Assessing Race, Ethnicity and Gender in Health., 978-0387324616
Assessing Race, Ethnicity and Gender in Health., 978-0387324616
Assessing Race, Ethnicity and Gender in Health., 978-0387324616
Visit the link below to download the full version of this book:
https://cheaptodownload.com/product/assessing-race-ethnicity-and-gender-in-healt
h-full-pdf-download/
Assessing Race, Ethnicity
and Gender in Health
Sana Loue
Case Western Reserve University
Cleveland, Ohio
Sana Loue
Case Western Reserve University
Cleveland, OH 44106–4945
USA
Cover illustration
C 2006 Superstock
9 8 7 6 5 4 3 2 1
springer.com
Acknowledgments
This book would not have been possible without the vision of my former editor,
Mariclaire Cloutier, and my current editor, Bill Tucker. Appreciation is due to
a number of individuals for their assistance with the research. Gary Edmunds,
Nancy Mendez, Ingrid Vargas, and Jenice Contreras deserve recognition for the
many hours that they spent tracking down references.
I owe special thanks to members of the communities with which I work, too
numerous to name individually, for their contributions to my education and to the
evolution of my understanding of identities.
v
List of Tables
vii
Contents
Prologue xi
PART I FOUNDATIONS
Index 153
ix
Prologue
American studies are five times more likely than European trials to report in pub-
lications the race or ethnicity of the study participants (Sheikh, Netuveli, Kai, and
Panesar, 2004). In a review of full-length articles appearing in three major pedi-
atric journals, for instance, it was found that over one-half of the published reports
contained data on participants’ race and/or ethnicity (Walsh and Ross, 2003).
Nearly 80% of hospitals in the United States collect data on race and ethnicity
(Runy, 2004). Despite the relatively high prevalence of collecting and reporting
race/ethnicity information, however, significant debate continues regarding the
wisdom of this practice. The following objections have been voiced with regard
to the collection and use of data relating to race and ethnicity.
r The categories once used are inadequate and categories change too frequently
to make the collection of the data worthwhile. As an example, during the period
from 1990 to 2003, hospitals in Rhode Island utilized three different classification
systems for the collection and recording of race and ethnicity data (Buechner,
2004).
r Categories that we construct may not be valid due to intermarriage. The most
recent census data indicates that over 2% of the American population, or more
than 7 million persons, now acknowledge a multiracial identity (Ahmann, 2005).
Children were more likely to be recorded as having a multiracial identity than
adults, which suggests an increase in interracial coupling.
r Data collection based on self-reports is not valid because individuals change
their self-identification depending upon the context in which they are asked to
so designate (Kaplan and Bennett, 2003).
r Individuals grouped into the same category demonstrate significant genetic di-
versity, so that the construction of the categories is questionable (Erickson, 2003;
Schultz, 2003). There are no gene variants that are present in all individuals of
one population and that are absent in the individuals of another population group
(Bonham, Warshauer-Baker, and Collins, 2005). For instance, Wilson and col-
leagues (2001) found in a study of drug-metabolizing enzymes and genotyping
in eight populations around the world that genotypes clustered into four groups,
xi
xii Prologue
but the four groups did not correspond to the populations from which they had
been drawn.
r Categorizations are too broad to have any definitive medical meaning (Schultz,
2003) and may obscure heterogeneity within groups (Kaufman, 1999; Williams,
2001). Gatrad and Sheikh (2000) have cautioned health care providers to re-
frain from making assumptions about patients’ willingness or unwillingness to
undergo particular screening tests or elective procedures on the basis of their
ethnicity.
r Socially determined categories cannot be applied to biological science (Schultz,
2003).
r Classifying individuals on the basis of socially constructed categories of race
and ethnicity serves to reinforce racial and ethnic divisions that already exist
(Azuonye, 1996; Bogue and Edwards, 1971; Fullilove, 1998; Stolley, 1999).
r The categories developed for race and ethnicity are often used to compare mi-
nority groups to the majority population and these comparisons often focus on
the negative aspects of the health and lives of minority group members (López,
2003).
r A focus on ethnic or racial groups may lead researchers to believe or encourage
them to disregard relevant social or cultural processes that are shared across
group boundaries (Garro, 2001).
