlate with year s in social services (r = -.20, p = .01). The relationship is negative, indicati ng that participants with more tenure in social services were more likely to rate vignettes as less severe. Research Question 1 Do Child Protective Investigators rate seve rity of maltreatment differently based on victims race/ethnicity when victim's gender is controlled? This research question was

PAGE 48

48 investigated through the use of a one-way, between-subjects ANOVA. The means between groups for each vignette can be found in Table 31. The 8-item overall severity scale means for the different forms were compared for Form A/black child ( M = 5.49, SD = 0.82), Form B/white child ( M = 5.68, SD = 0.63), and Form C/No Race Given ( M = 5.49, SD = 0.79). A significant difference was not found between groups based on the ethnicity of the child reported in the vignette, F (2, 147) = 1.02, p = 0.36. Therefore, no evidence was found in support of hypothesis 1. Exploratory analyses were completed on th e 8-item vignettes separately and no significant differences were found. These resu lts can be found in Table 3-3. Research Question 2 Do Child Protective Investigators rate seve rity of maltreatment differently based on characteristics such as reporter race/ethnicity, reporter gender, amount of cultural sensitivity training of the reporter, and set ting of employment (i.e., rural versus suburban)? Multiple regression analysis was utilized in answering the second research quest ion. These results are illustrated in Table 3-4. Partial supp ort was found for this hypothesis. None of the variables investigated were found to be signi ficant predictors of the 8-item overall severity rating scale. Exploratory analyses were comple ted on each vignette separately. Since there were no apriori pred ictions, we used a more conser vative Bonferroni-corrected significance value corrected for 8 items ( p < .05/8 = .006). Years in so cial services was found to be a significant predictor of severity ratings on Vignette 2 ( b = -0.04, p = 0.00), vignette 3 ( b = 0.04, p = 0.00), and Vignette 6 ( b = -0.04, p = 0.00). The relationship between time in social services and severity rating is negative for each vignette, indicating a ge neral pattern that the longer the participants have been in their field th e less likely they are to rate the vignettes as severe.

PAGE 49

49 Research Question 3 Is there a difference in willingness to report based on the childs race/ethnicity, reporter race/ethnicity, reporter gender, amount of cultura l sensitivity training of the reporter, or the populations of the county in which the reporter works? Descriptiv e statistics for the percentage of vignettes on which a referral was made can be found in Table 3-1. Logistic regression was utilized regressing the outcome willingness to re port onto six predictors: child ethnicity (3 levels), reporter ethnicity (4 levels), reporter gender (2 levels), amount of cultural sensitivity training, years in social services and the population of the county in which the reporter works. Table 3-5 illustrates the results of the analysis and contains both b weights and odds ratios. This analysis was exploratory in na ture and did not utilize any ap riori predications. Thus, the significance value must be conservative and a Bonferroni correction was used for 8 items ( p < .05/8 = .006). Unfortunately, only five vignettes were able to be used in th is analysis due to lack of variance in the sample. For Vignettes 1 and 6, a lack of variance on the willingness to report outcome caused quasi-complete separation of data points and thus regressi on coefficients could not be reported for these vignettes. This lack of variance can be seen in the very high reporting percentages given in Table 3-1 for these vigne ttes. For Vignette 5, a ll respondents chose to report the incident, leav ing no variance to predict. Also, one predictor variable, reporter race/ethnicity, was removed from further analyses b ecause lack of sample size at certain levels of the ethnicity variable caused qua si-complete separation of data points in all 8 vignettes. Results indicate that time in so cial services was found to si gnificantly predict willingness to report for Vignette 3 ( b = -.07, p = 0.00). The relationship is negative, indicating that investigators with more time spent working in soci al services are less like ly to report abuse. However, the odds ratio of 0.93 i ndicates that individuals with higher tenure are approximately

PAGE 50

50 just as likely to report as those with lower te nure, so this effect may not have practical significance.

PAGE 51

51Table 3-1. Descriptive statistics for vignettes by form Severity mean rating Percent of respondents reporting Vignette Black child White child No race given Black child White child No race given All 8 vignettes 5.49 5.68 5.49 81% 80% 77% 1: Child wets bed and is immersed in hot water 6.46 6.55 6.26 100% 98% 88% 2: Father punches drunk teen 5.39 5.66 5.24 79% 83% 69% 3: Mother calls child evil and hits with broom 5.31 5.17 5.33 71% 66% 78% 4: Parents lock 14-year-old in his room 5.31 5.35 4.92 79% 68% 65% 5: Parents beat 11-year-old with cane and burn a mark on his arm for stealing 6.57 6.74 6.65 100% 100% 100% 6: Child mumbled, parent banged child against wall, bruising his shoulders 5.83 6.14 6.14 96% 96% 96% 7: Parents make child kneel in closet 4.77 4.94 4.67 69% 60% 59% 8: 9-year-old left alone most of day and night, parents working 4.27 4.86 4.69 52% 68% 63% Note. black child: N=52, white child: N=47, no race given: N=51

PAGE 52

52 Table 3-2. Correlation matrix Variables of interest 1 2 3 4 5 1. Overall severity rating 2. Gender -0.03 3. Ethnicity -0.06 0.10 4. Rural vs. urban county -0.04 0.05 0.06 5. CST hours 0.16 0.02 -0.10 -0.05 6. Years in social services -0.20* 0.09 0.19* 0.02 0.08 Note CST = cultural sensitivity training. p < .05.

PAGE 53

53Table 3-3. ANOVA an alyses for severity rating by form Vignette SS df MS F p Overall severity rating 1.16 2 0.58 1.02 0.36 1: Child wets bed and is immersed in hot water 2.15 2 1.08 1.33 0.27 2: Father punches drunk teen 4.47 2 2.24 1.34 0.26 3: Mother calls child evil and hits with broom 0.74 2 0.37 0.26 0.77 4: Parents lock 14-ye ar-old in his room 5.61 2 2.81 1.28 0.28 5: Parents beat 11-year-old with cane and burn a mark on his arm for stealing 0.78 2 0.39 0.88 0.42 6: Child mumbled, parent banged child against wall, bruising his shoulders 3.28 2 1.64 1.82 0.17 7: Parents make chil d kneel in closet 1.80 2 0.90 0.42 0.66 8: 9-year-old left alone most of day and night, parents working 9.27 2 4.64 2.53 0.08

PAGE 54

54Table 3-4. Regression anal yses for severity rating Vignette Black child vs. no race given White child vs. no race given Gender Respondent ethnicity: black vs. white Respondent ethnicity: hispanic vs. white Respondent ethnicity: other vs. white CST hours Rural vs. urban county Years in social services Adjusted R2 Overall severity rating -0.04 0.18 -0.01 0.16 0.15 0.12 0.00 -0.05 -0.02 0.04 1: Child wets bed and is immersed in hot water 0.23 0.30 -0.20 0.12 0.20 0.33 0.00 -0.05 0.00 -0.01 2: Father punches drunk teen 0.11 0.38 0.10 -0.07 1.15 0.56 0.00 -0.28 -0.04* 0.09 3: Mother calls child evil and hits with broom -0.14 -0.19 -0.13 0.36 0.18 0.28 0.00 0.27 -0.04* 0.06 4: Parents lock 14-year-old in his room 0.32 0.44 -0.07 0.21 -0.46 -0.52 0.00 0.10 -0.02 0.01 5: Parents beat 11-year-old with cane and burn a mark on his arm for stealing -0.07 0.08 0.04 -0.08 0.16 0.23 0.00 0.12 0.00 -0.02 6: Child mumbled, parent banged child against wall, bruising his shoulders -0.41 -0.01 0.01 0.33 0.15 0.00 0.00 0.24 -0.04* 0.12 7: Parents make child kneel in closet 0.07 0.22 0.05 -0.06 -0.25 0.08 0.00 -0.52 -0.02 0.04 8: 9-year-old left alone most of day and night, parents working -0.42 0.18 0.16 0.43 0.06 -0.03 0.00 -0.25 -0.01 -0.00 Note CST = cultural sensitivity training; values are b-weights. p < .05/8 = .006.

PAGE 55

55Table 3-5. Logistic regression an alyses for willingness to report Reporting of vignette Black child vs. no race given White child vs. no race given Gender CST hours Rural vs. urban county Years in social services Percentage of vigne ttes reported overall -0.15 0.74 -0.14 0.75 -0.11 0.80 0.00 1.00 0.16 1.38 -0.03 0.97 2: Father punches drunk teen -0.29 0.56 -0.39 0.46 0.06 1.31 -0.00 1.00 0.22 1.57 0.01 1.01 3: Mother calls child evil and hits with broom 0.33 1.94 0.33 1.94 -0.28 0.57 -0.00 1.00 -0.31 0.54 -0.07* 0.93 4: Parents lock 14-ye ar-old in his room -0.31 0.54 -0.06 0.89 -0.08 0.86 0.00 1.00 0.10 1.23 -0.04 0.96 7: Parents make chil d kneel in closet -0.27 0.58 -0.03 0.93 -0.13 0.76 0.01 1.01 0.36 2.04 -0.00 1.00 8: 9-year-old left al one most of day and night, parents working 0.21 1.53 -0.10 0.83 -0.13 0.78 0.00 1.00 0.17 1.42 0.00 1.00 Note CST = cultural sensitivity training; first values in each cell are b-weights, second values are odds ratio estimates. p < .05/8 = .006.

PAGE 56

56 CHAPTER 4 DISCUSSION This study achieved the goals of investigati ng whether (1) Child Protective Investigators rate severity of maltreatment di fferently based on victims race/ethnicity; (2) Child Protective Investigators rate severity of maltreatment differently based on characteristics such as race/ethnicity, gender, amount of cultural sensitivity training of the reporter, and setting of employment; and (3) if the child s race/ethnicity, reporter race/e thnicity, reporter gender, amount of cultural sensitivity training of the reporter, or the population of the county in which the reporter works affects the investig ators willingness to re port. While some st atistically significant results were found among the exploratory an alyses, none of the above hypotheses were supported. The results demonstrated that one of the reporter characteristics examined during exploratory analyses, years sp ent in social services, or tenure on the job, was found to significantly predict severity rati ngs across three different vigne ttes as well as the decision to report on one vignette. The relationship was nega tive, indicating that participants with more tenure in social services were more likely to rate vignettes as less severe and were less likely to report the abuse. This reporter ch aracteristic has not received much attention in the literature and thus was not one of central focus before co mpleting this study. Research completed by Crenshaw, Crenshaw and Lichtenberg (1995) inve stigated the reporting behaviors of teachers, counselors, school psychologists, and principles. The results of that study showed no relationship between educator tenur e and reported severity among five scenarios of child abuse. Similarly, Warner-Rogers, Hansen, and Spieth (1996) who sampled third and fourth year medical students, found no influence of experience. Uniquely, the current sample was comprised of a large range of experience le vels, ranging from 1 month on the job to 38 years of experience.

PAGE 57

57 The present findings could indicat e that Child Protective Investig ators become desensitized to situations involving chil d maltreatment and consequently rate them as less severe over time. Along with results found in Ashton (1999), Ha nsen et al. (1997), and Warner-Rogers, Hansen, & Spieth (1996), who utilized medical st udents, psychologists, social workers, and graduate students, the current study found no evidence of gender differences among reported severity ratings. Previous studies that have not found gender to be a significant predictor of severity ratings have utilized overwhelmingly fe male and Caucasian populations, all calling for a more culturally diverse sample. The sample for the present study was more diverse than samples have been in previous studies. Additionally, the significance of this resear ch is its focus on Child Protective Investigators, the individuals deemed with the responsibility of investigating and intervening in cases of child abuse. Thus, this finding may further indicate that gender differences may not exist within the system a nd should be used to further evaluate evidence among this population. Similarly, reporter ethnicity was not found to be a significant predictor of severity ratings. Evidence that reporter ethnicity does not affect judgments co ncerning child maltreatment is inconsistent with research completed by As hton (2004) and Hansen et al. (1997). One explanation for this inconsistency may be that previous studies that ha ve found differences in reporting rates based on reporter ethnicity have only sampled ma ndated reporters and not Child Protective Investigator s, the individuals responsible fo r verification. Thus, one might hypothesize that mandated reporters, and not Child Protective Investigators, are responsible for the overrepresentation of minority children in the reporting rates. Additi onally, reporter ethnicity had to be removed as a predictor variable for willingness to report due to lack of sample size at certain levels of ethnicity. Thus while this study aimed to uti lize a more diverse sample, there

PAGE 58

58 were not enough participants within each ethnic ity to analyze whether or not it significantly predicted willingness to report. The population of the county in which one wo rks was not found to significantly predict reporting behavior. This result is in contrast to previous research conducted by Webster, OToole, OToole, & Lucal (2005) who found an effect for reporting behavior among teachers in rural and suburban districts, with rural teachers being less likely to report. Traditionally, rural areas have a higher percentage of families livin g in poverty than that of urbanized areas. Researchers have hypothesized that people living in rural areas may have more well-developed social networks and a strong sense of community and thus may rely on informal channels of support when dealing with child maltreatment (S chuck, 2005). With such a limited number of Child Protective Investigators be ing employed within rural countie s, it made recruiting an equal number of participants for this study impossible. Future rese arch should focus on recruiting an equal number of participants from each area to further investigate this variable as a possible predictor of reporting behavior. Finally, no evidence was found to support that th e ethnicity of the child depicted in the scenarios affects the severity ratings or decision to report among the Child Protective Investigators sampled. The finding that race of th e child does not affect judgments of abuse is consistent with previous research complete d by Egu & Weiss (2003) and Turbett & OToole (1983). In the previous studies that did find the race of the child as a significant factor in reporting, the populations sampled included physicia ns, clinical psycholo gists, and social workers (Nalepka, OToole & Turbett, 1981; Ha nsen et al., 1997). While being mandated reporters, these professionals have not had the specific training that actual Child Protective Investigators complete. This finding provides evidence against current theory concerning the

PAGE 59

59 overrepresentation of minority chil dren within the child welfare system. Barth (2005) maintains that the race of the child play s a role in every decision making point throughout the child welfare system. The current study found that the ethnicit y of the child was one of the only characteristics investigated that did not affect reporter behavior. Additionally the reported am ount of cultural sensitivity training was not found to be a signific ant predictor of severity ratings or willingness to report. Thus, it did not appear that the amount of cultural sensitivity training received affected ones judgment of child maltreatment. Limitations First, it is not known whether the present findings generalize to other locations. Census data suggest that the demographic characteristics of residents of Florida are similar to the nation as a whole; however, Florida ha s one of the highest child victimization rates in the nation (Schuck, 2005). Additionally, only seventeen of Floridas sixtyseven counties were sampled in this study. Furthermore, the sample was not random ly selected and utilized the first 150 Child Protective Investigators agreeing to participate within the seve nteen most immediate counties surrounding the University of Florida. The present study utilized vignette and self -report methodology. Real-life situations contain many additional facts and ambiguities that affect persona l judgments about suspicion and reporting of child maltreatment. Anecdotally, a large percentage of the participants reported feeling uncomfortable with making a decision based on such ambiguous information. Although use of hypothetical cases of ma ltreatment presented in vignettes permit the systematic manipulation of case variables, not all important case variables can be included in one study. Thus, a limitation of the current study was using gender as a control in order to isolate the ethnicity of the child. The child in the scenarios was always male, thus limiting the

PAGE 60

60 generalizability of the study. Future research should replicate this study investigating if an effect of child gender is present within this population. While no significant effects of child ethnic ity were found, particip ants demonstrated a tendency to give less severe ratings on the vigne ttes depicting the victim as African-American. This trend is consistent with the results found in the literature. While a between-subjects design brings certain advantages to this study (e.g., ra ndom assignment, reductio n of effects of demand characteristics and social desirability), it may ha ve masked true differences because participants were reluctant to answer honestly for fear of being perceived as racist when all the vignettes depicted minority children. Other studies, on the othe r hand, utilized withinsubjects designs that may overestimate effects. Thus, a confounding issu e in completing research of this nature is getting accurate and honest responses from worker s within the field. Anecdotally, many of the participants in this study expre ssed concern upon receiving Form A with the children depicted as black. This may be representative of what an emotionally charged issue culture is within the child welfare system. Future research should work on finding an alternate method of gathering reporter data which allows the re spondent to answer honestly and without the influence of social desirability. Additionally, only five vignettes we re able to be used in analysis of the third hypothesis due to lack of variance in the sample on the wi llingness to report outcome for these vignettes. For vignette 5, all respondents chos e to report the incident, leaving no variance to predict. Also, one predictor variable, ethnicity, was removed from further analyses because lack of sample size at certain levels of the ethnic ity variable. While years in so cial services was found to be statistically significant, only a small amount of the variance in predicting severity ratings and willingness to report was accounted for by this variable. Future research should focus on

PAGE 61

61 increasing the effectiveness of th e vignettes used in the study. The present study yielded results which showed a restriction in range of the resp ondents reporting of seve rity levels. Future studies should utilize vignettes which allow the respondents to use the lower values of the scale, creating more variab ility in responses. Conclusion Despite its limitations, this study provides important information about a previously unstudied population. The significance of this research is its focus on Child Protective Investigators, the individuals responsible fo r investigating and veri fying cases of child maltreatment. While CPS workers comprise a negl ected source of data, their reporting behaviors have gained much attention and speculation am ong researchers. There have been increasing concerns about the overrepresent ation of minority children within the child welfare system in this country. Researchers have spent mu ch time and attention conducting studies with undergraduates, graduate students, teachers, physicians, and social workers and have extrapolated results to hypothesize reporting behavi ors of CPS workers. In his theoretical model pertaining to this overrepresentati on, Barth (2005) maintains that ag ency characteristics allow the race of the child to play a role in every decision making point throughout the child welfare system. More specifically, factors such as staffing patterns, culturally incompetent staff, institutionalized racism, and inadequate dura tion or configuration of services (p.36) are provided as possible contributing characteristics. The present re search study provides direct evidence that the race of the child do es not play a role in the severi ty ratings or decision to report among actual Child Protective Investigators. The present study also adds to our understanding of what factors may influence identification and reporting of maltreatment. Tenure on the job ha s received little to no previous attention in the literature as a reporter characteri stic of interest. Thus, this characteristic should

PAGE 62

62 be the focus of future research among this populati on to help in further u nderstanding the effect it has on reporting behavior. The findings presented here should be viewed as exploratory in nature and warrant replication w ith larger groups of investigat ors. Readers are reminded the demographic variables chosen for this study repres ent only a part of an individuals psychosocial characteristics. Further research is needed to better understand the underlying factors within cultures that affect differe nces in reporting.

