The National Children’s Study Workshop on Measuring Racial / Ethnic Discrimination and Racism from a Developmental Perspective, was initiated by the Health Disparities and Environmental Justice Working Group of the National Children’s Study Advisory Committee. The goal was to summarize findings concerning how best to measure racism and discrimination in multiple racial / ethnic groups from a longitudinal, epidemiological perspective.
Background
In April, 2002, The Office of Behavioral and Social Sciences Research of the National Institutes of Health convened a meeting of approximately 100 scientists to present scientific evidence of the effects of racial / ethnic bias on physical and mental health (1). In addition to highlighting findings related to associations between racism / discrimination and health outcomes, this comprehensive review of existing literature focused on directions for future research. Among their findings was the conclusion that the majority of empirical research had been done on African Americans and that few studies have systematically addressed how prejudice and discrimination affect other racial / ethnic minority groups (1), and that there is no consensus in the literature as to the optimal measures for capturing exposure to discrimination (2). The latter point is especially relevant for children, where the least amount of research has been done.
A report from the National Research Council, compiled by the Panel on Methods for Assessing Discrimination of the Committee on National Statistics, notes that racial ethnic disparities in health outcomes are well known. The fact that they exist is not proof that racism or discrimination is a causal factor, however it does motivate further investigation into the issue. That report uses a definition of discrimination which goes beyond the legal definition to encompass differential treatment on the basis of race/ ethnicity that disadvantages any racial / ethnic group. The purpose is to include behavior that has negative consequences but may not be unlawful. It categorizes discrimination into four areas: discrimination that is intentional and explicit; discrimination that is subtle, unconscious and automatic; statistical discrimination and profiling; and exclusionary organizational processes. This report further reflects the inherent difficulties in measurement and reiterates the difficulties in defining racial discrimination in a clear way so that credible ways of measuring can be found (3). It says that "For the purpose of understanding and measuring racial discrimination, race should be viewed as a social construct that evolves over time." However, there is no single approach to measuring racial discrimination that allows researchers to address all the important measurement issues or to answer all the questions of interest. Because measurement of both race/ethnicity and racial discrimination can vary with the method used, it recommends that whenever possible, multiple methods be employed in the same study so that comparisons can be made across methods.
The purpose of the National Children’s Study Workshop on Measuring Racial / Ethnic Discrimination and Racism was to address the issues of measurement from a lifespan perspective, identifying important contextual factors and feasibility issues relevant to a prospective epidemiologic study of children.
Constructs Relevant to Various Ethnic / Racial Populations over Time
The Report from the National Research Council defined race as a social construct that evolves over time. This concept was echoed in comments made by participants related to the "concept of self" and racial identity in children, where it was noted that it is important to put racism in a context. The context of change over time is important not only within individuals, i.e., as children grow and develop, but also with respect to societal trends as concepts of race / ethnicity evolve over time. Given the longitudinal nature of the National Children’s study, societal changes are likely to occur and should be documented, especially those events that differentially impact racial / ethnic groups.
From a developmental perspective, attributions such as racial/ ethnic identity evolve and change as children grow older. Part of this process stems from learning racial language as they begin to socialize outside the home in day care settings and schools. The implication is that ethnic identity may be formulated as a result of racism. However, a sense of "place" is also an important issue for racial / ethnic identity. In its most basic context, place simply means a geographic location. But from a social perspective it is also a nexus where social life is initiated and engaged, and involves the values, traditions and history we use for organizing our experiences. A sense of place gives children a sense of security and engagement. Exclusion from social networks can be a form of denying children a sense of place. Children who are caught between the values of two cultures, the dominant white culture to which they do not belong, and their own race / ethnic group, also have more complex issues with respect to a sense of identity than children from the dominant culture. The stability of a child’s concept of self may also vary by immigration status. African Americans have a history of discrimination in this country and become aware of it at an early age. It is a stable cultural reality for African Americans in our society. Whereas for immigrants, a child’s concept of self and who s/he is may be more inclined to vary over time, depending on their particular culture and the attitudes of Americans towards that culture. A case in point is Muslim children who may be experiencing much more negative feedback about their identity since the events of 9/11 than before.
