Research Using Intentional Deception
Ethical Issues Revisited
Diana Baumrind
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
II I Ill
Illllll Illll
University of California, Berkeley
Illllll Illllll
ABSTRACT: Ethical issues concerning the use of
intentional deception in research with human participants are revisited 10 years after the publication of
the 1973 APA guidelines on the conduct of research
with human subjects. Intentional deception is defined.
The present status of guidelines concerning intentional
deception and the incidence and extremity of deception being used are reviewed, leading to the conclusion
that the former has not decreased the latter. My
position proscribing intentionally deceptive research
is grounded in rule-utilitarian metaethics by contrast
to act-utilitarian and deontological metaethics; three
ethical rules proscribing intentional deception in the
research setting are presented. The costs to subjects,
the profession, and society are reviewed, and arguments are brought to bear against the scientific
benefits claimed for deceptive instructions, In this
context, the scientific problems with social psychological experimentation are discussed. Finally, certain
recommendations for research strategies in lieu of
deception paradigms, and for appropriate debriefing,
are offered.
In a series of articles (Baumrind, 1964, 1971, 1972,
1975a, 1975b, 1978, I979), beginning with a critique
of the Milgram (1964) paradigm, I argue that the
use of intentional deception in the research setting
is unethical, imprudent, and unwarranted scientifically. In response to my latest article (Baumrind,
1979), Baron (1981) offered "'an openly optimistic
rejoinder." He claimed that deception research is
necessary to accomplish beneficial scientific ends
and that as a result of the guidelines researchers
now use informed consent and thorough debriefing.
Two respondents took exception to Baron's optimistic
rejoinder. Dresser (1981) pointed out that in distorting a study's true purpose the investigator necessarily grounds participants" willingness to cooperate
on misinformed consent. Goldstein (1981) argued
that participants' rights to autonomy, dignity, and
privacy are necessarily violated by deceptive research
practices and rejected Baron's assurance that experimenters were now sensitized to ethical issues. Surveys of the major social psychological journals suggest
that Goldstein is correct and Baron's optimism is
unwarranted. Ten years after publication of the
February 1985 • American Psychologist
Cofs~ti~t 1985 by the American Ps~hok~ical Association, Inc. ~3-066X/85/$00.75
Vol. 40, No. 2, 165-174
II
I Illllllll
Ethical Principles in the Conduct of Research With
Human Participants (American Psychological Association, 1973), I was invited to revisit these issues
in the context of a symposium on ethics of deception
research delivered to the International Society for
Research on Aggression.
By intentional deception I mean withholding
information in order to obtain participation that the
participant might otherwise decline, using deceptive
instructions and confederate manipulations in laboratory research, and employing concealment and
staged manipulations in field settings. Because perfect
communication between human beings is impossible
to achieve, there will always be some degree of
misunderstanding in the contract between researcher
and subject. Full disclosure of everything that could
possibly affect a given subject's decision to participate
is not possible, and therefore cannot be ethically
required. Provided that participants agree to postponement of full disclosure of the purposes of the
research, absence of full disclosure does not constitute intentional deception. The investigator whose
purpose is "to take the person unaware by trickery"
or to "'cause the person to believe the false" in order
to minimize ambiguity about causal inference is
intentionally deceiving subject-participants to further
the investigator's scientific ends or career goals.
Investigators continue to use intentional deception and to justify its use. Epistemological superiority
is accorded to deceptive methods as a means of
controlling the demand characteristics of the setting
by assuring that all subjects believe the situation to
be realistic and perceive it in the same way. If
intentional deception does not accomplish this objective, its epistemological superiority is doubtful,
which in turn casts doubt on the benefits of intentional deception and thus on the justification for its
use. Although the examples I offer are drawn largely
from aggression research, the justification for deceptive practices arises from the research paradigm that
guides experimental social psychology as a whole. It
is necessary, therefore, to examine that paradigm.