PAGE 63

63 APPENDIX A DEMOGRAPHIC QUESTIONNAIRE Answer the following questions: 1) Please circle the grou p that best describes your own cultural identification? African-American or Black Caucasian, or White Hispanic, or Latino Some other group (specify) ____________________________________ 2) What is your gender? Male or Female 3) What country were you born in? _______________________ 4) If you were not born in the U.S., how long have you lived here? __________ 5) What is your religious affiliation? _________________ 6) What type of work does (did) your father do? Professional Skilled Semi-skilled Unskilled 7) What is your current age? __________ 8) Do you have a childhood hi story of harsh discipline? _____________ 9) What is your marital status? Single Married Divorced Widowed Living with Significant Other 10) Do you have any children? Yes (if yes, how many?) ________ No 11) What is your highest level of education?

PAGE 64

64 High School Diploma Some College Bachelors Degree Masters Degree Ph.D. 12) What field is your highe st degree in? (e.g., social work, education, etc.) _____________________________ 13) How long have you worked in social serv ices (include time with present employer as well as other agencies)? ________ years ________ months 14) Please indicate your current st atus with the Department of Children and Familes: Child Protective Investigator Child Protective Service Worker Foster Care Worker Adoption Worker Licensing Worker Other (explain)__________________ 15) What ethnicity/race is the majori ty of the clientele you serve? _____________ 16) Which best describes the county in which you work? Rural Urban/Suburban 17) How much cultural sensitivity training ha ve you received (include time with present employer as well as other agencies)? ________________ hours 18) Please identify where you received the above training (Circl e all that apply) Preservice: Undergraduate Entire course Part of a course Graduate School (Before current employment) Entire course Part of a course Inservice (During current employment)

PAGE 65

65 Workshop Conference Other __________________

PAGE 66

66 APPENDIX B VIGNETTES Please imagine that you are a case worker appr oached by a teacher who discloses the following information. 1. During a parent-teacher conferen ce, a mother discloses to a 2nd grade teacher that her eight-year-old son wets the bed occasionally. The last time he wet the bed his mother immersed him in a tub of very hot water. Hi s mother explained that the child will not wet the bed again if he associates bed wetting with the punishment. How serious do you feel the above incident is on a Likert scale of seriousness From 1 to 7, with being not serious and being very serious? 1 2 3 4 5 6 7 How would you recommend that the educat or respond given the above scenario? (1) no action, nothing needs to be done about the situation (2) encourage the family to seek professional help (3) notify the child protection agency to investigate the situation 2. A high school teacher found out that a 17-ye ar-old boy in her class came home drunk and began to argue with his mother last weekend. His father came into the room just as he yelled an obscene remark at his mother. His father punched him in the face with his fist. How serious do you feel the above incident is on a Likert scale of seriousness From 1 to 7, with being not serious and being very serious? 1 2 3 4 5 6 7 How would you recommend that the educat or respond given the above scenario? (1) no action, nothing needs to be done about the situation (2) encourage the family to seek professional help (3) notify the child protection agency to investigate the situation 3. During outside play, a six-year-old boy re veals to his teacher that her parents are separated. The child lives with his mother and occasionally visits his father. Whenever he returns from a visit with his father and talks about the visit, his mother tells him he is evil and hits him with a broom. How serious do you feel the above incident is on a Likert scal e of seriousness From 1 to 7, with being not serious and being very serious? 1 2 3 4 5 6 7

PAGE 67

67 How would you recommend that the educat or respond given the above scenario? (1) no action, nothing needs to be done about the situation (2) encourage the family to seek professional help (3) notify the child protection agency to investigate the situation 4. A 14-year-old boy tells his history teacher th at his parents found out that he is dating outside of his race, which they dont approve of. He has lied about meeting his girlfriend and he has snuck out of the house on several occasions to see her. When his parents caught him trying to sneak back into the house one morning, they sent him to his room and locked him in for two days, allowi ng him out only to go to the bathroom. How serious do you feel the above incident is on a Likert scale of seriousness From 1 to 7, with being not serious and being very serious? 1 2 3 4 5 6 7 How would you recommend that the educat or respond given the above scenario? (1) no action, nothing needs to be done about the situation (2) encourage the family to seek professional help (3) notify the child protection agency to investigate the situation 5. An 11-year-old boy stole candy from a store. He had been caught st ealing before. When his parents found out that he had been stea ling again, they beat him with a switch and burned a mark on his arm. They said the ma rk would remind him not to steal again. He told his P.E. teacher on Monday morning wh ile he was dressing out for gym class. How serious do you feel the above incident is on a Likert scale of seriousness From 1 to 7, with being not serious and being very serious? 1 2 3 4 5 6 7 How would you recommend that the educat or respond given the above scenario? (1) no action, nothing needs to be done about the situation (2) encourage the family to seek professional help (3) notify the child protection agency to investigate the situation 6. The parents of an 8-year-old boy want him to answer clearly when spoken to and have warned him in the past about mumbling. Recently when his mother asked him a question, he mumbled the answer under his breath. The moth er banged the child against the wall, bruising his shoulders. He told th e speech pathologist what happened during their weekly session. How serious do you feel the above incident is on a Likert scale of seriousness

PAGE 68

68 From 1 to 7, with being not serious and being very serious? 1 2 3 4 5 6 7 How would you recommend that the educat or respond given the above scenario? (1) no action, nothing needs to be done about the situation (2) encourage the family to seek professional help (3) notify the child protection agency to investigate the situation 7. A family recently arrived in the metropolitan area from a rural location. They are very strict with their ten-ye ar-old son. They want him to go to school and come right home. Recently their son went to a school friends house instead of coming straight home. His parents disciplined him in their usual way, whic h is to make him kneel in the closet for several hours. The parents believe the closen ess of the closet and the discomfort of kneeling for long periods of time reinforce the household rules. How serious do you feel the above incident is on a Likert scale of seriousness From 1 to 7, with being not serious and being very serious? 1 2 3 4 5 6 7 How would you recommend that the educat or respond given the above scenario? (1) no action, nothing needs to be done about the situation (2) encourage the family to seek professional help (3) notify the child protection agency to investigate the situation 8. Both parents work long hours; they leave very early in the morning and come home late at night. Their nine-year-old son is ofte n on his own. The boy gets himself ready for school in the morning and lets himself in afte r school. The parents tell their son to eat food prepared and left in the refrigerator, wa rming it up if he wants. He usually eats it cold. He goes to bed by himself because his parents are usually not back by his bedtime. How serious do you feel the above incident is on a Likert scale of seriousness From 1 to 7, with being not serious and being very serious? 1 2 3 4 5 6 7 How would you recommend that the educat or respond given the above scenario? (1) no action, nothing needs to be done about the situation (2) encourage the family to seek professional help (3) notify the child protection agency to investigate the situation

PAGE 69

69 APPENDIX C INFORMED CONSENT Protocol Title: The effects of reporter and child char acteristics on the identification and willingness to report child maltreatment Please read this consent document carefully be fore you decide to participate in this study. Purpose of the research study: The purpose of this study is to examine the eff ects of reporter and chil d characteristics on the identification and willingness to report child maltreatment. What you will be asked to do in the study: You will be asked to complete a packet that cont ains eight vignettes depicting different scenarios followed by a demographic questionnaire. No id entifying information (name, social security number, etc.) should be included in order to ensure confidenti ality. Each vignette depicts a fictitious description of problematic parental be haviors. You will then be asked to read each vignette and rate the information on a Likert scale of seriousness from 1 to 7, with being not serious and being very serious. After you rate the seriousness of the scenario, you will then be asked to indicate how you would tell the educator to respond by selecting one of three options: (1) no action, nothing needs to be done a bout the situation; (2) en courage the family to seek professional help; or (3) notify the child protection agency to investigate the situation. Time required: 15 minutes Risks and Benefits: We do not anticipate any benefits or risks associated with participation in this study. Compensation: No compensation will be given fo r participation in this study. Confidentiality: No identifying information will be gathered duri ng data collection. Your demographic data and vignette response sheet will be kept anonymously and assigned a code number to ensure confidentiality. Voluntary participation: Your participation in this study is completely voluntary. There is no penalty for not participating.

PAGE 70

70 Right to withdraw from the study: You have the right to withdraw from th e study at anytime without consequence. Whom to contact if you have questions about the study: Danielle Madera, M.Ed., School Psychology Gr aduate Student, University of Florida, Gainesville, FL 32608; (352) 334-1300. Whom to contact about your rights as a research participant in the study: UFIRB Office, Box 112250, University of Flor ida, Gainesville, FL 32611-2250; ph 392-0433. Agreement: I have read the procedure described above. I volunt arily agree to pa rticipate in the procedure and I have received a copy of this description. Participant: _____________________________________ Date: _________________ Principal Investigat or: _____________________________Date: _________________

PAGE 71

71 REFERENCE LIST American Psychological Association. (2002). Guidelines on multicultural education, training, research, practice, and organization al change for psychologists. Retrieved December 12, 2005, from http://www.apa.org/pi/multiculturalguidelines/homepage.html Ards, S., Chung, C., & Myers, S. L. (1998). The effects of sample selection bias on racial differences in child abuse reporting. Child Abuse & Nelgect, 22, 103. Ashton, V. (1999). Worker judgements of serious ness about and reporting of suspected child maltreatment. Child Abuse & Neglect, 23 (6), 539. Ashton, V. (2004). The effect of personal char acteristics on reporting child maltreatment. Child Abuse and Neglect, 28, 985. Barth, R. P. (1997). Effects of age and race on the odds of adoption versus remaining in longterm out-of-home care. Child Welfare, 76 285. Barth, R. P. (2005). Child welfare and race: Models of disproportionality. In D. M. Derezotes, J. Poertner, & M. F. Testa., Race matters in Child Welfare (pp. 25). Washington, DC: Child Welfare League of America, INC. Baumrind, D. (1995). Child maltreatment and optimal ca regiving in social contexts New York, NY: Garland Publishing, Inc. Behl, L.E., Crouch, J. L., May, P. F., Valente, A. L., & Conyngham, H. A. (2001). Ethnicity in child maltreatment research: A content analysis. Child Maltreatment, 6 (2), 143. Belsky, J. (1980). Child maltreatme nt: An ecological integration. American Psychologist,35 (4), 320. Bronfenbrenner, Y. (1979). The ecology of family development Cambridge, MA:Harvard University Press. Bui, K. & Takeuchi, D. T. (1992) Ethnic minority adolescents a nd the use of community mental health care services. American Journal of Community Psychology, 20 403. Champion, K., Shipman, K., Bonner, B., Hensley, L., Howe, A. (2003). Child maltreatment training in doctoral programs in clinical, counseling, and school psychology: Where do we go from here? Child Maltreatment, 8, 211. Cohen, J. A., Deblinger, E., Mannarino, A. P., & Arellano, M. A. (2001). The importance of culture in treating abused and neglecte d children: An empirical review. Child Maltreatment, 6 (2), 148. Coohey, C. (2001). The relationship between fa milism and child maltreatment in Latino and Anglo families. Child Maltreatment, 6 (2), 130.

PAGE 72

72 Crenshaw, W. B., Crenshaw, L. M., & Lichtenb erg, J. W. (1995). When educators confront child abuse: An analysis of the decision to report. Child Abuse & Neglect, 19 (9), 1095 1113. Deater-Deckard, K., Dodge, K. A., Bates, J. E., & Pettit, G. S. (1996). Physical discipline among African-American and European-American moth ers: Links to children's externalizing behaviors. Developmental Psychology, 32 1065. Drake, B., & Pandey, S. (1996). Understandin g the relationship between neighborhood poverty and specific types of child maltreatment. Child Abuse & Neglect, 20 (11), 1003. Dubowitz, E. (1999). Neglected children: Research, practice, and policy. Thousand Oaks, California: Sage Publications. Egeland, B. A history of abuse is a major risk f actor for abusing the next generation. In: R. J. Gelles, D. R. Loseke, Eds. Current controversies on family violence (pp. 197) Thousand Oaks, CA, Sage, 1993. Egu, C. L., & Weiss, D. J. (2003). The role of race and severity of abuse in teachers recognition or reporting child abuse. Journal of Child and Family Studies, 12 (4), 465. Ferrari, A. M. (2002). The impact of culture upon child rearing practic es and definitions of maltreatment. Child Abuse & Neglect, 26 793. Fluke, J. D., Yuan, Y. T., Hedderson, J., & Curtis P. A. (2003). Dispropo rtionate representation of race and ethnicity in child maltreatm ent: Investigation andvictimization. Children and Youth Services Review, 25 359. Fontes, L. A., Cruz, M., & Tabachnick, J. (2001). Views of child sexual abuse in two cultural communities: An exploratory study among African-Americans and Latinos. Child Maltreatment, 6 (2), 103. Garbino, J., & Sherman, D. (1980). High-risk neighborhoods and high-risk families:The human ecology of child maltreatment. Child Development, 51 (1), 188. Garland, A. F., & Beinger, B. A. (1997). Racial/e thnic differences in court referred pathways to mental health services for children in foster care. Children and Youth Services Review, 19 (8), 651. Gelles, R. J. (1992). Poverty and violence toward children. American Behavioral Scientist, 35 (3), 258. Green, J. W. (1982). Cultural awareness in the human services Englewood Cliffs, NJ:PrenticeHall.

PAGE 73

73 Hansen, D. J., Bumby, K. M., Lundquist, L. M., Cha ndler, R. M., Le, P. T., & Futa, K. T. (1997). The influence of case and professional vari ables on the identification and reporting of child maltreatment: A study of licensed ps ychologists and certifiedmasters social workers. Journal of Family Violence, 12 (3), 313. Hines, A. M., Lemon, K., Wyatt, P., & Merd inger, J. (2004). Factors related to the disproportionate involvement of children of color in the ch ild welfare system: A review and emerging themes. Children and Youth Services Review, 26, 507. Hinson, J., & Fossey, R. (2000). Child abuse: Wh at teachers in the s know, think, and do. Journal of Education for Student s Placed at Risk, 5(3), 251. Horton, C., & Cruise, T. (2001). Child abuse & neglect: The schools response. New York, NY: The Guilford Press. Kempe, C. H., Silverman, F. N., Steele, B. F ., Droegmueller, W. & Silver, H. K. (1962). The Battered Child Syndrome. Journal of the American Medical Association 181, 17. Kenny, M. (2001). Child abuse reporting: Teachers perceived deterrents. Child Abuse & Neglect, 25, 81. Korbin, J. E. (2002). Culture and child maltr eatment: Cultural competence and beyond. Child Abuse & Neglect, 26 637. Korbin, J. E., Coulton, C. J., Lindstrom-Ufuti, H., & Spilsbury, J. (2000) Neighborhood views on the definition and etiology of child maltreatment. Child Abuse & Neglect, 24 (12), 1509. Levesque, R. J. (2000). Cultural eviden ce, child maltreatment, and the law. Child Maltreatment, 5 (2), 146. Lu, Y. E., Landsverk, J., Ellis-Macleod, E., Ne wton, R., Ganger, W., & Johnson, I. (2004). Race, ethnicity, and case outcomes in child protective services. Children and Youth Services Review, 26 447. Maynard, C., & Wiederman, M. (1997). Undergra duate students perceptions of child sexual abuse: Effects of age, sex, and gender-role attitudes. Child Abuse & Neglect, 21 (9), 833. McCabe, K. M., Yeh, M., Hough, R. L., Landsver k, J., Hurlburt, M. S., Culver, S. W., & Reynolds, B. (1999). Racial/eth nic representation across five public sectors of care for youth. Journal of Emotional and Behavioral Disorders, 7 72. Milner, J. S. (1998). Individual and family char acteristics associated with intrafamilialchild physical and sexual abuse. In P.K. Trickett & C.J. Schedllenbach (Eds.), Violence against children in the family and the community (pp. 141). Washi ngton, DC: American Psychological Association.