The example of Muslim children after 9/11 also illustrates that discrimination occurs in a social-cultural context and may manifest differently in different race / ethnic populations. Ways of coping with discrimination also vary by population. It has been demonstrated that passive, forbearance coping reactions are more effective in Southeast Asians with strong attachment to Asian ethnic values than in Koreans (4). In Korean immigrants, active, problem-focused coping was most effective, especially among those who were better acculturated. Although Vietnamese, Chinese and Korean immigrants share the same cultural traditions of Confucianism, Buddhism, and Taoism that would point to a predominance of forbearance coping, Korean immigrants are much more likely to be affiliated with protestant churches, have more members at higher SES levels and are better acculturated.
It was emphasized throughout the meeting that the association of discrimination with physical and mental health outcomes stems from it cumulative effect. Measuring at one point in time does not provide an adequate conceptualization for meaningful analyses (although at one interview one can ask about lifetime experiences). In the case of perceived (and typically interpersonal) discrimination, it is the cumulative effects that result in the primary emotions of anger, sadness and aggression and the cognitive sense of shame, powerlessness or lack of personal control, exclusion and discouragement. Other forms of harm arising from institutional and statistical discrimination likewise can be cumulative in impact, e.g., impact of environmental injustice, lack of asset accumulation due to low wage jobs, etc. The cumulative effects are of several types: generational, i.e., passed from parent to child; longitudinal within domains, such as discrimination within the school system over time from elementary grades through high school; and between domains, i.e., those stemming from different sources such as health care settings, work and school. Thus, data show that overt discrimination is significantly associated with a reduction of positive affect or sense of well-being. Subtle discrimination is not directly associated with a reduction in positive affect but is associated with depression (4).
Implications for measurement: The implication of these data for measurement is that discrimination should be clearly defined and measured in various situations and contexts, from individual situations to institutional settings. It should also be measured multiple times during development and its frequency of occurrence noted along with the multiple settings in which it occurs. However, to understand the impact of discrimination on physical and mental health outcomes, it is also important to measure how well a person copes. Successful coping will increase resilience and reduce, at least somewhat, the negative effects of discrimination. Therefore, it is also important to measure inter-generational experiences because parents pass their expectations with respect to discrimination and their methods of coping on to their children. Since means for successful coping vary across cultures, measurements must be done in a culturally sensitive manner. Emotion and cognition play key roles in coping strategies and should be included as part of this process.
Domains of Measurement
There are several ways of thinking about the domains of measurement from a theoretical perspective. One framework conceptualizes institutionalized vs. personally mediated racism / discrimination, and the extent to which these forms of discrimination have been internalized (5). In this context, institutionalized racism / discrimination is defined as differential access to the goods, services and opportunities of society based on race. Personally mediated racism is defined as prejudice and discrimination at an individual level based on assumptions about the motives, intentions and abilities of others according to race (5). When the discrimination is internalized, the members of the race or ethnic group which has been discriminated against, begin to believe the biases of others and come to believe it, feeling that they are somehow inferior because of their race. One definition of successful coping would be the ability of a child to withstand these assaults on his / her personal sense of self. Another way of conceptualizing race would be the perception of discrimination by the individual, the social distance between races and the racial climate created by relations between race / ethnic groups.
A number of domains and settings were enumerated where racism may occur: residential housing, the physical environment (e.g., neighborhood), schools, school clubs, day care, access to work and work environments, health care, research settings, media & entertainment, immigration, prisons and the criminal justice system, neighborhood surveillance (e.g, by the police), adequacy of nutrition, access to transportation, and internet sources of racism (now quite wide spread). It was emphasized that these various domains contribute to the overall cumulative burden of discrimination. Other contextual measures relate to distribution of local resources (are supermarkets available in the neighborhood or do they have to travel to get to one) and hazards (drugs, alcohol, weapons).