Additional concerns have been voiced with regard to the assessment of accul-
turation level and immigration status, two constructs that are relevant to race and
ethnicity.
r Categories of immigration status change too frequently to be useful over time. As
an example, attempts to assess immigrants’ utilization of publicly funded health
care for specific services must be cognizant of the fluctuations in participants’
eligibility for benefits as a result of changes in their immigration status and/or
changes in the relevant legislation (Loue, Cooper, and Lloyd, 2005).
r Self-reports of immigration status will be inaccurate, if they are provided, due
to fears of deportation.
r Questions relating to immigration status will result in poor recruitment and
retention due to fears of the consequences that might ensue following disclosure.
Similar objections have been raised to the collection of data relating to sexual
orientation and sexual identity:
r Sexual orientation and/or sexual identity are fluid over a lifetime and assessment
at one point in time may be inaccurate.
r How an individual chooses to identify him- or herself may not be reflective of
his or her true orientation because of concern about the consequences of self-
disclosure and, accordingly, collection of this information is not valuable.
r Assessment of sexual orientation tends to compare homosexuals and heterosex-
uals, with the resulting inference that homosexuality is somehow “less than” or
“worse than” or undesirable.
Prologue xiii
Indeed, with respect to any or all of these variables, there is little consensus among
researchers as to how categories should be defined or who should be assigned to
them. And, in view of such vociferous displeasure with the collection and recording
of these variables in the context of health care and research, one must necessarily
question why it is done.
Hospitals have been found to collect these data in order to meet the requirements
of a law or regulation, to improve the quality of care, to ensure the availability of
interpreter services, to improve or maintain community relations, to assist in tar-
geting marketing efforts, and/or because it is perceived as beneficial (Runy, 2004).
Pediatric researchers have reported collecting and reporting race and ethnicity
data because it was required by the institutional review board of their institution,
the National Institutes of Health, and/or the peer-reviewed journal in which they
wished to publish; to conform to a tradition of reporting race and ethnicity data;
to better describe the study population; and/or because they believed that it was
relevant (Walsh and Ross, 2003).
Scholars have suggested other reasons underlying the necessity for the collection
of data relating to race and ethnicity (Mays, Ponce, Washington, and Cochran,
2003). These reasons are relevant, as well, to data regarding sexual orientation.
(1) To describe vital and health statistics. These data provide information that
can be utilized by public health programming and planning to develop programs
targeting specific health issues of concern in a manner that is appropriate to the
affected communities.
(2) To identify risk indicators for specific health outcomes. It has been argued
that ethnicity itself constitutes a risk factor for specific diseases, such as Tay-
Sachs, which occurs predominantly in individuals of Eastern European Jewish
descent (Greenidge, 2004). Alleles associated with sickle cell anemia are not evenly
distributed across racial/ethnic groups, but are found more frequently in African-
American populations (Collins, 2004). The risk factor profile for breast cancer
among African-American women has been found to differ from that of white
women, although the underlying mechanisms of these differences require further
investigation (Bernstein, Teal, Joslyn, and Wilson, 2003).
(3) To improve the delivery of health care services (Hasnaian-Wynia and Pierce,
2005). Nonwhite patients have been found to rate the quality of and their satisfac-
tion with their health care lower than do whites (Haviland, Morales, Reise, and
Hays, 2003). To some extent, this difference may be attributable to disparate treat-
ment by health care providers associated with differences in patient self-identified
or perceived race or ethnicity (Bach, Pham, Schrag, Tate, and Hargraves, 2004;
van Ryn and Burke, 2000).
(4) To identify markers of unmeasured biological differences. Researchers have
reported slower metabolism of some drugs in persons of Asian ethnicity, compared
to white and blacks (Meadows, 2003). It has been hypothesized that this difference
may be attributable, in part, to genetic factors that have not yet been identified.
(5) To identify proxy variables for unmeasured social factors. Race, ethnicity, sex,
and sexual orientation may serve as markers for other variables that we are unable
xiv Prologue
References
Ahmann, E. (2005).Tiger Woods is not the only “Cablinasian”: Multi-ethnicity and health
care. Pediatric Nursing 31(2): 125–129.
Azuonye, I.O. (1996). Guidelines will encourage the thinking that underpins racism in
medicine. British Medical Journal 313: 426.
Bach, P.B., Pham, H.H., Scrag, D., Tate, R.C., Hargraves, J.L. (2004). Primary care physi-
cians who treat blacks and whites. New England Journal of Medicine 351: 575–584.
