HOW TO REVIEW THE LITERATURE AND CONDUCT ETHICAL STUDIESshowed 10 percent of the U.S. military to be a per-
son who is gay or lesbian and the military provided
no support for the banning of gays from the mili-
tary. In 2005, the White House threatened the head
of the little-known Bureau of Justice Statistics with
dismissal and eventually demoted him for releasing
law enforcement data on racial profiling. The gov-
ernment agency produces dozens of reports each
year on crime patterns, drug use, police tactics, and
prison populations. The data documented clear dis-
parities in how racial groups were treated once they
were stopped by the police. Political supervisors
demanded deleting references to the disparities
from reports. The data were based on interviews
with 80,000 people in 2002. It showed that White,
Black, and Hispanic drivers nationwide were
stopped by the police that year at about the same
rate, roughly 9 percent. However, once the police
had made a stop, what happened next differed
depending on driver’s race and ethnicity.^34
In sponsored research, we can negotiate con-
ditions for releasing findings prior to beginningthe
study and sign a contract to that effect. It may be
unwise to conduct the study without such a guar-
antee, although competing researchers who have
fewer ethical scruples may do so. Alternatively, we
can accept the sponsor’s criticism and hostility and
release the findings over the sponsor’s objections.
Most researchers prefer the first choice because the
second one may scare away future sponsors.
Social researchers sometimes restrict or delay
the release of findings to protect the identity of
informants, to maintain access to a research site, to
hold on to their jobs, or to protect the personal safety
of themselves or of family members.^35 This is a
less disturbing type of censorship because it is not
imposed by an outside power. It is done by someone
who is close to the research and who is knowledgeable
about possible consequences. Researchers shoulder
the ultimate responsibility for their research. Often,
they can draw on many different resources, but they
face many competing pressures as well. (See Expan-
sion Box 6, Common Types of Misuse in Evaluation
Research.)
Concealing the True Sponsor.Is it ethical to keep
the identity of a sponsor secret? For example, anabortion clinic funds a study on the attitudes of
religious groups opposed to abortion. We must bal-
ance the ethical value of making the sponsor’s iden-
tity public to subjects and releasing results against
the sponsor’s desire for confidentiality and the
likelihood of reduced cooperation from subjects. If
the results are published, there is a clear overriding
ethical mandate to reveal the true sponsor. There is
less agreement on the ethical issue of revealing the
true sponsor to subjects. Presser and colleagues
(1992) found that the answers given by survey
respondents may depend on its sponsor. If a respon-
dent believes a survey is conducted by a newspaper
that has taken a strong position on an issue, the
respondent is less likely to contradict the news-
paper’s public stand on the issue. This is less a
problem if the respondent believes the survey spon-
sor is a neutral academic organization. It is ethical
to inform the subjects of the sponsor unless one has
a good methodological reason for not doing so.EXPANSION BOX 6
Common Types of Misuse in
Evaluation ResearchAsking “wrong” research questions (e.g., asking sum-
mative yes/no questions when formulative questions
are most appropriate or asking questions that
exclude major stakeholders)
Requesting an evaluation study after a decision on a
program has been made, using the study only as a
way to delay or justify the decision already made
Demanding the use of a research design/data col-
lection technique that is inappropriate for the pro-
gram evaluation task
Interfering with the research design or data collection
process to ensure that it produces desired results
Continuing a program when the evaluation results
unambiguously show it to be ineffective or ending a
program when the results unambiguously show it to
be highly effective
Suppressing/deleting positive results to eliminate/
reduce a program, or suppressing/deleting negative
results to continue/expand a program
Source:Adapted from Stevens and Dial (1994), who also pro-
vide examples of misuse.