PAGE 74

74 Morrow, K. B., & Sorrell, G. T. (1989). Factor s effecting self-esteem, depression and negative behaviors in sexually abused female adolescents. Journal of Marriage and the Family, 51 677. Nybell, L. M., & Gray, S. (2004). Race, place, space: Meanings of cultural competence inthree child welfare agencies. Social Work, 49 (1), 17. Rao, K., DiClemente, R. J., & Ponton, L. E. (1992) Child sexual abuse in Asians compared with other populations. Journal of the American Academy of Child Adolescent Psychiatry, 31 (5), 880. Scannapieco, M., & Connell-Carrick, K. (2005). Understanding child maltreatment. New York, NY: Oxford University Press. Schuck, A. M. (2005). Explaini ng black-white disparity in ma ltreatment: Poverty, femaleheaded families, and urbanization. Journal of Marriage and Family, 67 543. Sullivan, P. M., & Knutson, J. F. (2000). Malt reatment and disabilities: A population-based epidemiological study. Child Abuse & Neglect, 24, 1257. Terao, S. Y., Borrego, J., & Urquiza, A. J. (2001) A reporting and response model for culture and child maltreatment. Child Maltreatment, 6 (2), 158. U.S. Department of Health and Human Se rvices, Administration on Children, Youth and Families. Child Maltreatment 2004 (Washington, DC: U.S. Government Printing Office, 2006). U.S. Department of Health and Hu man Services, Childrens Bureau. Child Maltreatment 1998: Reports from the States to the Natio nal Child Abuse and Neglect Data System (NCANDS). (Washington, D.C.: U.S. Government Printing Office, 2000). Warner-Rogers, J. E., Hansen, D. J., & Spie th, J. E. (1996). The influence of case and professional variables on id entification and reporting of physical abuse: A studywith medical students. Child Abuse & Neglect, 20 (9), 851. Webster, S. W., OToole, R., OToole, A. W., & Lucal (2005). Overreporting and Underreporting of child abuse: Teacher s use of professional discretion. Child Abuse & Neglect, (29) 1281. Wissow, L. S. (2001). Ethnicit y, income, and parenting contexts of physical punishment in a national sample of families with young children. Child Maltreatment, 6 (2),118.

PAGE 75

75 BIOGRAPHICAL SKETCH Danielle Dawn Madera wa s born in Valley Stream, NY on May 5, 1978, the daughter of Demetrius and Christine Madera. Danielle gradua ted from the University of Florida in May 2000, with a B.S. in psychology. In August 2000, she began working as a Family Service Counselor for the Department of Children and Fa milies in St. Petersburg, FL. After a year of experience working with child maltreatment, she d ecided to begin her gra duate education. In August 2001, Danielle entered the School Ps ychology Doctoral Training Program at the University of Florida. At the University of Fl orida, Danielle pursued research interests in child maltreatment. She received her Master of Education degree from the University of Florida in May 2005. Danielle then moved on to comp lete her advanced practicum placement with the University of Florida Child Protection Team After completing her coursework, Danielle obtained an APPIC internship at the Childrens Assessment Center in Houston, TX, where she works providing therapy and psychologi cal services to sexually abused children and their non-offending caregivers. Having successf ully defended her dissertation, Danielle graduated in May 2007 with her Doctor of Phil osophy degree from the University of Florida. She hopes to continue working with victims of child maltreatment and continue her research efforts in that field.


Permanent Link: http://ufdc.ufl.edu/UFE0019803/00001

Material Information

Title: Effects of Reporter and Child Characteristics on the Identification and Willingness to Report Child Maltreatment
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: UFE0019803:00001

Permanent Link: http://ufdc.ufl.edu/UFE0019803/00001

Material Information

Title: Effects of Reporter and Child Characteristics on the Identification and Willingness to Report Child Maltreatment
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: UFE0019803:00001


This item has the following downloads:


Full Text





EFFECTS OF REPORTER AND CHILD CHARACTERISTICS ON THE IDENTIFICATION
AND WILLINGNESS TO REPORT CHILD MALTREATMENT


















By

DANIELLE D. MADERA


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA INT PRACTICAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

2007






































O 2007 Danielle D. Madera









ACKNOWLEDGEMENTS

I want to extend the highest measure of gratitude to a small few of the many family,

friends, and colleagues who have helped me complete this great journey in scholarship. Let me

begin by thanking my mother, Christine Madera, and father, Demetrius Madera, who instilled in

me the drive, motivation, and pride necessary to successfully complete one's doctorate. I would

like also to thank Emily Hunter for her invaluable statistical assistance and collaboration

throughout this process. On this same front, a special thanks goes out to Susan Eichler of the

Hernando County Department of Children and Families for facilitating the use of her Child

Protective Investigators and assisting in the data gathering process.

A special thanks also goes out to each of the four members of my dissertation committee,

Dr. Tina Smith-Bonahue, Dr. Diana Joyce, Dr. Jaime Algina, and Dr. William Conwill, who

made unique and positive contributions to the present proj ect. I appreciate you all more than I

can say and hope to repay some small measure of what you have given me by always using my

degree to help others as you have helped me.












TABLE OF CONTENTS


page

ACKNOWLEDGE MENT S ............_...... ...............3....


LI ST OF T ABLE S ............ ..... .__ ...............6...


AB S TRAC T ......_ ................. ............_........7


CHAPTER


1 LITERATURE REVIEW ................. ...............9...............


Ecological Model of Child Maltreatment ......_._ ................ ........._._ ....... 1
Ontogenic Development ..........._...__.......... ...............10.....
M icrosy stem ..........._..._ ................. 11...._.._.....
Exo sy stem ....._ .. ................. ........._............1
M acro sy stem ................. ...............14................
Culture, Race, and Ethnicity ................. ...............15......... .....
Culture and Child Maltreatment ................. ...............15......... .....
Focus in the Literature ....._.. ................. ........._.. ........1
Effects of Culture on Treatment ................. ...............18.._.._._ ....
Effects of Culture on Child Maltreatment .............. ...............20....
Defining Child Maltreatment............... ..............2
Parenting Practices............... .. ..............2
Reporting Child Maltreatment ........_.................. ...............28......
Disproportionality in Child Maltreatment .............. ...............34....
Model s of Disproportionality .................... ...............3


2 M ETHOD .............. ...............41....


Participants .............. ...............41....
M measures ................. ........... ...............42.......

Demographic questionnaire............... .............4
Vignettes ................. ...............42.................
Procedure .............. ...............43....

A analysis .............. ...............44....

3 RE SULT S .............. ...............47....


Descriptive Statistics .............. ...............47....
Correl ati ons ................. ...............47........... ...
Research Question 1 .............. ...............47....
Research Question 2 .............. ...............48....
Research Question 3 .............. ...............49....














4 DI SCUS SSION ............. ...... .__ ...............56..


Limitations ............. ...... .__ ...............59...
Conclusion ............. ...... ...............61...


APPENDIX


A DEMOGRAPHIC QUESTIONNAIRE ............... .............6


B VIGNETTES .............. ...............66....


C INFORMED CONSENT ................. ...............69................


REFERENCE LIST .............. ...............71....


BIOGRAPHICAL SKETCH .............. ...............75....










LIST OF TABLES

Table page

2-1 Descriptive statistics for participants by form (n=150) ................. ................. ......46

3-1 Descriptive statistics for vignettes by form .........._._ ......_.. ......__ ..........5

3-2 Correlation matrix .............. ...............52....

3-3 ANOVA analyses for severity rating by form ................. ...............53..............

3-4 Regression analyses for severity rating ................. ...............54...............

3-5 Logistic regression analyses for willingness to report .............. ....................5









Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

EFFECTS OF REPORTER AND CHILD CHARACTERISTICS ON THE IDENTIFICATION
AND WILLINGNESS TO REPORT CHILD MALTREATMENT

By

Danielle D. Madera

May 2007

Chair: Tina Smith-Bonahue
Major: School Psychology

This study examined the effects of reporter and child characteristics on the identification

and willingness to report child maltreatment among currently employed Child Protective

Investigators. The subjects consisted of 150 Child Protective Investigators employed in

seventeen different counties in North Central Florida. All participants completed a demographic

questionnaire along with eight different vignettes depicting problematic parental behavior. The

vignettes were presented in three different forms, Form A depicted the child victim as black,

Form B as white, and Form C did not provide a description of the child's ethnicity. The

participants were then asked to read each vignette and rate the information on a Likert scale of

"seriousness" from 1 to 7, with "1" being not serious and "7" being very serious. After the

respondents rated the seriousness of the scenario, they were then asked to indicate how they

would tell the educator to respond by selecting one of three options: (1) no action, nothing needs

to be done about the situation, (2) encourage the family to seek professional help, or (3) notify

the child protection agency to investigate the situation. The research questions were analyzed

using a one-way, between subj ects analysis of variance, multiple regression, and logistic

regression. Findings did not support that Child Protective Investigators rate scenarios of









maltreatment differently based on the child's ethnicity. Partial support was found for reporter

characteristics affecting the overall severity rating and willingness to report child maltreatment.

More specifically, years in social services of the reporter was found to significantly predict the

overall severity ratings given to several vignettes as well as willingness to report.









CHAPTER 1
LITERATURE REVIEW

Despite decades of attention and numerous prevention efforts, child abuse and neglect

remains a serious social problem. An estimated 3 million children were alleged to have been

abused or neglected and received investigations or assessments by state and local child protective

service agencies during the Federal fiscal year 2004 (U. S. Department of Health and Human

Services, 2006). Out of those 3 million children, an estimated 872,000 were determined to be

victims of child abuse or neglect. More than 60% of child victims were neglected by their

parents or other caregivers. Approximately 18% were physically abused, 10% were sexually

abused, and 7% were emotionally maltreated.

Child maltreatment itself was not identified until 1962 when Dr. Henry C. Kempe first

introduced what he termed "battered child syndrome" (Kempe, Silverman, Steele,

Droegemueller, & Silver, 1962). Kempe's groundbreaking work helped to raise awareness and

legislation in the fight against child abuse and neglect. Kempe coined the term "battered child

syndrome" to describe "a clinical condition in young children who received serious physical

abuse, generally from a parent or foster parent." Furthermore, Kempe stated that the syndrome

should be considered in any child exhibiting abuse or neglect in the presence of a discrepancy

between the explanation and clinical findings. Kempe also suggested that alcoholism,

promiscuity, unstable marriages, and criminal activity are factors that may increase the risk for

"battered child syndrome." Since that time, many researchers have attempted to create a model

complex enough to capture the multifaceted nature of child maltreatment. This paper reviews

the literature in order to provide a theoretical model for conceptualizing child maltreatment, as

well as the specific factors related to the complex social, environmental, and cultural variables

that underlie child abuse and neglect.









Ecological Model of Child Maltreatment

The most comprehensive and widely-accepted model in explaining the etiology of child

maltreatment is that proposed by Belsky (1980), a hybrid of the model first introduced by

Bronfenbrenner (1979), to explain the ecological perspective of human development. Belsky's

addition of a fourth level, the ontogenic level, in describing the context of child maltreatment

differentiates it from Bronfenbrenner's earlier work. Belsky's model describes four

interconnected levels: (1) ontongenic, (2) microsystem, (3) exosystem, and (4) macrosystem.

Within this framework, understanding child maltreatment is contingent on examining the

relationship and interaction of all four levels; examining any one level in isolation yields an

inconclusive explanation. Moreover, each level is influenced by the family's culture

(Scannapieco & Connell-Carrick, 2005) suggesting that exploration of the cultural context of the

family is necessary to understand any child or parent behavior.

Ontogenic Development

In Belsky's (1980) description of ontogenic development, the investigation of the

childhood history of an abusive parent is necessary for developing a more comprehensive

understanding of how their parenting behaviors may have been shaped and reinforced. Previous

research has shown that a childhood history of abuse alone does not determine whether victims

will ultimately abuse their own children. While approximately 70% of parents who abuse

children were neglected or abused in their childhood (Egeland, 1993), other factors have been

identified to contribute to this high statistic. Parents with a history of childhood abuse, reported

feeling depressed and angry over their childhood, having substance abuse problems, as well as

having limited parenting skills. Thus, a childhood history of abuse alone will not result in

becoming an abusive parent, but is a crucial factor in understanding the etiology of child









maltreatment and how it may predispose parents to react to certain stimuli in the microsystem

and exosystem (Scannapieco & Connell-Carrick, 2005).

Microsystem

Consistent with Bronfenbrenner' s model, in Belsky's model, the Microsystem is composed

of the family unit and the immediate context in which the abuse occurs. This level of the

ecological model of abuse incorporates the family, characteristics of the parent, and

characteristics of the child (Belsky, 1980). Many researchers have investigated parent and child

characteristics that are related to an increased risk of abuse and neglect. For example,

psychological characteristics such as low self-esteem and poor impulse control have been shown

to exist among parent perpetrators (Dubowitz, 1999; Milner, 1998). Additionally, parenting

skills have been a noted deficiency in many offending parents. Abusive parents are more likely

to utilize punishment, threats, and coercion and less likely to use reasoning and positive

reinforcement (Milner).

While the responsibility for abuse should never fall on a child, research has shown that

some child characteristics increase the likelihood that a child will suffer maltreatment at the

hands of their caregiver. Child characteristics such as difficult temperament and disabilities

(e.g., learning disability, speech or language delay, emotional/behavioral problems) have been

shown to leave children at an increased risk of experiencing maltreatment (Sullivan & Knutson,

2000). Thus, one could imagine how the combination of a child with special needs and a parent

with poor impulse control or other limitations could create an interactive relationship with an

increased risk of abusive parental responses.

Exosystem

Belsky (1980) describes the exosystem as the larger social structures that encompass the

family unit incorporated with the mesosystem described in Bronfenbrenner' s original model.









Research has indicated many family characteristics strongly correlate with the occurrence of

child maltreatment. While child maltreatment transcends all socioeconomic levels, poverty has

been the most noted risk factor. Sedlack and Broadhurst (1996) found that family income was

the strongest correlate of incidence among all categories of maltreatment. Additionally, children

from lower income families had the highest likelihood of experiencing neglect and serious injury

or death. Poverty may be related to such factors as unemployment, lack of education, substance

abuse, and social isolation, all of which contribute to the overall risk of maltreatment (Horton &

Cruise, 2001).

One of the seminal studies in the field conducted by Gelles (1992) examined the

relationship between poverty and violence towards children. The data utilized in the study were

collected in two national surveys. The first survey, The First National Family Violence Survey,

was conducted in 1976 while the second survey, The Second National Family Violence Survey,

was conducted in 1985. The concept of "violence" was operationalized using the Conflict

Tactics Scales (CTS). This measure contains 18 items that ask the respondent how often they

use each technique when angry at a family member. The First National Family Violence Survey

sampled 2,143 households in which there was at least one child at home between the ages of 3

and 17 years. The 1985 National Family Violence Survey utilized a sample of 3,233 households

in which there was at least one child under the age of 18 years. The results for the 1976 survey

found that the rate of severe violence was 62% higher in poverty-income families compared to

non-poverty income families. Moreover, the rate of severe violence was 250% higher in

poverty-income families. The 1985 survey produced similar results. Severe violence was found

to be 46% higher in poverty-income families and very severe violence was 100% higher in










poverty-income families. This study was one of the first in the field to provide support for the

claim that violence is more likely to occur in low income families.

In a similar study, Drake and Pandey (1996) examined the relationship between

neighborhood poverty and rates of reported and substantiated cases by Child Protective Services.

The data utilized in this study were from the 1992 Child Abuse and Neglect database in Missouri

along with the 1990 census data to assist in categorizing neighborhood poverty levels. Families

were divided into three categories based on the location of their homes: least poverty subgroup,

median poverty subgroup, and highest poverty subgroup. The study examined 481,722 total

families, across all three groups and found that for all types of abuse, the least poverty subgroup

had the lowest number of reported and substantiated cases of child maltreatment. The reporting

rates for physical abuse grew significantly across the three poverty levels starting from the

lowest to highest. The substantiation rates for physical abuse ranged from 8% to 23% across the

three poverty levels moving from lowest to highest. Similar results were found for child neglect.

Both reporting and substantiation rates increased along with the poverty level. Consistent with

the results reported in Gelles (1992), this study provides more support for poverty as a risk factor

for all types of child maltreatment.

Understanding how the social context in which a family lives may affect child rearing

practices. Many researches have investigated how residing in certain neighborhoods may affect

the occurrence of child maltreatment. When socioeconomic status of neighborhoods was held

constant, differences in maltreatment rates were still found between the neighborhoods (Garbino

& Sherman, 1980), suggesting the importance of factors beyond basic resource levels. Upon

further investigation, formal and informal social supports were related to neighborhoods with a

lower rate of maltreatment.