Social support systems of all kinds are important in helping to cope with racial discrimination. Family networks and faith based communities were identified as possible support systems in this regard. Self-efficacy, hope, optimism, resiliency and personal control were all seen as being influenced by the extent of an individual’s exposure to racism and discrimination.
Several participants asked whether the study of discrimination would be limited to race / ethnicity. They said discrimination was also important with respect to gender and sexuality (e.g., heterosexual vs. lesbian/gay/bisexual/transgender (LGBT)), and pointed out that this may be particularly important in instances where family structure in the racial / ethnic groups is confounded by discrimination based on either of these other issues. They are also important forms of discrimination in and of themselves, regardless of race.
Implications for Measurement
Conceptually, there are several ways of gathering data about discrimination in epidemiological studies (6). It can be inferred indirectly at the individual level; measured directly by self-report of discrimination in the individual; and thirdly, at a group level. The latter method involves investigating whether group-level measures of discrimination are associated with population rates of health outcomes.
Data bases: Data bases are one way of inferring exposure to racial discrimination without using self-report. Macro trends in social attitudes can be obtained through sources such as Gallup polls. Geocoding to census tracts can be utilized to obtain information on poverty percentages, median housing values, racial / ethnic composition, etc., from which an index of dissimilarity can be created as one measure of institutional discrimination (7). The Home Mortgage Disclosure Act (HMDA) data base that includes information on loan type, purpose, loan amount requested, applicant income, reason for denial, gender, race and co-applicant characteristics has revealed inequities in lending practices that are objective measures of racial climate and institutionalized racism in specific neighborhoods. Another method of providing context to individual measures of discrimination is the examination of the racial composition of school boards and teacher ratios for representational parity. The General Accounting Office (GAO), the National Academies and over 200 other organizations and individuals are also working together to develop a national indicators system with a broad range of information covering the economy, society, and the environment (8).
Individualized measures of racism: The group expressed consensus that it would be important for any instruments selected for the NCS to be composed of scales that have been tested and shown to have internal consistency, reliability, and / or been validated as predictors of specific health outcomes. This would enable results from the NCS to be compared to previously published literature and assure that if lack of association were found, it would not be attributable to unreliable measures.
Multiple measures were suggested that are related to the above domains. One such measure that measured discrimination in multiple domains was developed for the longitudinal Coronary Artery Risk Development in Young Adults Study (9,10) and has been used in many other studies (6). This questionnaire measures whether the participant has "ever experienced discrimination, been prevented from doing something or been hassled or made to feel inferior... because of their race or color" in any of seven situations: "at school, getting a job, at work, getting housing, getting medical care, on the street or in a public setting, and from the police or in the courts". This variable can be coded according to how many situations discrimination has been experienced. Its internal consistency is high, and additional psychometric properties pertaining to validity and reliability are currently being assessed in a sample of working class African Americans and Latinos. Also included are questions on how the participant responds to unfair treatment ("accept it as a fact of life" vs. "do something about it;" "speak to others vs. "keep to oneself"). The instrument is intended for use in epidemiologic studies and takes little time to administer.
A telephone administered perceived racism scale (TPRS) was developed from focus groups of African Americans who evaluated the content and face validity of the questions. Examples of questions include: "Because I am Black, I feel..;." "Whites often assume that blacks work in lower status jobs and therefore, treat them as such" (11). Factor analyses revealed five factors: passive emotions (feelings of powerlessness and hopelessness), active emotions (angry, frustrated, anxious, sad), passive behaviors (don’s speak up), internal active behaviors (praying) and external active behaviors (working harder to prove them wrong). Tests of internal consistency indicated good reliability for all scales. This instrument has the advantage of being designed for telephone interviews, which can be important in epidemiologic studies.
Another scale analyzing coping methods in response to discrimination has also demonstrated good internal consistency (12). This measure showed that in Korean immigrants in Toronto, active, problem-focused coping styles were more effective in reducing the impact of discrimination on depression than frequent use of passive, emotion-focused coping, which had a debilitating effect on mental health.