Bernstein, L., Teal, C.R., Joslyn, S., Wilson, J. (2003). Ethnicity-related variation in breast
cancer risk factors. Cancer 97 (1 Suppl): 222–229.
Bogue, G., Edwards, G.F. (1971). How to get along without race in demographic analysis.
Social Biology 18: 387–396.
Bonham, V.L., Warshauer-Baker, E., Collins, F.S. (2005). Race and ethnicity in the genome
era: The complexity of the constructs. American Psychologist 60(1): 9–15.
Buechner, J.S. (2004). Hospitalizations by race and ethnicity, Rhode Island, 1990–2003.
Medicine and Health/Rhode Island 87(7): 220–221.
Collins, F.S. (2004). What we do and don’t know about ‘race’, ‘ethnicity’, genetics and
health at the dawn of the genome era. Nature Genetics Supplement 36(11): S13–S15.
Erickson, A.K. (2003). Ethnicity puts clinical trials to the test. Nature Medicine 9(8): 983.
Fullilove, M.T. (1998). Comment abandoning “race” as a variable in public health research:
An idea whose time has come. American Journal of Public Health 88: 1297–1298.
Garro, L.C. (2001). The remembered past in a culturally meaningful life: Remembering as
cultural, social, and cognitive processes. In C. Moore, H. Mathews (Eds.), The Psychology
of Cultural Experience (pp. 105–147). U.K.: Cambridge University Press.
Gatrad, A.R., Sheikh, A. (2000). Birth customs: Meaning and significance. In A. Sheikh,
A.R. Gatrad (Eds.), Caring for Muslim Patients. Oxford, UK: Radcliffe Medical Press
Ltd.
Prologue xv
Greenidge, K.C. (2005). Race, politics, ethnicity, and science. American Journal of Oph-
thalmology 139: 704–706.
Hasnaian-Wynia, R., Pierce, D. (2005). HRET Disparities Toolkit: A Toolkit for Collect-
ing Race, Ethnicity, and Primary Language Information from Patients. The Health Re-
search and Educational Trust. Available at http://www.hretdisparities.org. Last accessed
November, 2005.
Haviland, M.G., Morales, L.S., Reise, S.P., Hays, R.D. (2003). Do health care ratings differ
by race or ethnicity? Joint Commission Journal on Quality and Safety 29(3): 134–145.
Kaplan, J.B., Bennett, T. (2003). Use of race and ethnicity in biomedical publication. Journal
of the American Medical Association 289(20): 2709–2716.
López, S.R. (2003). Reflections on the Surgeon General’s Report on Mental Health, Culture,
Race, and Ethnicity. Culture, Medicine, and Psychiatry 27: 419–434.
Loue S., Cooper, M., Lloyd L.S. (2005). Welfare and immigration reform and use of prenatal
care among women of Mexican ethnicity in San Diego, California. Journal of Immigrant
Health 7(1): 37–44.
Mays, V.M., Ponce, N.A., Washington, D.L., Cochran, S.D. (2003). Classification of race
and ethnicity: Implications for public health. Annual Review of Public Health 24: 83–110.
Meadows, M. (2003). FDA issues guidance on race and ethnicity data. FDA Consumer
May–June: 36.
Runy, L.A. (2004). Collecting race & ethnicity data. Hospitals & Health Networks 78(8):
30.
Sheikh, A., Netuveli, G., Kai, J., Panesar, S.S. (2004). Comparison of reporting of ethnicity
in US and European randomised controlled trials. British Medical Journal 329 (7457):
87–88.
Schultz, J. (2003). FDA guidelines on race and ethnicity: Obstacle or remedy? Journal of
the National Cancer Institute 95(6): 425–426.
Stolley, P.D. (1999). Race in epidemiology. International Journal of Health Services 29:
905–909.
Van Ryn, M., Burke, J. (2000). The effect of patient race and socio-economic status on
physicians’ perceptions of patients. Social Science and Medicine 50: 813–828.
Wash, C., Ross, L.F. (2003). Whether and why pediatric researchers report race and ethnicity.
Archives of Pediatric and Adolescent Medicine 157: 671–675.
Wilson, J.F., Weale, M.E., Smith, A.C., Gratrix, F., Fletcher, B., Thomas, M.G., Bradman,
N., Goldstein, D.B. (2001). Population genetic structure of variable drug response. Nature
and Genetics 29(3): 265–269.