Macrosystem

The Macrosystem is the larger cultural framework that encompasses the individual, family,

and community (Belsky, 1980). For example, physical punishment of children has been shown

to have differential support across various cultural groups. Researchers have consistently shown

that "harsh" physical punishment was more prevalent among African-American families than in

European families (Baumrind, 1995; Deater-Deckard, Dodge, Bates, & Pettit, 1996).

Differences have also been reported between Hispanic, African-American, and Caucasian

parents' use of parental nurturing behaviors with their children, with African-American parents

displaying more nurturing behaviors than their Hispanic counterparts (Ferrari, 2002).

Racism is another component described as part of the macrosystem (Scannapieco &

Connell-Carrick, 2005). As institutional racism continues to block opportunities for increased

educational and employment prospects for some, the frustration and stress among these cultural

groups may become heightened creating an increased risk of child maltreatment. Recently, more

attention has been given to the overrepresentation of certain minorities in both reporting statistics

and substantiation of child maltreatment cases, lending support for the existence of institutional

racism (Fluke, Hedderson, & Curtis, 2003).

In subscribing to the ecological model of abuse, understanding the role of culture is critical

to understanding each level of the model. The characteristics of the parent, child, and family all

are contingent upon the culture of the family, neighborhood, and society as a whole. Despite the

model's difficulty to be tested empirically, the ecological model of maltreatment has been noted

as the best theoretical model for explaining child maltreatment (Garbarino, 1977). Instead of

trying to focus on a single process or explanation, it provides a model for the comprehensive

interaction of all the factors involved in increasing the risk of child maltreatment.









Culture, Race, and Ethnicity

Throughout the literature, definitions of the terms race, culture, and ethnicity are discussed

extensively, with a number of authors highlighting the lack of clarity in the use of these terms.

The American Psychological Association (2002) defined culture as "the belief systems and value

orientations that influence customs, norms, practices, and social institutions" (p.2). Furthermore,

all individuals are cultural beings and have a "cultural, ethnic, and racial heritage" (p.3). The

APA argues that the term "race" is a socially constructed category by which society judges and

stereotypes individuals. On the other hand, ethnicity is defined as "the acceptance of the group

mores and practices of one' s culture of origin" (p.3). These terms have no agreed upon

definitions among researchers and thus are used interchangeably throughout the literature base.

Therefore, in the present literature review the terms race, ethnicity, and culture may be seen as

interchangeable, as the terms were used interchangeably during the search of the literature.

Moreover, the term "cultural competence" will refer to an awareness of the individual's race,

ethnicity, and/or cultural and how those belief systems may influence their customs, norms,

practices, and social institutions.

Culture and Child Maltreatment

As our society becomes increasingly culturally diverse, the growing need for cultural

competence and cultural sensitivity has come to the forefront as a national concern. In no area is

the need more evident than child abuse and child protection. Culture provides the context for all

interactions and events affecting children within their family structure. It also provides the basis

for defining and responding to child maltreatment. As current demographic trends have

predicted, the changing ethnic composition of this country is a growing reality that further calls

for a high standard of cultural competence (Korbin, 2002).









Cultural competence was first introduced to child protection by anthropologist James

Green (1982) as the term "ethnic competence," which referred to the growing need for human

service workers to be trained in cultural sensitivity (Dubowitz, 1999). According to Green,

cultural competence is a repertoire of skills and knowledge that enables one not only to acquire

sensitivity to cultural diversity but also to transcend cultural boundaries. In other words,

culturally competent practices prioritize children's well-being and protection while

understanding the cultural context of the child (Dubowitz, 1999). Thus the culturally competent

Child Protection Investigator would have the ability to differentiate between cultural practices

and cultural practices which are harmful to the child.

Within the field of child protection, increased attention has also surrounded the use of

cultural evidence to defend child rearing practices that differ from the mainstream definitions of

child maltreatment. Cultural evidence is defined as information offered by immigrant or

nonmainstream parents to influence the legal ramifications following allegations of child abuse

and neglect (Levesque, 2000). This creates another layer of complication for professionals

working in the field. Child protection workers must be informed and educated on how the laws

differ from state to state and what may be permissible as cultural evidence. Most states have

identified exceptions for cultural practices in such areas as religious beliefs in seeking medical

attention, corporal punishment, and poverty (Terao, Borrego, & Urquiza, 1999). Thus, assuming

variability in clinicians' understanding of cultural issues, vast differences may exist in the

decision-making process surrounding identifying and reporting child maltreatment.

Focus in the Literature

With such an increased focus on cultural competence and sensitivity, Nybell and Gray

(2004) conducted a study investigating the process of culturally competent organizational change

among social service workers, supervisors, and managers from three nonprofit child and family










agencies. Interviews and focus groups were performed with each agency to gain an

understanding of how employees viewed cultural competence and its development within their

agency. The results were also compared by race, agency position, and program affiliation of the

interviewee. The case studies highlighted disparate definitions of cultural competence between

individuals as well as an overwhelming need for more training on cultural differences. More

specifically, staff discussed the need for increased diversity in the hiring process within the

agency, along with the cultural clash between predominantly white workers and minority clients.

Workers indicated the desire for more open dialogue in the workplace pertaining to racial and

cultural differences. Also mentioned was the need for minority speakers to assist in training the

staff towards becoming more culturally competent. While this study is limited in that the

findings cannot be generalized to other agencies, it suggests the need for additional training

opportunities within and between agencies.

Researchers have also examined the extent to which child maltreatment literature has

attended to the issue of ethnicity in the design and sampling of research studies. In their 2001

study, Behl, Crouch, May, Valente, and Conyngham provided a content analysis of 1,133 articles

published between 1977 and 1998 in the child maltreatment literature. The authors sought to

identify which studies provided a focus on ethnicity in their sampling, participants, and design.

The articles were taken from three specialty journals in child maltreatment, Child Abuse &

Neglect, Child2\a~ltreatment, and Journal of Child SexualAbuse, and randomly assigned to three

coders. Coders independently reviewed each article for the presence of a focus on ethnicity,

analysis by ethnicity, or ethnicity held constant. Reliability was calculated across 20% of the

articles resulting in interrater agreement ranging from .77 to .95.









The researchers found that of the 1,133 articles analyzed, 6.7% included a focus on

ethnicity. The percentage of articles that reported data on the ethnic composition of the

participants increased significantly across the years examined. Surprisingly, only 50% of the

most recent articles examined (1995-1998) included any information on ethnicity of the subj ects.

In addition, approximately 12% of the articles used ethnicity as a control variable in the analyses.

Those studies generally held ethnicity constant in order to investigate the relationship among the

variables of interest. Surprisingly, of the most recent articles (1995-1998), approximately 75%

did not include ethnicity as a variable in the design or analyses. Implications of this study would

be the increased need for researchers to include ethnicity at a minimum as a descriptive variable

if not a variable of interest. Limitations of this study include the focus on only three of the most

prominent j ournals in the area of child maltreatment to the exclusion of literature in other fields

that may be of relevance. Also, the data examined provide trends in the research available only

to the year of 1998, excluding the last six years of research.

Effects of Culture on Treatment

Researchers have also examined the importance of culture in treating abused and neglected

children. Cohen, Deblinger, Mannarino, and Arellano (2001) reviewed the extent to which

cultural issues have been addressed in the empirical treatment outcome literature for child

maltreatment, the measurement limitations, and possible explanations for the results found.

Some evidence in the literature suggests that ethnicity may play a role in the severity of

symptoms exhibited by child victims of maltreatment. Several researchers report that minority

children suffer more lasting detrimental effects following child maltreatment as compared with

Caucasian children (Cohen et al.). For example, Morrow and Sorrell (1989) found that in

children having suffered child sexual abuse, Hispanic and African-American children displayed

significantly more behavioral problems and depressive symptoms than their Caucasian









counterparts. Rao, DiClemente, and Ponton (1992) found that Asian children who had suffered

sexual abuse were significantly less likely to display anger and sexualized behavior but more

likely to display suicidal ideation versus African-American, Hispanic, and Caucasian children.

However, the lack of studies that control for treatment received among the children is a

confounding variable requiring much more research before interpreting these results too broadly.

Cohen et al. (2001) also reviewed the literature surrounding minority use of treatment

services. Several researchers cite the growing evidence that Asian American and Hispanic

children are less likely to receive mental health treatment regardless of need and socioeconomic

status (Bui & Tekeuchi, 1992; McCabe et al., 1999). Surprisingly, African-American children

covered by Medicaid were shown to receive a disproportionately large amount of mental health

treatment, while those that were not covered by any insurance received the least amount of

services (McCabe et al.).

In understanding the relationship between culture and child maltreatment, several

methodological designs have been used. Methods have included ethnographic approaches,

vignettes, focus groups, clinical studies, and population surveys (Korbin, 2002). Most often,

ethnographic portraits are used in studying culture to allow for a comprehensive view of a

specific culture. This type of methodology often provides information that cannot be gained

through a statistical analysis. Conversely, these types of qualitative methods, rich with

descriptive information, often lack the generalizability that a sound empirical study would

produce.

Many of the ethnographic studies involve much smaller sample sizes and labor intensive

analyses that make replication virtually impossible. Other types of studies such as survey work

or aggregate analyses provide much larger sample sizes, more powerful results, and more










generalizable findings. However, these types of studies also lack the information on the cultural

context and processes that the individual families endure (Korbin, 2002). Thus, Korbin calls for

researchers to provide a more comprehensive design when initiating research that allows for both

qualitative and quantitative methodology within each study. In addition, the importance for

researchers to include ethnicity as a "must have" variable in all research on child maltreatment

from this point forward is discussed (Korbin).

Effects of Culture on Child Maltreatment

Understanding the importance of culture in the etiological model of abuse, researchers

have investigated the effect culture has on abuse and its reporting. Culture has been shown to

affect the way in which lay people define child maltreatment, the reporting characteristics of

both lay people and professionals deemed as mandatory reporters of child maltreatment, and the

reporting and substantiation rates of actual cases of child maltreatment. As mentioned

previously, ethnicity of victims of maltreatment as a variable of interest examined in the

literature is relatively new with only 6.7% of the articles published before 1998 including it as a

focus of research (Behl, Crouch, May, Valente, & Conyngham, 2001). The few studies that have

examined culture and ethnicity since then have yielded differing results, small sample sizes, and

studies that are difficult or impossible to replicate. Thus, the current research leaves many

unanswered questions concerning the effect of culture on all aspects of child maltreatment. In

the following literature review, the areas of defining child maltreatment, parenting practices, and

reporting child maltreatment are discussed.

Defining Child Maltreatment

Previous research focusing on the etiology of child maltreatment has suggested that

cultural differences in child rearing beliefs may influence specific definitions of child abuse and

neglect. Researchers using vignette studies have indicated that minority groups may provide









different definitions of child maltreatment, yet do not regard scripted scenarios with any less

significance than the mainstream culture (Korbin, Coulton, Lindstrom-Ufuti, and Spilsbury,

2000). Korbin et al. investigated the link between definitions of child maltreatment among

parents residing in different neighborhoods with differing levels of risk for child maltreatment.

The sample included 400 parents taken from 20 different neighborhoods in Cleveland, Ohio,

representative of all levels of the US census. Households were eligible for participation in the

study if there was at least one child under the age of 18 living in the home and one parent or

guardian. The sample was mostly female (81.8%) with the following racial diversity: African-

American (54.8%), Caucasian (36.3%), Hispanic (6.8%), Asian and Native American (2. 1%).

Individual variables such as age, race, income, education, and marital status were collected as

well. In an attempt to investigate the presence of neighborhood definitions of child

maltreatment, the respondents were first asked to identify three things they considered to be child

abuse and neglect. Participants were then asked to rate 13 factors drawn from the current

literature on the etiology of child maltreatment on a scale from 1 (contributes nothing) to 10

(contributes a lot) as to how much they thought the factor may contribute to child maltreatment.

Two measures were utilized for analyses in this study, a census-based measure of the level of

impoverishment in the neighborhood and a respondent based measure of perceptions of the

neighborhood conditions.

While the researchers found an overall agreement on the definitions of child maltreatment,

several significant differences were also discovered. African-Americans were more likely to list

behaviors of neglect while Caucasians were more likely to list behaviors of physical abuse

during the free response exercise. Women were also more likely than men to include lack of

supervision as an example of child maltreatment. Neighborhood perceptions of factors









contributing to the etiology of child abuse and neglect also yielded significant findings.

Respondents living in poorer, more chaotic, high crime neighborhoods were more likely to

attribute the etiology to poverty and family structure. Conversely, respondents living in less

poor, less chaotic neighborhoods were more likely to attribute the etiology to a lack of family

values. Limitations of this study include the overwhelming representation of women in the

sample as well as selection bias based on recruiting of subj ects.

Another related study completed by Fontes, Cruz, and Tabachnick (2001) explored the

qualitative differences between African-American and Latino views of child sexual abuse. The

authors utilized focus group interviews to address any differences in definitions of child sexual

abuse, signs that it is occurring, whether it occurs in their communities, and if men and women

hold different views. The participants included 34 men and 24 women divided into eight

separate focus groups defined by older and younger African-American women, older and

younger African-American men, Latino English-speaking men, Latina English-speaking women,

Latino Spanish-speaking men, and Latina Spanish-speaking women. The only demographic data

obtained was age, ethnicity, and that generally participants were from working class or poor and

have low levels of academic education. All the information available on recruiting describes that

all participants who volunteered and met the criteria of being over 18 and of African-American

or Latino descent were accepted. The Latino group was comprised of mostly Puerto Ricans with

a few people from the Caribbean and Central America. The focus groups lasted between 45 to

90 minutes and were videotaped for later analysis.

The results showed several differences between the responses of the focus groups. Both

Latino men and women spoke more about sexual abuse as an issue within their own family or

community while in African-American groups almost no personal stories were shared. In the









rare occurrence that they were shared, the story was ignored by the groups. Male participants

tended to be more detailed in descriptions of child sexual abuse while females were hesitant to

name specific acts. Only Latinos, both men and women, described changes in the family,

culture, and community as factors contributing to child sexual abuse. In English-speaking Latina

groups, participants described social and economic pressure to keep a man around the house as a

risk factor for child sexual abuse. African-American groups, both men and women, referred to

drug use and abuse as a risk factor for child sexual abuse (Fontes et al., 2001).

While the results obtained provide an in depth understanding of the participants' beliefs,

generalizability of the results is limited. First, the participants were not randomly selected and

the sample size was relatively small. Additionally, only limited demographic data was collected.

The duration of the focus group was short, allowing limited time for the participants to feel

comfortable sharing such intimate information. Finally, the focus groups were all led by men,

which the authors state could have affected some of the participants' willingness to share

personal information.

In summary, results of studies examining how culture affects definitions of child

maltreatment suggest that there may be differences in the type of abuse certain cultures are

willing to discuss or admit occur within their communities. Findings have also suggested that

the neighborhood in which one resides may have an affect on how one conceptualizes the

etiology of child maltreatment. Additionally, gender differences have been reported to affect

both of the aforementioned Eindings, however; these results are extracted from studies with an

overwhelmingly female sample. Thus, completing more studies incorporating a larger sample of

both men and women along with a more diversified group of participants are needed to further

the research in this area.









Parenting Practices

Researchers have also focused on cultural differences and predictors of parenting attitudes

and behaviors, hypothesizing their potential impact on child maltreatment. With such a high

percentage of U. S. parents utilizing physical punishment, researchers have examined the long-

term effects of using such discipline practices. Some researchers maintain that simply using

physical punishment may result in later dysfunction while others suggest the detrimental effects

depend upon the context in which the punishment is used. Wissow (2001) analyzed data from

the Commonwealth Fund Survey of Parents with Young Children, a national sample of 2,017

parents with at least one child younger than three living in the home. The sample consisted of

1,320 mothers/female guardians and 687 fathers/male guardians. The parents were interviewed

to assess their use of six types of disciplinary practices, 3 negative (yelling, spanking, and

hitting) and 3 positive (time out, taking away toy, and explaining). Only the parents of 18-36-

month-olds were asked about neutral/positive practices. The parents were further questioned

about how often in the last week they engaged in specific nurturing activities (e.g., reading to

child). Finally, parents were given a self-report survey that measured depressive symptoms.

The study found several significant results. First, 40% of parents report spanking

sometimes or often, while another 40% report spanking at least once. The most common form of

neutral discipline was time out, with 62% of parents with a child between the ages of 18-23

months reporting use, while 73% of parents of 24-36 month olds reporting use. Boys were

significantly more likely to be yelled at or spanked than girls. Based on the amount of reported

spanking, the spankers were grouped into average, above average, and highest levels of

spanking. The average spankers were found to do so in context with a larger use of nurturing

interactions as well as both negative and neutral disciplinary interactions. The above average

spankers report less reading, listening to music, playing, and hugging their children and use less









of other disciplinary strategies. The highest levels of spanking were found at the two extreme

ends of the participant groups. The first group was comprised of the greatest proportion of low-

income parents and largest number of Hispanic and African-American parents. The second

group with the highest levels of spanking was comprised of the smallest proportion of low-

income parents and was nearly 80% Caucasian.