A question measuring race consciousness has also been pilot tested in the 2002 Behavioral Risk Factor Surveillance Survey in six states (13). This question asks how often an individual thinks about his or her race and shows large discrepancies between African American and white participants. Whites don’s think much about their race whereas African American and Hispanic respondents think about it often.
This is not a complete list of the measures discussed at the meeting. Others, along with their references, are being sent to the Program Office. One of these is the Minority Health Survey from the Commonwealth Survey of 2002, another is the MEUSS used by Larry Bobo, another is a measure of social exclusion (Elena Yu), and still another is a measure of "tokenism" by Pamela Jackson.
In the discussion of measurement, it was also suggested that institutionalized racism was important. Measurement can be made individually by questionnaires (such how one is treated in schools, health care settings, etc.), by examining differences in assets and resources by race, and by doing "audit" studies. An example of an audit study would be sending out the same resume to multiple businesses that have advertised for personnel, but varying the names on the resume to resemble distinct ethnic groups or genders. The purpose would be to see if there is any consistent variation in the ranking of the quality of the resume based on these factors. It was suggested that this may be an appropriate "in depth" study to do with the R01 grant mechanism once the study is established and all centers have been activated.
Measures in Children and the Life Span Perspective
With respect to measuring racism and discrimination in children, two important points were made: measurement in children requires different methods than in adults and although discrimination at all ages is detrimental, there are time points where a child may be more vulnerable to the negativity created by discrimination than at others.
Group discussion indicated that measurement in children cannot be totally separated from that in adults, due to the importance of parent/child interactions throughout childhood. Thus, measures of discrimination in the mothers of the NCS children’s cohort will be important both during and after pregnancy. They will be important during pregnancy as measures of stress and its resulting neuroendocrine consequences, as well as in the context of the expectancies that pregnant mothers have for their unborn children. It was noted that young, pregnant mothers in minority groups, especially in socioeconomically deprived areas, often fear for the survival of their unborn children and that a form of institutionalized racism begins in the clinics, where clinic personnel automatically assume that if a baby was unplanned, it is also unwanted. When mothers are asked whether the baby was planned, if they answer no, there is no follow-up question. Clinic staff automatically assume that it is also unwanted. Simply knowing whether a baby is being born into a welcoming situation or is unwanted might be important for children of all races.
However, measuring discrimination in the mother will also be important after birth, not only as a measure of stress, but as a reflection of the experiences and attitudes that the parent will be conveying to the child. It was noted that the resiliency of the child is to some extent dependent on the resiliency of the mother and other household or family members. Racial coping skills are critical competencies for African American (and probably other minorities) children to have (14). If the mother and other adults who care for the child are resilient and capable of teaching successful coping skills to the child, he or she will have a better chance of faring well despite discrimination.
With respect to domains of measurement in children, it was emphasized that the cumulative trajectory over time is extremely important. The domain of education is an example where discrimination may begin early, taking the form of subtle assumptions about children on the part of teachers, based on the child’s race / ethnicity or socioeconomic status. Whether the trajectory assumes the form of cumulative burden, depends on how often and in what forms discrimination occurs during the progression from elementary through middle school, high school and college.
Objective measures of inequities in education include demographics such as student / teacher ratios in schools, teacher attendance, school resources and physical conditions of schools, as well as standardized test scores and drop out rates. There is also a series of age appropriate measures for primary school children which has been developed to examine school and teacher climate, student racial coping, as well as a child’s self-efficacy and self-concept, that has been developed and validated in the Comer school intervention project (15). The Comer intervention model emphasizes positive, collaborative partnerships among teachers, administrators, families and students (16). The measures developed to assess coping and well-being include: what I think of school, dialogues about family and friends, a pictorial scale of perceived competence and social acceptance for young children, a self-perception profile for children, a culture-free self-esteem inventory, racial coping measures and a self-efficacy inventory. An important finding of this study is that these concepts are not stable over time but change as the children get older. Experience from this study indicates that effects of discrimination on self-esteem begin as early as 2 nd grade.