Part I
Foundations
1
Constructing Categories: Context
and Consequences
are human mental constructs . . . they are intellectual boundaries we put on the world in order
to help us apprehend it and live in an orderly way . . . [N]ature doesn’t have categories; people
do (Stone, 1988: 307).
Categories have been distinguished from groups and classes. Jenkins (2003) has
argued that the formation of groups and classes is rooted in the processes of internal
definition, by which individuals signal to others both inside and outside of their
group their self-definition of identity. These internal definitions and designations
may be critical, depending upon the goals of a particular study. As an example, it
may be important that a researcher studying social networks in the LGBT (lesbian,
gay, bisexual, and transgender) communities understand how individuals self-
classify because such distinctions establish their own position within a group and
signal their group membership to others; these distinctions may, for instance,
denote the degree of masculinity or femininity claimed, such as in the distinction
made between butch and femme lesbians or bear and twinkie gay men. (For a
discussion of some of these intra-group classifications, see Gibson and Meem,
2002; Stoller, 1991; Wright, 1997).
In contrast, categories are externally defined, although in actuality, the processes
of defining the “us” and the “them” is an interactive one: the categorization of
“them” assists in the definition of “us”, while the definition of “us” is a function
of the history of relationships with others (Hagendoorn, 1993; Jenkins, 2003). We
form these categories based on perceived commonalities that we believe somehow
distinguish those objects or persons that belong to a specified category and those
that do not (Jay, 1981).
3
4 1. Constructing Categories: Context and Consequences
An unauthentic word, one which is unable to transform reality, results when dichotomy
is imposed upon its constitutive elements . . . . Either dichotomy, by creating unauthentic
forms of existence, creates also unauthentic forms of thought, which reinforce the original
dichotomy.
We see, then, that how we construct our categories and labels frames our world,
and that frame further reinforces our use of those categories. And, although our
construction of categories varies and shifts over time, place, and purpose, we often
act as if the categories to which we assign certain attributes and the members to
whom we attribute these attributes are static and remain so. As an example, research
indicates that the designation of an individual’s race or ethnicity may vary over the
course of his or her life depending upon who is charged with the responsibility or
authority to make this designation: the individual’s parent at the time of his or her
birth, the individual him- or herself, the employer reporting on the composition of
the workforce, and the coroner at the time of death (Yanow, 2003). Additionally,
whether an individual self-identifies as a member of a particular group, whatever
its nature, may vary depending upon the social, historical, and political context in
which the designation is to be made.
The referent category often takes on a special significance in that it is isolated from
all else [and] is A and pure. Not-A is necessarily impure, a random catchall, to which
nothing is external except A and the principle of order that separates it from Not-A (Jay,
1981: 45).
As a result, members of those categories that are not-A, that are not part of the
referent group and are perceived as different, also may be perceived as being
deviant. This is best understood through an examination of the “social audience”
approach to deviance.
The social audience approach defines who and what are deviant as a function
of the viewers:
[W]hether or not an act is deviant depends on how others who are socially significant in
power and influence define the act. One could commit any act, but it is not deviant in its
social consequences if no elements of society react to it (Bell, 1971: 11).
[A]cts and actors violating the norms of society will be termed “rule-breaking behavior”
and “rule breakers,” while the terms “deviant behavior and “deviant” will be reserved for
acts and actors labeled as deviant by a social audience (Cullen and Cullen, 1978: 8).
6 1. Constructing Categories: Context and Consequences
Ben-Yehuda (1990: 36) has asserted that “who interprets whose behavior, why,
where, and when is very crucial . . . .” He offers two examples in support of his
view. The first is that of Joan of Arc, who was executed in 1431 after having been
convicted as a heretical deviant; she was later canonized by Pope Benedict XV.
The second example is that of the Nobel Laureate chemist Louis Pauling. Pauling
lobbied the United Nations during the late 1950s to end nuclear weapons testing.
Despite the widespread support that he received from other scientists, his stance
led to his interrogation by the U.S. Senate and a prohibition against his attendance
at various international scientific meetings. In 1962, he was awarded the Nobel
Peace Prize in recognition of his efforts to promote peace (Ben-Yehuda, 1990).
Depending upon the construction of the categories to be used in research and
the portrayal of members of those groups, the construction of categories may
also result in the marginalization of their members. Tucker (1990: 7) has defined
marginalization as
the complex and disputatious process by means of which certain people and ideas are privi-
leged over others at any given time . . . [and] the process by which, through shifts in position,
any given group can be ignored, trivialized, rendered invisible and unheard, perceived as
inconsequential, de-authorized, “other,” or threatening, while others are valorized.