These results indicate that negative disciplinary practices are not just defined by ethnicity

or income level but a more complex constellation of factors. Use of spanking was found to be

strongly associated with parental reports of depressive symptoms. Consequently, groups of

parents with higher rates of physical punishment seem to be more defined by parental mood than

by race, ethnicity or socioeconomic status. A limitation of the study was the lack of information

on how the term "spanking" was defined and explained to the participants. Also limiting the

results found in the study is the lack of supplemental information on the family environment and

context within which the discipline occurs. Finally, only one parent was randomly chosen from

two parent homes to participate without question as to the consistency of disciplinary techniques

across caregivers.

Ferrari (2002) examined the relationship between culture and parenting behaviors and

attitudes. This study also examined the impact a childhood history of abuse would have on the

parenting beliefs. This study hypothesized that cultural variables would be more predictive of

parenting practices than ethnicity as a demographic variable. The study included 75 fathers and

75 mothers who were recruited as nontraditional students attending night classes at a community

college. The sample was comprised of African-Americans (33%), Hispanics (33%), and

Caucasians (33%). The participants were given a booklet containing seven scales or checklists

in a randomized order: The Machismo Scale, Familism Scale, Valuing Children Scale, Conflict









Tactics Scale, Nurturance Scale, vignettes of child maltreatment, and the Childhood Trauma

Questionnaire.

The study found that regardless of ethnic status, fathers who held familism in low regard

were more likely to use physical punishment to discipline their children than fathers who valued

familism more. Machismo predicted more use of physical punishment from fathers, but not for

mothers when controlling for ethnicity. For all parents, valuing children more was associated

with lower tolerance for mistreating behaviors. A history of childhood abuse or neglect was

predictive of the mother' s use of physical and verbal punishment. Interestingly, the fathers who

reported a history of child abuse were less likely to use physical punishment. The data suggests

that the intergenerational transmission of abuse may hold true for mother but not for fathers, at

least within this population. No ethnic differences were found in the seriousness ratings of child

abuse and neglect. African-American parents were found to report participating in more

nurturing behaviors than Hispanics, while also using more physical discipline than both Hispanic

and Caucasian participants. African-American mothers were found to use significantly more

verbal and physical abuse than African-American fathers. The findings indicate that both

ethnicity and gender are complex factors which are comprised of social roles, gender roles, and

norms, all of which may impact parenting styles and definitions of child maltreatment. This

study is limited by the use of only college students with no measure of socioeconomic status or

acculturation.

Investigating similar constructs, Coohey (2001) concentrated on different dimensions of

familism and the relationship to child maltreatment in Latino and Anglo families. Familism is

the most important culture-specific set of attitudes and beliefs held by Mexican, Puerto-Rican,

Cuban, Central and South Americans. These culture specific attitudes and beliefs emphasize the










importance of maintaining a supportive extended family network. Coohey hypothesizes that less

familistic attitudes would thus be linked to child maltreatment among Latinos. The study sought

to investigate if varying degrees of familism existed between and within Latino and Anglo

families and if these differences were in fact related to child maltreatment. The participants

included Anglo and Latina mothers taken from Child Protection Services records and matched

with public school parents from Chicago on mothers' ethnicity, country of origin and income

level. The sample included 51 nonabusive Anglos, 51 abusive Anglos, 35 nonabusive Latinas,

and 35 abusive Latinas. The study utilized a network circumscription approach in capturing the

maj ority of mothers' informal supports. The mothers were asked to name formal and informal

supports and evaluate each person on a variety of characteristics measuring familism. The three

dimensions of familism examined were attitudinal familism, behavioral familism, and structural

familism. Attitudinal familism is the idea that relatives possess more positive characteristics

than nonrelatives and a preference for exchanging valued resources with relatives over

nonrelatives. Behavioral familism is the actual receipt of support from relatives as opposed to

attitudes or beliefs about what others should give or get. Moreover, it focuses on support Latinos

receive from extended kin and not everyone in their informal support network. Finally,

structural familism is the actual size and proximity of kinship networks.

Several differences were found across all dimensions of familism. Angles, both abusive

and non-abusive, were found to be more likely to report at least one of their kin members were

critical of them as well as having at least one unreliable relative. A significant interaction was

found for the abusive Anglo mothers who reported they had at least one kin member who

couldn't control his or her anger. Nonabusive mothers, regardless of ethnicity, report more kin

who were warm and caring. Main effects for both abuse status and ethnicity were found on the










level of interest in getting emotional support from friends. Abusive mothers and non-abusive

Anglo mothers were more interested in obtaining emotional support from friends.

Significant results were also found across the dimensions of behavioral and structural

familism. A significant main effect for abuse status on the number of kin members was found.

Nonabusive mothers had significantly more kin in their social networks than abusive mothers.

Abusive mothers had fewer contacts with their kin than nonabusive mothers. Abusive mothers

also have fewer friends than nonabusive mothers. While the author's hypotheses were partially

supported by this study, several implications for culturally relevant practices were found. Based

on the results, practitioners should inquire as to the supportive or unsupportive nature of the

mother' s kin when either ascertaining level of support or investigating possible out of home

placements for children. Also, consistent with previous research, a pattern of disconnectedness

from kin could put poor Anglo mothers at risk for maltreatment. Practitioners should investigate

both kin and friends' level of support, proximity, and availability for Anglo mothers. A pattern

of receiving more emotional support from friends versus kin may characterize a family cultural

pattern that negatively influences parenting for Latinas in this study. Some limitations of this

study include the participation of only women in the sample. Also, during the interview process

questions were translated for non English-speaking participants, possibly losing some meaning

during the questionnaire process.

Reporting Child Maltreatment

Identification and willingness to report child maltreatment have gained a large focus in the

child maltreatment literature. With such a diverse number of professionals designated as

mandatory reporters of child maltreatment under state statutes, many different groups have

subsequently been studied in the literature. Without the ability to control and deviate case

variables such as the child's race, SES, or age within real life child maltreatment cases,









researchers have overwhelming adopted the use of simulated case vignettes to access the effect

of reporter characteristics and the characteristics of the scenario on the reported severity of abuse

and willingness to report.

The process of identifying and reporting child maltreatment is a comprehensive procedure

involving three distinct phases (Ashton, 2004). The reporter will first gather data about a given

situation, assess the behavior, and respond in a specific manner (Ashton). Personal

characteristics of the reporter, such as beliefs, attitudes, and opinions developed through

experiences may play a role in each step of the above process. Ashton examined the relationship

between personal characteristics of 276 entry-level workers and their decision to report child

maltreatment. The participants in this study were a group of culturally diverse college seniors

from an urban area all majoring in the social service field. The study provided the participants

with a self-administered questionnaire comprised of case vignettes depicting different child

maltreatment scenarios. Eight independent variables were assessed for each participant: age,

gender, parenthood, mother and father' s education, college maj or, ethnicity, immigration status,

and whether or not the respondent was born in the United States. The sample consisted of 30%

African Caribbean, 24% African-American, 25% Hispanic, 9% White, 6% Asian, and 6% other.

The participants were presented with 12 vignettes and asked to decide between three outcomes:

nothing needed to be done, the family needs to seek mental heath assistance, or the family needs

to be reported to child protection. The respondents also needed to rate each situation on a Likert

scale from 1 to 7 for the seriousness of the situation.

Ashton's study found that out of all eight characteristics, only ethnicity and immigration

status were related to the likelihood of the participants to report the abuse. Caucasians were

found to be significantly more likely to report abuse, while Asians were the least likely to report









the incident to child protection. The only other significant result found was that respondents who

were born in the United States were significantly more likely to report versus immigrants who

were born elsewhere. This study was limited in that it was a self-report measure which only

required college seniors to hypothesize if they would report, not if they actually did in a real life

situation. Furthermore, college seniors are not actual mandated reporters of abuse thus limiting

the generalizability of the results reported in this study.

In a similar study completed by Maynard and Wiederman (1997), a group of 413

undergraduate students recruited from an introductory psychology course were presented with

one of eight vignettes depicted child sexual abuse and asked to rate the severity of the scenario,

and respond to questions examining their perceptions of the situation and any attributions of

responsibility or blame. The sample consisted of 208 men (50.4%) and 205 women (49.6%).

Furthermore, 87.2% of the respondents were White, 6.5% were African-American, 2.4% were

Asian, and 1.9% were Hispanic. The vignettes depicting a 15-year-old child were rated as less

severe than those situations depicting a 7-year-old victim. Additionally, ratings of severity of

abuse and attributions of blame were significantly lower when the child and the victim were the

same sex. Gender-role attitudes were not significantly related to ratings of abusiveness or

attributions of blame. Limiting this study was the inclusion of undergraduate students instead of

mandated reporters of abuse. Moreover, the sample was overwhelmingly Caucasian further

highlighting the need for more culturally diverse samples in the child maltreatment literature.

In addition to research focused on undergraduate responders, graduate students in the area

of social work were sampled in a study completed by Ashton (1999) using 12 vignettes depicting

problematic parental behavior. The participants were then asked to rate the seriousness of the

situation and whether or not they would report the incident. The sample consisted of 86 first-










year graduate students majoring in social work. The sample was overwhelmingly female

(83.6%) and predominately Caucasian. Vignettes depicting physical violence, imminent harm,

or young children were rated more serious as well as being more likely to be reported. This

study did not find any relationship between respondent data, such as age, gender, or ethnicity,

and the ratings of severity and willingness to report. The author provides the reasoning that the

sample was overwhelmingly homogeneous, thus again calling for a more culturally diverse

sample. While sampling first year graduate students in a social work program increases the

generalizability versus undergraduate samples, it is still not representative of actual mandated

reporters.

Additionally, previous research has shown that graduate programs in clinical, counseling,

and school psychology did not meet minimum standards imposed by the American Psychological

Association (APA) in the area of child maltreatment training (Champion, Shipman, Bonner,

Hensley, & Howe, 2003). A survey assessed coursework, practice, and research experience in

the area of child maltreatment and found that all programs fall short of minimal levels of

competence, with no increased attention or focus over the last ten years. For most of the

programs, child maltreatment was briefly covered as a subj ect in law and ethics seminars offered

during a semester long course. Thus, research sampling college students may not provide as

much relevant information for understanding the reporting characteristics of professional

mandatory reporters.

While detecting and reporting child maltreatment is not a mandatory requirement of

students, it is a legal responsibility placed on school personnel. Thus, research has also focused

on the reporting characteristics of educators and other school personnel. While this

responsibility is of crucial importance, studies show that teachers feel unprepared, unsupported,









and ill-informed when it comes to reacting in situations that call for clinical and professional

judgment.

Kenny (2001) surveyed 197 teachers on their knowledge of child abuse laws, procedures

for reporting, and ability to detect child maltreatment in two vignettes. Of the sample surveyed,

1 1% of the teachers stated that they failed to report incidents of abuse at their school. When

asked what impacted their decision not to report, 39% stated that they feared making an

inaccurate report. Thirty-four percent of both special education and regular education teachers

report their pre-serviced training to properly handle child maltreatment as being inadequate. The

respondents identified a lack of preparation and training in their course work, as well as a lack of

knowledge of specific legal requirements. The teachers responded to the vignettes, which both

described obvious scenarios of child maltreatment, as reporting only 26% for the first case and

11% for the second case. Kenny (2001) also emphasizes the ethnic differences observed among

teacher respondents. None of the African-American or the African Caribbean teachers reported

the need to alert child protection in the scenarios presented.

The results reported in Kenny (2001) are echoed throughout the child maltreatment

literature. In a similar study conducted by Hinson and Fossey (2000), elementary classroom

teachers were administered a survey on their abilities to recognize abuse, perceptions of the

reporting process, and their understanding of liability. The results showed many of the teachers

were unable to recognize the signs of abuse and were unaware of the reporting procedures at

their schools. Further, most of the respondents were unaware of immunity rights in the law,

indicating that they felt fear and anxiety over possible retribution from the abuser. In addition,

educators reported considering child protective services ineffective and unable to help the

children (Hinson & Fossey, 2000).









In a study completed by Webster, O'Toole, O'Toole, and Lucal (2005), a sample of 480

teachers responded to vignettes in which the case characteristics, including socioeconomic status

and ethnicity, were systematically manipulated. The teachers responded to whether they

believed the situation to be child abuse and if they felt they would report the incident. The

sample was overwhelmingly Caucasian (93.1%) as well as largely female (74.2%). The results

showed that in the sample of teachers, overreporting was rare, occurring in only 4.2% of the

cases, with underreporting occurring in one third of the cases. A difference was found between

teachers working in rural schools versus suburban schools, with those in rural schools showing

an increased likelihood of underreporting the abusive situations. This study provided a relatively

untapped characteristic of child maltreatment, the difference between professionals employed in

rural and suburban settings and the effect that may have on reporting rates. Again, a more

culturally diverse sample needs to be examined as well as one incorporating an equal amount of

men as women.

Along with professional educators, other professionals have also been studied throughout

the literature. In a study completed by Warner-Rogers, Hansen, and Spieth (1996) a sample of

sixty 3rd and 4th year medical students were provided a set of sixteen vignettes depicting

scenarios of child maltreatment that systematically manipulated the injury severity, injury

explanation, and delay in seeking medical attention while holding SES and ethnicity constant.

The sample consisted of 40 males (66.7%) and 20 females (34.6%) with no information on the

respondents' ethnicities. The study found a significant three-way interaction between injury

severity, explanation, and delay in seeking treatment. Reporter gender or amount of clinical

experience was not shown to affect identification of abuse. Limiting the findings of this study is

the use of medical students. While the students in this study were in their 3rd or 4th year of









medical school, the ability to generalize to licensed physicians and other mandated reporters is

limited.

Licensed psychologists and certified Masters social workers were surveyed by Hansen,

Bomby, Lundquist, Chandler, Le, and Futa (1997) in an effort to identify case and professional

variable on the identification and reporting of child maltreatment. The sample consisted of 125

licensed psychologists, 69 (55.2%) female and 56 (44.7%) male. The sample was also

predominantly Caucasian (95.2%). Also surveyed were 85 Certified Master Social Workers

(CMSWs), with 45 being female (52.9%), 40 being male (47.1%). The maj ority of the CMSWs

were also Caucasian (91.8%). The respondents were presented with five case vignettes that

systematically manipulated three variables, the child's age, race, and SES. Consistent with

previous research, the vignettes depicting a younger child received a higher severity of abuse

rating. An effect of socioeconomic status was found, but only related to the sexual abuse

vignettes. The scenarios depicting families of lower socioeconomic status were associated with

higher ratings of abuse. The factor that produced the strongest effect on ratings of abuse was the

ethnicity of the parent and child described. The vignettes depicting African-American families

were rated as being less severe and less likely to be reported than the same identical scenario

depicting a Caucasian family. No significant differences for gender were found in the study.

Replicating this study with a more culturally diverse sample would increase the understanding of

the complex issue of race when investigating child maltreatment.

Disproportionality in Child Maltreatment

The occurrence of disproportionate representation of race and ethnicity in actual reporting

statistics has been examined in the child maltreatment literature. Fluke, Hedderson, and Curtis

(2003) investigated the racial and ethnic representation of child maltreatment reports using

National Child Abuse and Neglect Data System (NCANDS) data for more than 700,000 children









across five states. The data were collected in year 2000, from one state in the mid-Atlantic

region, one in the Midwest, two from the South, and one from the Southwest. The states were

specifically chosen based on their overall population and diverse racial populations. For each

state, a disproportionality representation index (DRI) was constructed for both investigation and

victim data. The Investigation DRI is defined as the ratio calculated by dividing the percentage

of children in a specific racial group who were subj ect of an investigation by the percentage of

children of the same racial group in the population in the state. The Victim DRI uses the same

equation to calculate the number of children of a specific racial group that were found to be

victims of maltreatment by the percentage who were subj ect of an investigation. The results

showed that across all five states, the Investigation DRI for African-American children who were

subj ect of an investigation is greater than 1. In addition, across all five states, the Victim DRI for

African-American children was greater than 1. The Victim DRI was substantially less than the

Investigation DRI for African-American children suggesting a bias in reporting procedures.

Great variation among both the Investigation and Victim DRI for all other ethnic groups was

found (Asian/Pacific Islander, White, Hispanic, and Native American), with no consistent results

greater than one.

In a similar study, researchers investigated the records of 3,936 children receiving care in

out-of-home placement for suspected child maltreatment (Lu, Landsverk, Ellis-Macleod,

Newton, Ganger, & Johnson, 2004). The study examined the relationship between child

characteristics such as age, gender, race, and reason for referral with case outcome decisions.

The analysis revealed significant differences among case outcomes for children of different

racial backgrounds. African-American children, when compared to relevant census data, were

over-represented in the child protective system as well as out-of-home care. Furthermore,










African-American children were significantly less likely to reach the goal of reunification with

their family than all other ethnicities examined. The study also investigated the initial child

abuse reporting rate, uncovering another significant effect of race. African-American children

were found to be over-reported while Asian and Hispanic children were found to be under-

represented.

The above study is consistent with results found by Barth (1997). Barth utilized a database

that included all children who were placed in out-of-home care in California since the year 1988.