In younger children, the advantages of projective techniques were also emphasized. These techniques involve having children make up a story related to a picture, having the child respond to a hypothetical situation of another child in a story, or giving the child a camera and telling him or her to go outside of their residence and take a picture, then tell about the picture.
Measures of school climate determined from ratings of parents, teachers and students have also been developed and validated in middle school and high school populations (17), showing pervasive racial differences among school staff. Additional measures suggested by workshop participants as being relevant for children include: children’s aspirations and the way they change over time; the point at which a child first recognizes race; stereotype threat; homework stress; and white privilege. White privilege is defined as the ability of whites in America to ignore the issue of race because it privileges them in ways that are most often unrecognized. Things that white children take for granted, such as fair treatment in school, are not privileges granted to minority students. However, whites, not confronted with these issues themselves, are often not aware that they exist. Access to "cultural capital" - the extent to which children are given piano and ballet lesson, taken to the theater, and given horse back riding, tennis or golf lessons was also thought to be important. These cultural factors can further serve to separate children along racial lines.
Access to mental health services for children, access to medications (e.g., asthma medication for children who need inhalers is sometimes taken away from the child and required to be kept in the nurse’s office), stigmatization about being obese, what a child knows about sexuality (the group realized this may be difficult), and access to physical activity in schools were also considered important.
Community Partnerships
One area that was greatly emphasized in this workshop was the importance of partnering with communities. An observational study such as the NCS must establish an ongoing partnership with the involved communities from the very beginning. The participants and the community need to receive regular feedback concerning study findings and issues related to their communities and individual children (e.g., health related feedback). Other suggested partnerships included the Office of Civil Rights, the Department of Housing and the Department of Labor.
References
- Cain VS, Kington RS. Investigating the role of racial / ethnic bias on health outcomes. Am J Pub Health 2003;93:191-192.
- Williams DR, Neighbors HW, Jackson JS. Racial / ethnic discrimination and health: Findings from community studies. Am J Pub Health 2003;93:200-208.
- Measuring Racial Discrimination: Executive Summary. Panel on Methods for Assessing Discrimination, Committee on National Statistics, Division of Behavioral and Social Sciences and Education. The National Academies Press, Washington, D.C.
- Wu Z, Noh S, Kaspar V, Schimmele CM. Race, ethnicity, and depression in Canadian society. J Health Soc Behav 2003;44:426-441.
- Jones CP. Levels of racism: A theoretic framework and a gardener’s tale. Am J Pub Health 200;90:1212-1215.
- Krieger N. Discrimination and Health. In Berkman L & Kawachi I (eds.). Social Epidemiology. New York. Oxford University Press, 2000;36-75.
- Gee GC. A multilevel analysis of the relationship between institutional and individual racial discrimination and health status. Am J Public Health 2002;92:615-623.
- Forum on Key National Indicators: Assessing the Nation’s Position and Progress. 2003, GAO-03-672SP. In cooperation with The National Academies.
- Krieger N. Racial and gender discrimination; Risk factors for high blood pressure? Social Sci Med 1990;30:1273-1281.
- Krieger N. Racial and gender discrimination and blood pressure: The CARDIA study of young black and white adults. Am J Public Health 1996;86:1370-1378.
- Vines AI, McNeilly MD, Stevens J, Hertz-Picciotto I, Bohlig M, Baird DD. Development and reliability of a telephone-administered perceived racism scale (TPRS: A tool for Epidemiological use. Ethn Dis 2001;11:251-262.
- Noh S, Kaspar V Perceived discrimination and depression: moderating effects of coping, acculturation and ethnic support. Am J Public Health 2003;93:232-238.
- Jones C. Unpublished data.
- Garcia Coll C, Crnic K, Lambery G, Wasik BH, Jenkins R, Garcia HV, & McAdoo HP. An integrative model for the study of developmental competencies in minority children. Child Development 1996;67:1891-1914.
- Slaughter DeFoe D. Reference pending.
- Comer J. School Power: Implications of an intervention project. 1995. New York: Free Press.
- Aber MS. A report of primary findings of the school Climate Study for Champaign Community Unit School District 4.