In this way, the adherence to a belief that humans must necessarily be biolog-
ically of one sex or another reinforces and is then further reinforced by our con-
struction of maleness and femaleness. We construct categories and then engage in
activities to eliminate what is perceived to fall outside of these categories. We point
to these categories that we have constructed as if they had naturally arisen and then
proceed to interpret reality in the context of these “naturally-arising” categories.
A further example of the consequences of categorization of a group by those
holding power and authority is provided by the previous psychiatric designation of
homosexuality per se as a mental disorder. The second edition of the Diagnostic
and Statistical Manual (1968), previously utilized by the American Psychiatric
Association as a guide to the diagnosis of mental disorders, referred to homosexu-
ality as a mental illness. The inclusion of homosexuality in this nosology removed
it from the litany of behaviors that had previously fallen within the jurisdiction
of the church as the guardian of morality and, to some extent, the legal system,
which had viewed homosexuality as criminal in nature and therefore deserving of
punishment (Bayer, 1981).
Accordingly, the psychiatric profession viewed itself as the protector “of de-
viants who had suffered at the hands of society and the more traditional forces of
social control” (Bayer, 1981:11). However, its increasing assumption of responsi-
bility for the control of behaviors previously viewed as immoral and/or criminal,
such as substance abuse and sexual deviance, was viewed by others as an attempt
to widen the scope of its authority and foster a therapeutic state (Kittrie, 1972).
Further, in the context of the civil rights movement of the 1960s and 1970s, ho-
mosexuals began to see themselves as an oppressed minority that had been and
continued to be oppressed by social institutions and ideological standards and
began to campaign actively and vociferously to have homosexuality deleted as a
class of mental illness (Bayer, 1981: 12–13). Bayer has explained the significance
of this action:
To dismiss the significance of the debate over whether homosexuality ought to be included
in the APA’s nosological classification . . . is to miss the enormous importance it carried for
American society, psychiatry, and the homosexual community. By investing the dispute
with great meaning, the participants had themselves transformed it from a verbal duel into
a crucial, albeit symbolic, conflict. The gay community understood quite well the social
consequences of being labeled and defined by others, no matter how benign the posture of
those making the classification. A central feature of its struggle for legitimation therefore
entailed a challenge to psychiatry’s authority and power to classify homosexuality as a
disorder.
We are told, from the time that we first recognize our homosexual feelings, that our love for
other human beings is sick, childish and subject to “cure.” We are told that we are emotional
cripples forever condemned to an emotional status below that of the “whole” people who
run the world. The result of this in many cases is to contribute to a self-image that often
lowers the sights we set for ourselves in life, and many of us asked ourselves, “How could
8 1. Constructing Categories: Context and Consequences
anybody love me?” or “How can I love somebody who must be just as sick as I am?”
(Quoted in Bayer, 1981: 119).
Many times, these attributions and the associated stigmatization are premised
on stereotypes. Those who are stigmatized
are pejoratively regarded by the broader society and [are] devalued, shunned or otherwise
lessened in their life chances and in access to the humanizing benefit of free and unfettered
social intercourse (Alonzo and Reynolds, 1995: 304).
particularly susceptible to certain genetic disorders. Plans to conduct such research should
be sensitive to such consideration, to the need to maintain confidentiality during and after
the study, and for the need to publish the resulting data in a manner that is respectful of the
interests of all concerned, or in certain circumstances not to publish them. The ethical review
committee should ensure that the interests of all concerned are given due consideration; often
it will be advisable to have individual consent supplemented by community consultation.
Summary
It is critical that researchers consider the larger ethical, social, and political im-
plications of their categorization of populations and research participants. Efforts
must be made to reduce the likelihood that these categorizations, which may be
formulated with the best of intentions to improve health, will result instead in the
reinforcement of stereotypes and the marginalization and/or devaluation of specific
groups.
References
Alonzo, A.A., Reynolds, N.R. (1995). Stigma, HIV and AIDS: An exploration and elabo-
ration of a stigma trajectory. Social Science and Medicine, 41(3), 303–315.
American Psychiatric Association. (1968). Diagnostic and Statistical Manual, DSM-II.
Washington, D.C.: Author.