Present
Status
of Deception
Research
If deceit is used to obtain consent, by definition it
cannot be informed. Deceptive instructions logically
165
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
contradict the informed consent provision contained ognized hearing deficit on the development of parain all federal and professional ethical guidelines. Yet noia; Zimbardo et al. then misinformed subjects as
these guidelines do permit each o f the provisions to the purpose of the experiment and what experiguaranteeing informed consent to be waived provided ences they would undergo and recruited subjects
that some or all considerations such as the following with the promise that they would continue to play
a part in his research, a promise that was not kept;
pertain:
and (d) Marshall and Zimbardo (1979) used multiple
(a) The research objective is of great importance and high-magnitude deceptions to study the affective
cannot be achieved without the use of deception; (b) on consequences of inadequately explained physiological
being fully informed later (Principle E), participants are
expected to find the procedures reasonable and to suffer arousal; they misinformed subjects about the purpose
no loss of confidence in the integrity of the investigator of the experiment; they manipulated physiological
or of others involved; (c) research participants are allowed arousal via injection o f epinephrine or placebo after
t o withdraw from the study at any time (Principle F), and telling subjects that they would receive a vitamin
are free to withdraw their data when the concealment or injection; and they misled subjects into thinking
misrepresentation is revealed (Principle H); and (e) inves- their responses were not being monitored when they
tigators take full responsibility for detecting and removing were. They then postponed debriefing for six weeks
stressful aftereffects of the experience (Principle I). (Amer- until all subjects had been tested. The IRB permitted
ican Psychological Association, 1982, p. 41)
all of the above manipulations, balking only at a
No strategic guidelines are included to assure that planned "angry" condition on the basis that it was
these considerations pertain. Institutional review unethical to induce anger in unsuspecting subjects.
boards (IRBs), as well as investigators, are at liberty
to set their own.
Rule-Utilitarian Objections
Neither the incidence nor the magnitude of de- to Deception Research
ception reported in social psychological research
appears to have decreased since 1973.)Thus, the The cost-benefit analysis arising from the metaethical
APA guidelines appear to serve an expressive rather justification called act-utilitarianism (Frankena,
than a deterrent function. McNamara and Woods 1963) is used to justify these exceptions. By contrast,
(1977) reported a rise to 57% in a survey covering deontological or rule-utilitarian metaethical positions
the years 1971 to 1974. In the most recent study do not lend themselves to justification of the use of
(Smith & Richardson, 19837, approximately half of deceptive research practices. From both these more
the 464 psychology undergraduates surveyed reported stringent metaethical positions, if informed consent
that the experiment in which they had participated is a right of participants and intentional deception
used deception. Both figures exceed Seeman's (1969) a necessary violation of that right, then intentional
figures of 18% in 1948 and 37% in 1963. The deception is to be avoided in the research setting.
The thesis of act-utilitarianism is that a particmaximum magnitude of reported deception has not
ular
act is right if, and only if, no other act the
decreased as four exemplars published since the
agent
could perform at the time would have, on the
1973 APA guidelines will illustrate: (a) Milgram's
agent's
evidence, better consequences. From an actparadigm was duplicated by Shanab and Yahya
(1977) with children as young as six; graphic reports utilitarian stance, if in the opinion of the investigator,
of the children's reactions document trembling, lip the requirements of the research demand that the
biting, and nervous laughter; (b) White (1979) used participants be kept unaware that they are being
a typical aggression paradigm in which a confederate studied or deception must be used to create a
angered real subjects by evaluating them personally psychological reality in order to permit valid inferin a highly negative and insulting manner; subjects ence, then failure to obtain informed consent, conwere then asked to administer shocks to the confed- cealment, and deception could be justified. The
erate, after being given a false cover story as to the decision would be made by the investigator applying
purpose of the experiment; (c) Zimbardo, Andersen, a cost-benefit calculus to the specific situation. Actand Kabat (1981) induced partial deafness through utilitarianism, by comparison with rule-utilitarianposthypnotic suggestion to study the effect of unrec- ism, falls short as a metaethical system of justification
because (a) it fails to consider the substantive rights
of the minority, (b) it fails to take long-range costs
While writing this article the author was supported by grants into account, and (c) it is subjective and not generfrom the W. T. Grant Foundation, the John D. and Catherine T. alizable. It takes little to convince a researcher or a
MacArthur Foundation, and a Research Scientist Award from the review board of his or her peers that the long-range
National Institute of Mental Health.
Requests for reprints should be sent to Diana Baumrind, benefits of a clever bit of deceptive manipulation
Institute of Human Development, 1203 Edward Chace Tolman outweigh the short-range costs to participants of
Hall, Universityof California, Berkeley,California94720.
being deceived. The long-range costs to subjects and
166
February 1985 • American Psychologist
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
society are unknown and therefore are easy for
investigators and review boards to dismiss. It is
difficult to imagine more extreme instances of deception than those provided by Zimbardo's experiments, and yet both were approved by the Stanford
IRB, subsequent to 1973, just as Milgram's experiments had been reviewed favorably at Yale prior to
1973.