The sample studied was a cohort of 3,873 children who were selected because it was the first

time they entered out-of-home care, were less than six years old at placement, were African-

American, Caucasian, or Latino, and were placed in a nonkinship foster family. The analysis

demonstrated that African-American children were less likely than Caucasian children to be

reunited with their families (41% v. 58%). Furthermore, African-American children are more

likely to remain in long-term care while Caucasian children are more likely to be adopted than

stay in long-term care. Latino children were found to be equally as likely to be placed in either

setting. Barth expounds as to possible contributing factors in this overwhelming

disproportionality in child welfare outcomes for African-American children. He states that

hiring "additional ethnically competent social workers" (p.6) may increase adoption rates for

African-American children.

McCabe et al. (1999) investigated representation of different ethnic/racial groups across

Hyve public youth service sectors. The sample analyzed in the study was drawn from the Patterns

of Youth Mental Health Care in Public Services Systems Proj ect (POC) which includes children

(birth to 18 years) who were active in one or more of Hyve areas of care (drug treatment, child

welfare, juvenile justice, mental health, and SED services in public schools). The sample









consisted of a cohort of 11,515 children who resided in San Diego during the 1996-1997 year.

African-Americans were found to be overrepresented in the child welfare system, mental health

services, juvenile justice, and SED. The only sector that they were not shown to be

overrepresented was in drug treatment. Caucasian children were found to be underrepresented in

child welfare, mental health, and juvenile justice. Furthermore, Caucasian children were

represented at expected rates in drug treatment and were overrepresented in SED within the

public schools. Latinos were underrepresented across all Hyve public sectors. With African-

American children being present in the child welfare system at 264% of the expected rate, the

authors suggest that "overidentifieation" may be to blame.

Investigating the same problem, with a slightly different focus, Garland and Besinger

(1997) examined court referrals for mental health services between children of different

ethnicities. The sample was taken from a longitudinal study, the Foster Care Mental Health

(FCMH) research project, and included 184 randomly selected cases. The authors analyzed the

children's child welfare case Hiles and found that Caucasian children were most likely to have

received counseling before entering the system. Moreover, they were most likely to receive

counseling within the first eight months of out-of-home care as compared to the other ethnicities.

Upon further analysis, African-American and Hispanic youth were less likely to use therapy

services when controlling for age, type of maltreatment, and previous use. Garland and Besinger

discuss possible causes of this disparity in the system. They hypothesize that in addition to

family cultural differences, "service system factors" (p.663) may play a role in these findings.

More specifically, they state that "service systems may not make appropriate services available

and accessible to minority families" (p.663).









The aforementioned studies indicate that ethnicity plays a significant role in the decision

making process at all levels throughout the child welfare system. It suggests the need to

reevaluate the current model of child protective services in an effort to understand exactly how

and why the needs of minority clients are not being met. More specifically, the workers,

agencies and policies that hold the responsibility for decision making need to be addressed more

directly within the relevant literature. In a recent review of the research available on minority

children in the child welfare system, Hines, Lemon, Wyatt, and Merdinger (2004) conclude that

ethnicity and/or SES is clearly a factor contributing to the disproportionate representation of

minority children in the child welfare system. More specifically, the authors state, "research in

the area of bias in reporting remains scant and inconclusive" (pg. 522).

Models of Disproportionality

While overrepresentation of certain ethnicities throughout the different stages of the child

welfare system has been researched and discussed for years, large scale studies of

disproportionality in the reporting and substantiation of child maltreatment cases are few and far

between (Barth, 2005). Barth proposes several possible explanations for the overrepresentation

of African-American children and families in the child welfare system. In his model entitled the

Child Welfare Services Decisionmaking Model, Barth describes the disproportionality in the

population being primarily attributed to agency characteristics that allow the race of the child to

play a role in every decision making point throughout the child welfare system. More

specifically, factors such as "staffing patterns, culturally incompetent staff, institutionalized

racism, and inadequate duration or configuration of services" (p.36) are provided as possible

contributing characteristics. Barth also cautions that this model is only in the infancy of

development due to the difficulty of teasing out race and ethnicity and being able to control and

manipulate such variables within the research. Furthermore, Ards et al. (1998) reports that









cultural differences in child-rearing beliefs between the reporter and the perpetrator along with

an unwillingness to report perpetrators of one' s own socioeconomic status as a possible reason

for differing reporting rates.

This model provides a framework for understanding bias in the reporting and

substantiation rates for child maltreatment by questioning the cultural competence and training of

child protection workers. However, as shown in the preceding review of the literature, few

researchers have actually surveyed the individuals responsible for substantiating the allegations

of child maltreatment. Regardless of this fact, the researchers continue to extrapolate the results

and hypothesize that bias in the system may be a contributing factor.

In summary, this literature review has described the effects that reporter characteristics

have on the seriousness that reporters rate vignettes depicting child maltreatment as well as the

willingness to report these cases to the Child Protective Investigators. Such characteristics

include gender, ethnicity, profession, and county in which one works (Ferrari, 2002; Fontes,

Cruz, & Tabachnick, 2001; Korbin, Coulton, Lindstrom-Ufuti, & Spilsbury, 2000). In addition

to reporter characteristics, the characteristics of the child have been shown to play a significant

role. The child's age, gender, ethnicity, and socio-economic status have all been shown to have

an effect on reporter responses (Maynard & Wiederman, 1997). Along with this research, a

growing awareness of the disproportionality in the ethnicity of children being reported and

substantiated as victims of child maltreatment has increased the importance of such studies

(Fluke, Hedderson, & Curtis, 2003; Lu, Landsverk, Ellis-Macleod, Newton, Ganger, & Johnson,

2004). Previous researchers have examined the impact of both reporter and victim

characteristics among professionals such as doctors, nurses, medical students, undergraduates

and teachers (Hansen et al., 1997; Warner-Rogers, Hansen, and Spieth, 1996; Webster, O'Toole,









O'Toole, and Lucal, 2005). The results of such research have been used in hypothesizing the

reason for the growing disparity in the child maltreatment statistics. Researchers have suggested

that Child Protective Investigators' lack of cultural sensitivity training along with reporter biases

have been contributing factors (Barth, 2005). In order to isolate ethnicity as the variable of

interest, gender of the child will be held constant in the following study.









CHAPTER 2
METHOD

The purpose of this study is to examine how reporter characteristics and child

characteristics affect reporter response among currently employed Child Protective Investigators.

To this end, the following questions will be addressed:

1. Do Child Protective Investigators rate severity of maltreatment differently based on victim' s
race/ethnicity when gender is controlled?

2. Do Child Protective Investigators rate severity of maltreatment differently based on
characteristics such as race/ethnicity, gender, amount of cultural sensitivity training of the
reporter, and setting of employment (i.e., rural versus suburban)?

3. Is there a difference in willingness to report based on the child's race/ethnicity, reporter
race/ethnicity, reporter gender, amount of cultural sensitivity training of the reporter, or the
populations of the county in which the reporter works?

Participants

A total of 162 participants completed the study. After examining the data for missing

values, twelve participants were removed from analyses due to missing independent variables.

Therefore, the final sample consisted of 150 Child Protective Investigators employed by the

Department of Children and Families (DCF) who are currently working in seventeen different

counties in North Central Florida. The participants were asked to identify themselves as

African-American or Black (B), Caucasian or White (W), Hispanic or Latino (LA), or belonging

to some other cultural group (OT). Table 2-1 illustrates the frequency distribution of gender,

ethnicity, and county among the participants within the three forms. The entire sample was

comprised of 111 females (74%) and 39 males (26%), ranging in age from 22 to 64 years.

Thirty-two percent of the sample was African-American, 66% was Caucasian, 6% was Hispanic,

and 5% identified themselves as Other. Twenty-seven percent of the sample worked in a rural

county, while 72% reported working in an urban/suburban county. The participants reported

time on the j ob ranged from 1 month to 3 8 years of experience.










The participants were recruited through their DCF supervisors. Every Child Protective

Investigator in the sampled 16 counties was asked to participate in the study. The first 150

participants who volunteered were included in the study. Each supervisor was given a summary

of the purpose of the study, type of subj ects needed, confidentiality of information, and

procedures. Additionally, a less detailed summary of the study was also distributed to the

supervisors to be disseminated among the participants. The supervisors were consulted

regarding the least disruptive manner in which to conduct the study. Dates for data collection

were scheduled with the supervisors who were instructed to disperse the summary of the study

one day prior. Supervisors were informed that a copy of the results will be available for all

participating DCF facilities. Informed consent was read aloud and obtained from each

participant prior to the participant' s data being used for the study.

Measures

Demographic questionnaire

The measures used in this included a demographic questionnaire (see Appendix A), along

with eight vignettes adapted from Ashton (1999) (see Appendix B). The demographic

questionnaire was specifically developed for this proj ect and requested participants provide the

following information: gender, ethnicity, age, degree, marital status, religion, childhood history

of abuse, number of children, population of county in which they work, amount of training, and

amount of time with current employer.

Vignettes

Following completion of the demographic data sheet, participants then read and completed

several questions pertaining to eight vignettes. Each vignette depicted a Sictitious description of

problematic parental behaviors. The participants were then asked to imagine that as caseworkers

they were approached by an educator who became aware of certain circumstances. The










participants were then asked to read each vignette and rate the information on a Likert scale of

"seriousness" from 1 to 7, with "1" being not serious and "7" being very serious. After the

respondents rated the seriousness of the scenario, they were then asked to indicate how they

would tell the educator to respond by selecting one of three options: (1) no action, nothing needs

to be done about the situation; (2) encourage the family to seek professional help; or (3) notify

the child protection agency to investigate the situation. The vignettes were developed by Vicki

Ashton (1999) who administered a total of 12 vignettes to first-year graduate students. Ashton

tested the vignettes for face content validity of "seriousness" by presenting them to a group of

social work professionals. They were asked to rate the seriousness of the vignettes on a scale of

0 to 6, according the amount that each vignette depicted a form of child maltreatment. Ashton

also tested the reliability of the vignettes by a 1-week test/retest on a sample of undergraduate

social work maj ors. She reported the median correlation coefficient (Pearson r) for the vignettes

as .61. Ashton cites the restricted range of responses as contributing to such a low estimate.

All four vignettes were used from the low and moderate level of severity to construct the

eight vignettes that will be utilized in this study. In order to isolate and examine ethnicity as the

variable of interest, the gender of the child in the vignettes was controlled for. Some

modification of the wording was necessary within each chosen vignette in order to change the

victim's gender. Furthermore, the ethnicity of the child was manipulated and three forms of the

vignettes will be created. One form will contain all black children, one form will contain all

white children, and one form will provide no information concerning the child's ethnicity. Each

form will be randomly assigned among the participants.

Procedure

Each Department of Children and Families supervisor decided which day was optimal for

data collection at their respective agency. On the day scheduled, the principal investigator









handed out consent forms that were read aloud to the participants. The participants then had the

opportunity to ask any questions they had. Participants were told that if they wished to

participate in the study, they should sign the form to indicate that they understood what I read.

The consent form instructed the participants not to put their names on any of the study materials

in order to protect their privacy. Participants who did not wish to complete the study were free

to leave.

Each participant was given a packet containing a demographic questionnaire along with

eight vignettes depicting different fictitious cases of child maltreatment. Following each vignette

were questions pertaining to the seriousness of the scenario and if the Child Protective

Investigator would advise an educator to report the incident. Three forms of the measure were

created to allow for changing the ethnicity of the children in the scenarios. Each form was

randomly assigned among the participants. Participants were then asked to refrain from

discussing the study with or around other Child Protective Investigators.

Analysis

The goal of this study was to determine if reporter characteristics and/or case

characteristics was related to the reported level of severity and/or the decision to report the

incident to Child Protective Services. A one-way, between-subjects ANOVA was utilized to

investigate whether the reported rate of severity differed based on the child's ethnicity. The

independent variable was the form code (three levels) and the dependent variable was the

reported level of severity per vignette. Multiple regression analyses were utilized in analyzing

the second research question for each of the eight vignettes. The dependent variable was the

reported level of severity, with the independent variables being child ethnicity, reporter

race/ethnicity, reporter gender, population of the county in which the reporter worked, years in

social services, and the amount of cultural sensitivity training. Logistic regression was used to










analyze the third research question. The dependent variable was the decision to report the

vignette to Child Protective Services. The independent variables were again child ethnicity,

reporter race/ethnicity, reporter gender, population of the county in which the reporter works,

years in social services, and the amount of cultural sensitivity training.










Table 2-1. Descriptive statistics for participants by form (n=150)
Form A Form B Form C
Variable n= 52 n= 47 n= 51

Gender
Male 14 11 14
Female 38 36 37
Ethnicity
Black 12 7 13
White 35 34 31
Hispanic 3 1 6
Other 2 5 1
County
Rural 16 12 13
Suburban 36 35 38









CHAPTER 3
RESULTS

The results of this study are presented in four sections. The first section presents the

results of the descriptive statistics for all of the variables investigated in this study. Included in

this section are the results of correlation analyses between the variables investigated in this

study. The following three sections present the results for each of the main hypotheses of the

study. The implications of these findings are discussed in Chapter 4.

Descriptive Statistics

Descriptive statistics illustrating the mean severity rating of each vignette, as well as the

percentage of workers who chose to report each vignette are presented in Table 3-1. Mean

severity ratings, across each vignette, ranged from a low of 4.27 to a high of 6.74. Standard

deviations for the 8 incidents ranged from a low of .60 to a high of 1.65. Internal consistency

reliability was calculated for the 8-item overall severity scale using Cronbach's Coefficient

Alpha and the following alphas were found: a = .79 for all forms combined, a = .84 for form 1,

a = .75 for form 2, and a = .77 for form 3.

Correlations

The correlations between overall severity rating, gender, ethnicity, county, cultural

sensitively training, and years in social services are given in Table 3-2. The overall severity

rating was found to significantly correlate with years in social services (r = -.20, p = .01). The

relationship is negative, indicating that participants with more tenure in social services were

more likely to rate vignettes as less severe.

Research Question 1

Do Child Protective Investigators rate severity of maltreatment differently based on

victim's race/ethnicity when victim's gender is controlled? This research question was










investigated through the use of a one-way, between-subj ects ANOVA. The means between

groups for each vignette can be found in Table 3-1. The 8-item overall severity scale means for

the different forms were compared for Form A/black child (M~= 5.49, SD = 0.82), Form B/white

child (M~= 5.68, SD = 0.63), and Form C/No Race Given (M~= 5.49, SD = 0.79). A significant

difference was not found between groups based on the ethnicity of the child reported in the

vignette, F (2, 147) = 1.02, p = 0.36. Therefore, no evidence was found in support of hypothesis

1. Exploratory analyses were completed on the 8-item vignettes separately and no significant

differences were found. These results can be found in Table 3-3.

Research Question 2

Do Child Protective Investigators rate severity of maltreatment differently based on

characteristics such as reporter race/ethnicity, reporter gender, amount of cultural sensitivity

training of the reporter, and setting of employment (i.e., rural versus suburban)? Multiple

regression analysis was utilized in answering the second research question. These results are

illustrated in Table 3-4. Partial support was found for this hypothesis.

None of the variables investigated were found to be significant predictors of the 8-item

overall severity rating scale. Exploratory analyses were completed on each vignette separately.

Since there were no apriori predictions, we used a more conservative Bonferroni-corrected

significance value corrected for 8 items (p < .05/8 = .006). Years in social services was found to

be a significant predictor of severity ratings on Vignette 2 (b = -0.04, p = 0.00), vignette 3 (b = -

0.04, p = 0.00), and Vignette 6 (b = -0.04, p = 0.00). The relationship between time in social

services and severity rating is negative for each vignette, indicating a general pattern that the

longer the participants have been in their field the less likely they are to rate the vignettes as

severe.









Research Question 3

Is there a difference in willingness to report based on the child's race/ethnicity, reporter

race/ethnicity, reporter gender, amount of cultural sensitivity training of the reporter, or the

populations of the county in which the reporter works? Descriptive statistics for the percentage

of vignettes on which a referral was made can be found in Table 3-1. Logistic regression was

utilized regressing the outcome willingness to report onto six predictors: child ethnicity (3

levels), reporter ethnicity (4 levels), reporter gender (2 levels), amount of cultural sensitivity

training, years in social services, and the population of the county in which the reporter works.

Table 3-5 illustrates the results of the analysis and contains both b weights and odds ratios. This

analysis was exploratory in nature and did not utilize any apriori predications. Thus, the

significance value must be conservative and a Bonferroni correction was used for 8 items (p <

.05/8 = .006).

Unfortunately, only five vignettes were able to be used in this analysis due to lack of

variance in the sample. For Vignettes 1 and 6, a lack of variance on the willingness to report

outcome caused quasi-complete separation of data points and thus regression coefficients could

not be reported for these vignettes. This lack of variance can be seen in the very high reporting

percentages given in Table 3-1 for these vignettes. For Vignette 5, all respondents chose to

report the incident, leaving no variance to predict. Also, one predictor variable, reporter

race/ethnicity, was removed from further analyses because lack of sample size at certain levels of

the ethnicity variable caused quasi-complete separation of data points in all 8 vignettes.