10 1. Constructing Categories: Context and Consequences
Bayer, R. (1981). Homosexuality and American Psychiatry: The Politics of Diagnosis. New
York: Basic Books, Inc.
Bell, R.B. (1971). Social Deviance. Homewood, Illinois: Dorsey.
Ben-Yehuda, N. (1990). The Politics and Morality of Deviance: Moral Panics, Drug Abuse,
Deviant Science, and Diverse Stigmatization. Albany, New York: State University of
New York Press.
Comstock, G.D. (1991). Violence against Lesbians and Gay Men. New York: Columbia
University Press.
Cullen, F.T., Cullen, J.B. (1978). Toward a Paradigm of Labeling Theory. Lincoln, Nebraska:
University of Nebraska.
Erikson, K.T. (1962). Notes on the sociology of deviance. Social Problems, 9, 307–314.
Epstein, J. (1990). Either/or—neither/both: Sexual ambiguity and the ideology of gender.
Genders 7: 99–142.
Freire, P. (1970). Pedagogy of the Oppressed. (Trans. M.B. Ramos). New York: Continuum.
Gibson, M., Meem, D.T. (Eds.). (2002). Femme/Butch: New Considerations of the Way We
Want to Go. Binghamton, New York: Harrington Park Press.
Healey, D. (2001). Homosexual Desire in Revolutionary Russia: The Regulation of Sexual
and Gender Dissent. Chicago, Illinois: University of Chicago Press.
Jenkins, R. (2003). Rethinking ethnicity: Identity, categorization, and power. In J. Stone,
R. Dennis (Eds.), Race and Ethnicity: Comparative and Theoretical Approaches (pp. 58–
71). Malden, Massachusetts: Blackwell Publishing.
Kitsuse, J.L. (1962). Societal reaction to deviant behavior: Problems of theory and method.
Social Problems 9: 247–256.
Kittrie, N. (1972). The Right to Be Different. Baltimore, Maryland: Johns Hopkins Univer-
sity Press.
Luchetta, Y. (1999). Relationships between homophobia, HIV/AIDS stigma, and HIV/AIDS
knowledge. In L. Pardie, T. Luchetta (Eds.), The Construction of Attitudes Towards
Lesbians and Gay Men (pp. 1–18). Binghamton, New York: Harrington Park Press.
Mason, A., Palmer, A. (1996). Queer Bashing: A National Survey of Hate Crimes against
Lesbians and Gay Men. London: Stonewall.
Stoller, R.J. (1991). Pain and Passion: A Psychoanalyst Explores the World of S&M. New
York: Plenum Press.
Stone, D.A. (1988). Policy Paradox and Political Reason. Boston, Massachusetts: Little,
Brown.
Stone, J. (2003). Max Weber on race, ethnicity, and nationalism. In J. Stone, R. Dennis (Eds.),
Race and Ethnicity: Comparative and Theoretical Approaches (pp. 28–42). Malden,
Massachusetts: Blackwell Publishing.
Szasz, T. (1970). The Manufacture of Madness. New York: Delta Books.
Tucker, M. (1990). Director’s forward. In R. Ferguson, M. Gever, T.T. Minh-ha, C. West
(Eds.), Out There: Marginalization and Contemporary Cultures (pp. 7–8). New York:
The New Museum of Contemporary Art/MIT Press.
Wright, L. (1997). The Bear Book: Readings in the History and Evolution of a Gay Male
Subculture. Binghamton, New York: Harrington Park Press.
Yanow, D. (2003). Constructing “Race” and “Ethnicity” in America: Category-Making in
Public Policy and Administration. Armonk, New York: M.E. Sharpe.
2
Methodological Considerations
The selection of instruments for use in a particular study or the construction of new
instruments to assess ethnicity, race, gender, sex, and related constructs requires
a consideration of various methodological issues, in addition to the context in
which these tools are to be developed and the ethical implications of the categories
once they have been developed. Issues to be considered prior to deciding which
of existing instruments to use or whether and how to construct a new instrument
include the focus of the research question, the format to be used to collect the data,
and how the population of interest is to be sampled. The selection of an instrument
for use, or the development of a new instrument, also requires attention to the
instrument’s validity, reliability, and the possibility of misclassification associated
with its use (McDowell and Newell, 1996). A basic understanding of these issues
is important in order to better evaluate the literature that exists with regard to the
constructs that are the focus of this text. This is not, however, a comprehensive
discussion of these issues, which can be the focus of entire books themselves,
and the reader is urged to consult the sources listed at the end of this chapter for
additional guidance.
11
12 2. Methodological Considerations
to assess the individual’s identity over time and/or how an individual is perceived
by others in terms of his or her race, ethnicity, sexual orientation, etc.
This is an important consideration even in instances in which the researcher has
decided to rely on pre-formulated categories for the classification of the research
participants. For instance, the manner in which the federal government has defined
various ethnicities and races has changed over time. (See chapter 3.) A study that
spans time periods that use different classification systems may find that the choices
provided to respondents at the initiation of the study period may no longer be in
use towards the end, and researchers may have to reconcile responses to the newer
categories.
(2) The focus of the research question. The concepts of race, ethnicity, sex, gen-
der, sexual orientation and related concepts are multidimensional. As an example,
depending upon the focus of the research, a determination of ethnicity may require
an assessment of the ethnicity of an individual’s parents and grandparents in addi-
tion to a consideration of the origin of the individual research participant. A sexual
history that focuses on the number and sex of one’s sexual partners may be suffi-
cient to answer a question focusing on the sex of one’s sexual partners, but it may
not be adequate to determine an individual’s sexual orientation, which is a function
of emotional attraction, physical attraction, sexual fantasies, self-definition, and
opportunity.
These characteristics are also subject to identification not only by the individ-
ual who may be a participant in the research, but also by the observer as well.
For instance, an individual may self-identify his or her race (suspending, for the
moment, a discussion as to whether race exists), but an individual’s race is also
subject to the perception of the observer who, on the basis of criteria that he or
she as somehow developed, will make a judgment regarding the individual’s race.
Consequently, it is important to consider in framing the research question whether
participants’ self-identity as to race, ethnicity, sexual orientation, gender, etc. is
important or whether it is the perception of specified observers that is critical. For
instance, how an individual identifies him- or herself with respect to race may not
be as important as how the individual is viewed by others, if the focus of the study
is an exploration of the effects of political marginalization.
selected from among those that are eligible for inclusion in the study (Kelsey,
Thomson, and Evans, 1986). For example, if an investigator wished to know the
proportion of households of a particular ethnic group had health insurance, the
sample would consist of a portion of those households where a designated member
was of that ethnicity.
The sampling frame refers to the list of the population from which the sample
will be drawn. In some cases, this is unknown and unknowable. For instance, a
study focusing on the experience of homophobia by gay and lesbian individuals
would have a difficult, if not impossible, task to construct a sampling frame,
since it would not be possible to know of every individual who self-identified
as gay or lesbian, since many may not wish to acknowledge their orientation
publicly.
In instances in which it would be difficult to locate and recruit study partici-
pants due to the nature of the research, investigators often rely on a convenience
sample comprised of volunteers. This strategy, however, can introduce bias into
the selection process. Snowball sampling, in which already-recruited participants
identify other individuals as potential participants, permits investigators to more
easily locate and recruit “hidden” individuals, but may also introduce selection
bias because the individuals recruited through respondents are more likely to be
like the respondents.
Probability sampling is advisable when it is feasible, in order to reduce the
possibility of selection bias. There are four basic designs for probability sampling:
simple random sampling, systematic sampling, stratified sampling, and cluster
sampling.
Simple random sampling requires knowledge of the complete sampling frame
in advance (Kelsey, Thompson, and Evans, 1986). This strategy means that each
sampling unit in the population has an equal chance of being selected for partici-
pation. This method does not require advance knowledge of the population itself,
but may be very inefficient.
Systematic sampling refers to the selection of sampling unit, such as individuals
or households, at regularly spaced intervals within the sampling frame, such as
every third household. This method has several advantages in that it does not
require advance knowledge of the sampling frame, as it can be constructed as the
process progresses and is generally relatively simple to implement.
Stratified sampling requires the division of the population into strata and a
sample is selected from each such strata. This process is significantly more complex
than the other strategies, but offers increased precision and may facilitate the
inclusion of specific groups of persons.
Like stratified sampling, cluster sampling divides the sample into groups, such
as clusters of homes. A sample is then taken of these clusters for inclusion in
the study or, alternatively, a subsample of these sample clusters is utilized. As
an example, an investigator wishing to study the prevalence of violence in public
housing projects might divide all such projects into clusters by geographic area and
then take a sample of these clusters. The households within these clusters could
then be queried about the violence in the public housing projects.