Results indicate that time in social services was found to significantly predict willingness

to report for Vignette 3 (b = -.07, p = 0.00). The relationship is negative, indicating that

investigators with more time spent working in social services are less likely to report abuse.

However, the odds ratio of 0.93 indicates that individuals with higher tenure are approximately










just as likely to report as those with lower tenure, so this effect may not have practical

significance.










Table 3-1. Descriptive statistics for vignettes by form


Percent of respondents
reporting


Severity mean rating


Black White No race Black
child child given child

5.49 5.68 5.49 81%

6.46 6.55 6.26 100%

5.39 5.66 5.24 79%

5.31 5.17 5.33 71%

5.31 5.35 4.92 79%


White
child

80%

98%

83%

66%

68%


race
given

77%

88%

69%

78%

65%


Vignette

All 8 vignettes

1: Child wets bed and is immersed in hot water

2: Father punches drunk teen

3: Mother calls child evil and hits with broom

4: Parents lock 14-year-old in his room

5: Parents beat 11-year-old with cane and burn
a mark on his arm for stealing

6: Child mumbled, parent banged child against
wall, bruising his shoulders

7: Parents make child kneel in closet

8: 9-year-old left alone most of day and night,
parents working


6.57 6.74


6.65 100% 100% 100%


5.83

4.77


6.14

4.94


6.14

4.67


96%

69%


96%

60%


96%

59%


4.27 4.86


4.69 52% 68% 63%


Note. black child: N=52, white child: N=47, no race given: N=51










Table 3-2. Correlation matrix


Variables of interest 1 2 3 4 5

1. Overall severity rating

2. Gender-03

3. Ethnicity -.6 01

4. Rural vs. urban county-04 0.5 .6

5. CSThours0.16 0.02 -0.10 -0.05

-0.20* 0.09 0.19* 0.02 0.08
6. Years in social services
Note. CST = cultural sensitivity training.
*p < .05.










Table 3-3. ANOVA analyses for severity rating by form

Vignette SS

Overall severity rating1.6
2.15
1: Child wets bed and is immersed in hot water

2: Father punches drunk teen4.7
~0.74
3: Mother calls child evil and hits with broom
5.61
4: Parents lock 14-year-old in his room

5: Parents beat 11-year-old with cane and burn a
~0.78
mark on his arm for stealing

~n6: Child mumbled, parent banged child against
w. 3.28
wall, bruising his shoulders
~1.80
7: Parents make child kneel in closet

8: 9-year-old left alone most of day and night,
~9.27
parents working


MS
0.58

1.08

2.24

0.37

2.81



0.39



1.64

0.90


F

1.02

1.33

1.34

0.26

1.28



0.88



1.82

0.42


_P
0.36

0.27

0.26

0.77

0.28



0.42



0.17

0.66


2 4.64


2.53 0.08












Table 3-4. Regression analyses for severity rating
Black White Respondent Respondent Respondent Rural
child vs. child vs. ethnicity: ethnicity: ethnicity: vs. Years in
no race no race black vs. hispanic vs. Other vs. CST urban social Adjusted
Vignette given given Gender white white white hours county services R2


Overall severity rating

1: Child wets bed and is
immersed in hot water


2: Father punches drunk teen

3: Mother calls child evil and
hits with broom

4: Parents lock 14-year-old in
his room

5: Parents beat 11-year-old
with cane and burn a mark
on his arm for stealing

6: Child mumbled, parent
banged child against wall,
bruising his shoulders

7: Parents make child kneel in
closet

8: 9-year-old left alone most
of day and night, parents
working


-0.04


0.23


0.11


-0.14


0.32




-0.07




-0.41


0.07


0.18


0.30


0.38


-0.19


0.44




0.08




-0.01


0.22


-0.01


-0.20


0.10


-0.13


-0.07




0.04




0.01


0.05


0.16


0.12


-0.07


0.36


0.21




-0.08




0.33


-0.06




0.43


0.15


0.20


1.15


0.18


-0.46




0.16




0.15


-0.25




0.06


0.12


0.33


0.56


0.28


-0.52




0.23




0.00


0.08




-0.03


0.00 -0.05 -0.02


0.00 -0.05 0.00


0.04


-0.01


0.00 -0.28 -0.04* 0.09


0.00 0.27 -0.04* 0.06


0.00 0.10


-0.02


0.01




-0.02


0.00 0.12 0.00


0.00 0.24 -0.04* 0.12


0.00 -0.52 -0.02




0.00 -0.25 -0.01


0.04




-0.00


-0.42


0.18


0.16


Note. CST = cultural sensitivity training; values are b-weights.
*p < .05/8 = .006.










Table 3-5. Logistic regression analyses for willingness to report
Black child White child
vs. no race vs. no race Rural vs. Years in social
Reporting of vignette given given Gender CST hours urban county services
-0.15 -0.14 -0.11 0.00 0.16 -0.03
Percentage of vignettes reported overall 0.74 0.75 0.80 1.00 1.38 0.97
-0.29 -0.39 0.06 -0.00 0.22 0.01
2: Father punches drunk teen 0.56 0.46 1.31 1.00 1.57 1.01
3: Mother calls child evil and hits with 0.33 0.33 -0.28 -0.00 -0.31 -0.07*
broom 1.94 1.94 0.57 1.00 0.54 0.93
-0.31 -0.06 -0.08 0.00 0.10 -0.04
4: Parents lock 14-year-old in his room 0.54 0.89 0.86 1.00 1.23 0.96
-0.27 -0.03 -0.13 0.01 0.36 -0.00
7: Parents make child kneel in closet 0.58 0.93 0.76 1.01 2.04 1.00
8: 9-year-old left alone most of day and 0.21 -0.10 -0.13 0.00 0.17 0.00
night, parents working 1.53 0.83 0.78 1.00 1.42 1.00
SNote. CST cultural sensitivity training; first values in each cell are b-weights, second values are odds ratio estimates.
*p < .05/8 = .006.









CHAPTER 4
DISCUSSION

This study achieved the goals of investigating whether (1) Child Protective Investigators

rate severity of maltreatment differently based on victim's race/ethnicity; (2) Child Protective

Investigators rate severity of maltreatment differently based on characteristics such as

race/ethnicity, gender, amount of cultural sensitivity training of the reporter, and setting of

employment; and (3) if the child' s race/ethnicity, reporter race/ethnicity, reporter gender, amount

of cultural sensitivity training of the reporter, or the population of the county in which the

reporter works affects the investigators willingness to report. While some statistically significant

results were found among the exploratory analyses, none of the above hypotheses were

supported.

The results demonstrated that one of the reporter characteristics examined during

exploratory analyses, years spent in social services, or tenure on the job, was found to

significantly predict severity ratings across three different vignettes as well as the decision to

report on one vignette. The relationship was negative, indicating that participants with more

tenure in social services were more likely to rate vignettes as less severe and were less likely to

report the abuse. This reporter characteristic has not received much attention in the literature and

thus was not one of central focus before completing this study. Research completed by

Crenshaw, Crenshaw and Lichtenberg (1995) investigated the reporting behaviors of teachers,

counselors, school psychologists, and principles. The results of that study showed no

relationship between educator tenure and reported severity among five scenarios of child abuse.

Similarly, Warner-Rogers, Hansen, and Spieth (1996) who sampled third and fourth year

medical students, found no influence of experience. Uniquely, the current sample was comprised

of a large range of experience levels, ranging from 1 month on the job to 38 years of experience.










The present Eindings could indicate that Child Protective Investigators become desensitized to

situations involving child maltreatment and consequently rate them as less severe over time.

Along with results found in Ashton (1999), Hansen et al. (1997), and Warner-Rogers,

Hansen, & Spieth (1996), who utilized medical students, psychologists, social workers, and

graduate students, the current study found no evidence of gender differences among reported

severity ratings. Previous studies that have not found gender to be a significant predictor of

severity ratings have utilized overwhelmingly female and Caucasian populations, all calling for a

more culturally diverse sample. The sample for the present study was more diverse than samples

have been in previous studies. Additionally, the significance of this research is its focus on Child

Protective Investigators, the individuals deemed with the responsibility of investigating and

intervening in cases of child abuse. Thus, this finding may further indicate that gender

differences may not exist within the system and should be used to further evaluate evidence

among this population.

Similarly, reporter ethnicity was not found to be a significant predictor of severity ratings.

Evidence that reporter ethnicity does not affect judgments concerning child maltreatment is

inconsistent with research completed by Ashton (2004) and Hansen et al. (1997). One

explanation for this inconsistency may be that previous studies that have found differences in

reporting rates based on reporter ethnicity have only sampled mandated reporters and not Child

Protective Investigators, the individuals responsible for verification. Thus, one might

hypothesize that mandated reporters, and not Child Protective Investigators, are responsible for

the overrepresentation of minority children in the reporting rates. Additionally, reporter ethnicity

had to be removed as a predictor variable for willingness to report due to lack of sample size at

certain levels of ethnicity. Thus, while this study aimed to utilize a more diverse sample, there









were not enough participants within each ethnicity to analyze whether or not it significantly

predicted willingness to report.

The population of the county in which one works was not found to significantly predict

reporting behavior. This result is in contrast to previous research conducted by Webster,

O'Toole, O'Toole, & Lucal (2005) who found an effect for reporting behavior among teachers in

rural and suburban districts, with rural teachers being less likely to report. Traditionally, rural

areas have a higher percentage of families living in poverty than that of urbanized areas.

Researchers have hypothesized that people living in rural areas may have more well-developed

social networks and a strong sense of community and thus may rely on informal channels of

support when dealing with child maltreatment (Schuck, 2005). With such a limited number of

Child Protective Investigators being employed within rural counties, it made recruiting an equal

number of participants for this study impossible. Future research should focus on recruiting an

equal number of participants from each area to further investigate this variable as a possible

predictor of reporting behavior.

Finally, no evidence was found to support that the ethnicity of the child depicted in the

scenarios affects the severity ratings or decision to report among the Child Protective

Investigators sampled. The finding that race of the child does not affect judgments of abuse is

consistent with previous research completed by Egu & Weiss (2003) and Turbett & O'Toole

(1983). In the previous studies that did find the race of the child as a significant factor in

reporting, the populations sampled included physicians, clinical psychologists, and social

workers (Nalepka, O'Toole & Turbett, 1981; Hansen et al., 1997). While being mandated

reporters, these professionals have not had the specific training that actual Child Protective

Investigators complete. This finding provides evidence against current theory concerning the









overrepresentation of minority children within the child welfare system. Barth (2005) maintains

that the race of the child plays a role in every decision making point throughout the child welfare

system. The current study found that the ethnicity of the child was one of the only characteristics

investigated that did not affect reporter behavior. Additionally the reported amount of cultural

sensitivity training was not found to be a significant predictor of severity ratings or willingness

to report. Thus, it did not appear that the amount of cultural sensitivity training received affected

one's judgment of child maltreatment.

Limitations

First, it is not known whether the present Eindings generalize to other locations. Census

data suggest that the demographic characteristics of residents of Florida are similar to the nation

as a whole; however, Florida has one of the highest child victimization rates in the nation

(Schuck, 2005). Additionally, only seventeen of Florida' s sixty-seven counties were sampled in

this study. Furthermore, the sample was not randomly selected and utilized the first 150 Child

Protective Investigators agreeing to participate within the seventeen most immediate counties

surrounding the University of Florida.

The present study utilized vignette and self-report methodology. Real-life situations

contain many additional facts and ambiguities that affect personal judgments about suspicion and

reporting of child maltreatment. Anecdotally, a large percentage of the participants reported

feeling uncomfortable with making a decision based on such ambiguous information. Although

use of hypothetical cases of maltreatment presented in vignettes permit the systematic

manipulation of case variables, not all important case variables can be included in one study.

Thus, a limitation of the current study was using gender as a control in order to isolate the

ethnicity of the child. The child in the scenarios was always male, thus limiting the










generalizability of the study. Future research should replicate this study investigating if an effect

of child gender is present within this population.

While no significant effects of child ethnicity were found, participants demonstrated a

tendency to give less severe ratings on the vignettes depicting the victim as African-American.

This trend is consistent with the results found in the literature. While a between-subj ects design

brings certain advantages to this study (e.g., random assignment, reduction of effects of demand

characteristics and social desirability), it may have masked true differences because participants

were reluctant to answer honestly for fear of being perceived as racist when all the vignettes

depicted minority children. Other studies, on the other hand, utilized within- subj ects designs that

may overestimate effects. Thus, a confounding issue in completing research of this nature is

getting accurate and honest responses from workers within the Hield. Anecdotally, many of the

participants in this study expressed concern upon receiving Form A with the children depicted as

black. This may be representative of what an emotionally charged issue culture is within the

child welfare system. Future research should work on finding an alternate method of gathering

reporter data which allows the respondent to answer honestly and without the influence of social

desirability.

Additionally, only Hyve vignettes were able to be used in analysis of the third hypothesis

due to lack of variance in the sample on the willingness to report outcome for these vignettes.

For vignette 5, all respondents chose to report the incident, leaving no variance to predict. Also,

one predictor variable, ethnicity, was removed from further analyses because lack of sample size

at certain levels of the ethnicity variable. While years in social services was found to be

statistically significant, only a small amount of the variance in predicting severity ratings and

willingness to report was accounted for by this variable. Future research should focus on









increasing the effectiveness of the vignettes used in the study. The present study yielded results

which showed a restriction in range of the respondents reporting of severity levels. Future

studies should utilize vignettes which allow the respondents to use the lower values of the scale,

creating more variability in responses.

Conclusion

Despite its limitations, this study provides important information about a previously

unstudied population. The significance of this research is its focus on Child Protective

Investigators, the individuals responsible for investigating and verifying cases of child

maltreatment. While CPS workers comprise a neglected source of data, their reporting behaviors

have gained much attention and speculation among researchers. There have been increasing

concerns about the overrepresentation of minority children within the child welfare system in

this country. Researchers have spent much time and attention conducting studies with

undergraduates, graduate students, teachers, physicians, and social workers and have

extrapolated results to hypothesize reporting behaviors of CPS workers. In his theoretical model

pertaining to this overrepresentation, Barth (2005) maintains that agency characteristics allow the

race of the child to play a role in every decision making point throughout the child welfare

system. More specifically, factors such as "staffing patterns, culturally incompetent staff,

institutionalized racism, and inadequate duration or configuration of services" (p.36) are

provided as possible contributing characteristics. The present research study provides direct

evidence that the race of the child does not play a role in the severity ratings or decision to report

among actual Child Protective Investigators.

The present study also adds to our understanding of what factors may influence

identification and reporting of maltreatment. Tenure on the job has received little to no previous

attention in the literature as a reporter characteristic of interest. Thus, this characteristic should









be the focus of future research among this population to help in further understanding the effect

it has on reporting behavior. The findings presented here should be viewed as exploratory in

nature and warrant replication with larger groups of investigators. Readers are reminded the

demographic variables chosen for this study represent only a part of an individual's psychosocial

characteristics. Further research is needed to better understand the underlying factors within

cultures that affect differences in reporting.









APPENDIX A
DEMOGRAPHIC QUESTIONNAIRE

Answer the following questions:

1) Please circle the group that best describes your own cultural identification?

African-American or Black
Caucasian, or White
Hispanic, or Latino
Some other group (specify)

2) What is your gender? Male or Female

3) What country were you born in?

4) If you were not born in the U. S., how long have you lived here?

5) What is your religious affiliation?

6) What type of work does (did) your father do?

Professional
Skilled
Semi-skilled
Unskilled

7) What is your current age?

8) Do you have a childhood history of harsh discipline?

9) What is your marital status?

Single
Married
Divorced
Widowed
Living with Significant Other

10) Do you have any children?

Yes (if yes, how many?)
No


11) What is your highest level of education?










High School Diploma
Some College
Bachelors Degree
Masters Degree
Ph.D.

12) What field is your highest degree in? (e.g., social work, education, etc.)



13) How long have you worked in social services (include time with present employer as
well as other agencies)?

years months

14) Please indicate your current status with the Department of Children and Familes:

Child Protective Investigator
Child Protective Service Worker
Foster Care Worker
Adoption Worker
Licensing Worker
Other (explain)

15) What ethnicity/race is the maj ority of the clientele you serve?

16) Which best describes the county in which you work?

Rural
Urban/Suburban

17) How much cultural sensitivity training have you received (include time with present
employer as well as other agencies)?

hours


18) Please identify where you received the above training (Circle all that apply)

Preservice: Undergraduate
Entire course
Part of a course
Graduate School (Before current employment)
Entire course
Part of a course
Inservice (During current employment)










Workshop
Conference
Other









APPENDIX B
VIGNETTES

Please imagine that you are a case worker approached by a teacher who discloses the following
information.

1. During a parent-teacher conference, a mother discloses to a 2nd grade teacher that her
eight-year-old son wets the bed occasionally. The last time he wet the bed his mother
immersed him in a tub of very hot water. His mother explained that the child will not wet
the bed again if he associates bed wetting with the punishment.

How serious do you feel the above incident is on a Likert scale of "seriousness"
From 1 to 7, with "1" being not serious and "7" being very serious?

1 2 3 4 5 6 7

How would you recommend that the educator respond given the above scenario?

(1) no action, nothing needs to be done about the situation
(2) encourage the family to seek professional help
(3) notify the child protection agency to investigate the situation

2. A high school teacher found out that a 17-year-old boy in her class came home drunk and
began to argue with his mother last weekend. His father came into the room just as he
yelled an obscene remark at his mother. His father punched him in the face with his fist.

How serious do you feel the above incident is on a Likert scale of "seriousness"
From 1 to 7, with "1" being not serious and "7" being very serious?

1 2 3 4 5 6 7

How would you recommend that the educator respond given the above scenario?

(1) no action, nothing needs to be done about the situation
(2) encourage the family to seek professional help
(3) notify the child protection agency to investigate the situation

3. During outside play, a six-year-old boy reveals to his teacher that her parents are
separated. The child lives with his mother and occasionally visits his father. Whenever
he returns from a visit with his father and talks about the visit, his mother tells him he is
evil and hits him with a broom.

How serious do you feel the above incident is on a Likert scale of "seriousness"
From 1 to 7, with "1" being not serious and "7" being very serious?

1 2 3 4 5 6 7









How would you recommend that the educator respond given the above scenario?

(1) no action, nothing needs to be done about the situation
(2) encourage the family to seek professional help
(3) notify the child protection agency to investigate the situation

4. A 14-year-old boy tells his history teacher that his parents found out that he is dating
outside of his race, which they don't approve of. He has lied about meeting his girlfriend
and he has snuck out of the house on several occasions to see her. When his parents
caught him trying to sneak back into the house one morning, they sent him to his room
and locked him in for two days, allowing him out only to go to the bathroom.

How serious do you feel the above incident is on a Likert scale of "seriousness"
From 1 to 7, with "1" being not serious and "7" being very serious?

1 2 3 4 5 6 7

How would you recommend that the educator respond given the above scenario?

(1) no action, nothing needs to be done about the situation
(2) encourage the family to seek professional help
(3) notify the child protection agency to investigate the situation

5. An 11i-year-old boy stole candy from a store. He had been caught stealing before. When
his parents found out that he had been stealing again, they beat him with a switch and
burned a mark on his arm. They said the mark would remind him not to steal again. He
told his P.E. teacher on Monday morning while he was dressing out for gym class.

How serious do you feel the above incident is on a Likert scale of "seriousness"
From 1 to 7, with "1" being not serious and "7" being very serious?

1 2 3 4 5 6 7

How would you recommend that the educator respond given the above scenario?

(1) no action, nothing needs to be done about the situation
(2) encourage the family to seek professional help
(3) notify the child protection agency to investigate the situation

6. The parents of an 8-year-old boy want him to answer clearly when spoken to and have
warned him in the past about mumbling. Recently when his mother asked him a
question, he mumbled the answer under his breath. The mother banged the child against
the wall, bruising his shoulders. He told the speech pathologist what happened during
their weekly session.

How serious do you feel the above incident is on a Likert scale of "seriousness"









From 1 to 7, with "1" being not serious and "7" being very serious?


1 2 3 4 5 6 7

How would you recommend that the educator respond given the above scenario?

(1) no action, nothing needs to be done about the situation
(2) encourage the family to seek professional help
(3) notify the child protection agency to investigate the situation

7. A family recently arrived in the metropolitan area from a rural location. They are very
strict with their ten-year-old son. They want him to go to school and come right home.
Recently their son went to a school friend's house instead of coming straight home. His
parents disciplined him in their usual way, which is to make him kneel in the closet for
several hours. The parents believe the closeness of the closet and the discomfort of
kneeling for long periods of time reinforce the household rules.

How serious do you feel the above incident is on a Likert scale of "seriousness"
From 1 to 7, with "1" being not serious and "7" being very serious?

1 2 3 4 5 6 7

How would you recommend that the educator respond given the above scenario?

(1) no action, nothing needs to be done about the situation
(2) encourage the family to seek professional help
(3) notify the child protection agency to investigate the situation

8. Both parents work long hours; they leave very early in the morning and come home late
at night. Their nine-year-old son is often on his own. The boy gets himself ready for
school in the morning and lets himself in after school. The parents tell their son to eat
food prepared and left in the refrigerator, warming it up if he wants. He usually eats it
cold. He goes to bed by himself because his parents are usually not back by his bedtime.

How serious do you feel the above incident is on a Likert scale of "seriousness"
From 1 to 7, with "1" being not serious and "7" being very serious?

1 2 3 4 5 6 7

How would you recommend that the educator respond given the above scenario?

(1) no action, nothing needs to be done about the situation
(2) encourage the family to seek professional help
(3) notify the child protection agency to investigate the situation









APPENDIX C
INFORMED CONSENT

Protocol Title: The effects of reporter and child characteristics on the identification and
willingness to report child maltreatment

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 the effects of reporter and child characteristics on the
identification and willingness to report child maltreatment.

What you will be asked to do in the study:

You will be asked to complete a packet that contains eight vignettes depicting different scenarios
followed by a demographic questionnaire. No identifying information (name, social security
number, etc.) should be included in order to ensure confidentiality. Each vignette depicts a
fictitious description of problematic parental behaviors. You will then be asked to read each
vignette and rate the information on a Likert scale of "seriousness" from 1 to 7, with "1" being
not serious and "7" being very serious. After you rate the seriousness of the scenario, you will
then be asked to indicate how you would tell the educator to respond by selecting one of three
options: (1) no action, nothing needs to be done about the situation; (2) encourage the family to
seek professional help; or (3) notify the child protection agency to investigate the situation.

Time required:

15 minutes

Risks and Benefits:

We do not anticipate any benefits or risks associated with participation in this study.

Compensation:

No compensation will be given for participation in this study.

Confidentiality:

No identifying information will be gathered during data collection. Your demographic data and
vignette response sheet will be kept anonymously and assigned a code number to ensure
confidentiality.

Voluntary participation:

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










Right to withdraw from the study:

You have the right to withdraw from the study at anytime without consequence.

Whom to contact if you have questions about the study:

Danielle Madera, M.Ed., School Psychology Graduate Student, University of Florida,
Gainesville, FL 32608; (352) 334-1300.

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:

I have read the procedure described above. I voluntarily agree to participate in the procedure and
I have received a copy of this description.



Participant: Date:



Principal Investigator: Date:









REFERENCE LIST


American Psychological Association. (2002). Guidelines on multicultural education, training,
research, practice, and organizational change for psychologists. Retrieved December
12, 2005, from http ://www.apa.org/pi/multiculturalguidelines/oeaehm

Ards, S., Chung, C., & Myers, S. L. (1998). The effects of sample selection bias on racial
differences in child abuse reporting. Child Abuse & Nelgect, 22, 103-115.

Ashton, V. (1999). Worker judgements of seriousness about and reporting of suspected child
maltreatment. Child Abuse & Neglect, 23(6), 539-548.

Ashton, V. (2004). The effect of personal characteristics on reporting child maltreatment. Child
Abuse and Neglect, 28, 985-997.

Barth, R. P. (1997). Effects of age and race on the odds of adoption versus remaining in long-
term out-of-home care. Child Welfare, 76, 285-308.

Barth, R. P. (2005). Child welfare and race: Models of disproportionality. In D. M. Derezotes, J.
Poertner, & M. F. Testa., Race matters in Child Welfare (pp. 25-46). Washington, DC:
Child Welfare League of America, INTC.

Baumrind, D. (1995). Child maltreatment and optimal care giving in social contexts. New York,
NY: Garland Publishing, Inc.

Behl, L.E., Crouch, J. L., May, P. F., Valente, A. L., & Conyngham, H. A. (2001). Ethnicity in
child maltreatment research: A content analysis. Child2\a~ltreatment, 6(2), 143-147.

Belsky, J. (1980). Child maltreatment: An ecological integration. American Psychologist, 35(4),
320-335.

Bronfenbrenner, Y. (1979). The ecology of family development. Cambridge, MA:Harvard
University Press.

Bui, K. & Takeuchi, D. T. (1992). Ethnic minority adolescents and the use of communitymental
health care services. American Journal of Community Psychology, 20, 403-417.

Champion, K., Shipman, K., Bonner, B., Hensley, L., Howe, A. (2003). Child maltreatment
training in doctoral programs in clinical, counseling, and school psychology: Where do
we go from here? Child Maltreatment, 8, 211-217.

Cohen, J. A., Deblinger, E., Mannarino, A. P., & Arellano, M. A. (2001). The importance of
culture in treating abused and neglected children: An empirical review. Child
Maltreatment, 6(2), 148-157.

Coohey, C. (2001). The relationship between familism and child maltreatment in Latino and
Anglo families. Child2\a~ltreatment, 6(2), 130-142.










Crenshaw, W. B., Crenshaw, L. M., & Lichtenberg, J. W. (1995). When educators confront
child abuse: An analysis of the decision to report. Child Abuse & Neglect, 19(9), 1095-
1113.

Deater-Deckard, K., Dodge, K. A., Bates, J. E., & Pettit, G. S. (1996). Physical discipline among
African-American and European-American mothers: Links to children's externalizing
behaviors. Developmental Psychology, 32, 1065-1072.

Drake, B., & Pandey, S. (1996). Understanding the relationship between neighborhood poverty
and specific types of child maltreatment. Child Abuse & Neglect, 20(11), 1003-1018.

Dubowitz, E. (1999). Neglected children: Research, practice, and policy. Thousand Oaks,
California: Sage Publications.

Egeland, B. A history of abuse is a maj or risk factor for abusing the next generation. In: R. J.
Gelles, D. R. Loseke, Eds. Current controversies on family violence (pp. 197-208)
Thousand Oaks, CA, Sage, 1993.

Egu, C. L., & Weiss, D. J. (2003). The role of race and severity of abuse in teachers' recognition
or reporting child abuse. Journal of Child and Family Studies, 12(4), 465-474.

Ferrari, A. M. (2002). The impact of culture upon child rearing practices and definitions of
maltreatment. Child Abuse & Neglect, 26, 793-8 13.

Fluke, J. D., Yuan, Y. T., Hedderson, J., & Curtis, P. A. (2003). Disproportionate representation
of race and ethnicity in child maltreatment: Investigation andvictimization. Children
and Youth Services Review, 25, 359-373.

Fontes, L. A., Cruz, M., & Tabachnick, J. (2001). Views of child sexual abuse in two cultural
communities: An exploratory study among African-Americans and Latinos. Child
Maltreatment, 6(2), 103-117.

Garbino, J., & Sherman, D. (1980). High-risk neighborhoods and high-risk families:The human
ecology of child maltreatment. ChildDevelopnzent, 51(1), 188-198.

Garland, A. F., & Beinger, B. A. (1997). Racial/ethnic differences in court referred pathways to
mental health services for children in foster care. Children and Youth Services Review,
19(8), 651-666.

Gelles, R. J. (1992). Poverty and violence toward children. American Behavioral Scientist,
35(3), 258-274.

Green, J. W. (1982). Cultural awareness in the human services. Englewood Cliffs, NJ:Prentice-
Hall.










Hansen, D. J., Bumby, K. M., Lundquist, L. M., Chandler, R. M., Le, P. T., & Futa, K. T. (1997).
The influence of case and professional variables on the identification and reporting of
child maltreatment: A study of licensed psychologists and certifiedmasters social
workers. Journal ofFa~nily Violence, 12(3), 313-332.

Hines, A. M., Lemon, K., Wyatt, P., & Merdinger, J. (2004). Factors related to the
disproportionate involvement of children of color in the child welfare system: A review
and emerging themes. Children and Youth Services Review, 26, 507-527.

Hinson, J., & Fossey, R. (2000). Child abuse: What teachers in the '90s know, think, and do.
Journal of Education for Students Placed at Risk, 5(3), 25 1-66.

Horton, C., & Cruise, T. (2001). Child abuse & neglect: The school's response. New York,
NY: The Guilford Press.

Kempe, C. H., Silverman, F. N., Steele, B. F., Droegmueller, W. & Silver, H. K. (1962). The
Battered Child Syndrome. Journal of the American M~edical Association, 1 8 1, 1 7-24.

Kenny, M. (2001). Child abuse reporting: Teachers' perceived deterrents. Child Abuse &
Neglect, 25, 81-92.

Korbin, J. E. (2002). Culture and child maltreatment: Cultural competence and beyond. Child
Abuse & Neglect, 26, 637-644.

Korbin, J. E., Coulton, C. J., Lindstrom-Ufuti, H., & Spilsbury, J. (2000). Neighborhood views
on the definition and etiology of child maltreatment. Child Abuse & Neglect, 24(12),
1509-1527.

Levesque, R. J. (2000). Cultural evidence, child maltreatment, and the law. Child2\a~ltreatnzent,
5(2), 146-160.

Lu, Y. E., Landsverk, J., Ellis-Macleod, E., Newton, R., Ganger, W., & Johnson, I. (2004).
Race, ethnicity, and case outcomes in child protective services. Children and Youth
Services Review, 26, 447-461.

Maynard, C., & Wiederman, M. (1997). Undergraduate students' perceptions of child sexual
abuse: Effects of age, sex, and gender-role attitudes. ChildAbuse & Neglect, 21(9),
833-844.

McCabe, K. M., Yeh, M., Hough, R. L., Landsverk, J., Hurlburt, M. S., Culver, S. W., &
Reynolds, B. (1999). Racial/ethnic representation across five public sectors of care for
youth. Journal of Enotional and Behavioral Disorders, 7, 72-82.

Milner, J. S. (1998). Individual and family characteristics associated with intrafamilialchild
physical and sexual abuse. In P.K. Trickett & C.J. Schedllenbach (Eds.), Violence against
children in the family and the conanunity (pp. 141-170). Washington, DC: American
Psychological Association.










Morrow, K. B., & Sorrell, G. T. (1989). Factors effecting self-esteem, depression and negative
behaviors in sexually abused female adolescents. Journal of2arriage and the Family,
51, 677-686.

Nybell, L. M., & Gray, S. (2004). Race, place, space: Meanings of cultural competence inthree
child welfare agencies. Social Work, 49(1), 17-26.

Rao, K., DiClemente, R. J., & Ponton, L. E. (1992). Child sexual abuse in Asians compared with
other populations. Journal of the American Academy of Child Adolescent Psychiatry,
31(5), 880-886.

Scannapieco, M., & Connell-Carrick, K. (2005). Understanding child maltreatment.New York,
NY: Oxford University Press.

Schuck, A. M. (2005). Explaining black-white disparity in maltreatment: Poverty, female-
headed families, and urbanization. Journal of2arriage and Family,~FFF~~~FFF~~~F 67, 543-551.

Sullivan, P. M., & Knutson, J. F. (2000). Maltreatment and disabilities: A population-based
epidemiological study. Child Abuse & Neglect, 24, 1257-1273.

Terao, S. Y., Borrego, J., & Urquiza, A. J. (2001). A reporting and response model for culture
and child maltreatment. Child2\a~ltreatment, 6(2), 158-168.

U.S. Department of Health and Human Services, Administration on Children, Youth and
Families. Child2\a~ltreatment 2004 (Washington, DC: U.S. Government Printing Office,
2006).

U.S. Department of Health and Human Services, Children's Bureau. Child2\a~ltreatment 1998:
Reports from the States to the National Child Abuse and Neglect Data System
(NCANDS). (Washington, D.C.: U.S. Government Printing Office, 2000).

Warner-Rogers, J. E., Hansen, D. J., & Spieth, J. E. (1996). The influence of case and
professional variables on identification and reporting of physical abuse: A studywith
medical students. Child Abuse & Neglect, 20(9), 851-866.

Webster, S. W., O'Toole, R., O'Toole, A. W., & Lucal (2005). Overreporting and
Underreporting of child abuse: Teachers' use of professional discretion. ChildAbuse &
Neglect, (29), 1281-1296.

Wissow, L. S. (2001). Ethnicity, income, and parenting contexts of physical punishment in a
national sample of families with young children. Child2\a~ltreatment, 6(2),118-129.









BIOGRAPHICAL SKETCH

Danielle Dawn Madera was born in Valley Stream, NY on May 5, 1978, the daughter of

Demetrius and Christine Madera. Danielle graduated from the University of Florida in May

2000, with a B.S. in psychology. In August 2000, she began working as a Family Service

Counselor for the Department of Children and Families in St. Petersburg, FL. After a year of

experience working with child maltreatment, she decided to begin her graduate education. In

August 2001, Danielle entered the School Psychology Doctoral Training Program at the

University of Florida. At the University of Florida, Danielle pursued research interests in

child maltreatment. She received her Master of Education degree from the University of Florida

in May 2005. Danielle then moved on to complete her advanced practicum placement with

the University of Florida Child Protection Team. After completing her coursework,

Danielle obtained an APPIC internship at the Children's Assessment Center in Houston, TX,

where she works providing therapy and psychological services to sexually abused children and

their non-offending caregivers. Having successfully defended her dissertation, Danielle

graduated in May 2007 with her Doctor of Philosophy degree from the University of Florida.

She hopes to continue working with victims of child maltreatment and continue her research

efforts in that field.