Deontological moralists (e.g., Wallwork, 1975)
claim that the basic judgments of obligation are
perceived as being given intuitively without recourse
to consideration of what serves as the common good.
For deontologists such as Kant, the principle of
justice or truth or the value of life stands by itself
without regard to any balance of good over evil for
self, society, or the universe. By contrast, I hold,
with Waddington (1960), that the function of ethical
beliefs is to mediate human evolution so that there
can be no principles, including preservation of life,
distributive justice, and trustworthiness that are
absolutely inviolable.
I ground my judgment that intentional deception in the research setting is morally wrong not in
act-utilitarianism (which is too relativistic) or in a
deontological categorical imperative (which is too
dogmatic), but rather in rule-utilitarianism, the view
that an act is right if, and only if, it would be as
beneficial to the common good in a particular social
context to have a moral code permitting that act as
to operate under a rule that would prohibit that act.
In Western society, as a result of innate or learned
behaviors associated with optimum survival for our
community, most of us elevate rules of social living
that maximize the opportunity for self-determination
so that we may hold each other accountable for our
actions. Also, we share strong aversions to certain
types of actions. We are averse to hurting others,
killing others, and telling lies, and so we feel that
we must justify these acts if we intentionally commit
them. Those of us who are rule-utilitarians grant
that even actions to which we have strong moral
aversions are justifiable in certain contexts. Thus,
retaliative aggression may be justifiable provided
that it is proportionate to the grievance; killing may
be justifiable if our victim is an enemy or a murderer
or a fetus under three months; telling lies may be
justifiable if intended to benefit the recipient and
not the liar; hurting others may be justifiable if we
are dentists or surgeons. However, by justifying
morally aversive acts we legitimate them, and this,
too, has social consequences, because harm inflicted
self-righteously may appear to demand no reparation
and is not self-correcting.
Why is intentional deception in the context of
the research endeavor so wrong from a rule-utilitarian
perspective? At least three ethical rules generally
accepted in Western society proscribe deceitful reFebruary 1985 • American Psychologist
search practices: (a) the right of self-determination
within the law, which translates in the research
setting to the right of informed consent; (b) the
obligation of a fiduciary (in this case, the researcher)
to protect the welfare of the beneficiary (in this case,
the subject); and (c) the obligation, particularly of a
fiduciary, to be trustworthy in order to provide
sufficient social stability to facilitate self-determined
agentic behavior. Consistent with a rule-utilitarian
position, I propose to ground these rules teleologicaUy
by arguing that their adoption benefits modern Anglo-American society more than contradictory rules,
thus explaining their general acceptance.
The principle of informed consent is a manifestation of the basic right granted each individual in
Anglo-American political philosophy and tradition
to be self-determining and let alone so long as the
individual is not interfering with the rights of other
individuals or the public. According to Sir William
Blackstone (t 765-1769/1941), individuals surrender
to society many rights and privileges that they would
be free to exercise in a state of nature, in exchange
for benefits that each receives as a member of
society. Each citizen retains, however, certain fights
and privileges that the public may not abrogate
without the citizen's consent. Thus, subjects have
the right to judge for themselves whether being lied
to or learning something painful about themselves
constitutes psychological harm for them. A violation
of an individual's right of informed consent is a
breach of the social contract and thus legitimates
retaliative lawlessness because only in a rule-following
environment may we be held fully accountable for
the consequences of our actions. Therefore, social
scientists must exercise their right to seek knowledge
within the constraints imposed by the right to
informed consent of those persons from whom they
would obtain that knowledge.
Further, the basic rules of fiduciary law apply
to the researcher-subject and the teacher-student
relationships (Holder, 1982). A fiduciary obligation
pertains when a person, called the fiduciary, is
dealing with another under circumstances involving
the placing of a special confidence. The overriding
duty of the fiduciary is loyalty and trustworthiness
by contrast to the principle of "caveat emptor,"
which may apply to some sales relations. If challenged, the burden of proof is on the fiduciary to
show loyalty and trustworthiness. It is illogical for
an investigator to justify the use of deceit by appealing to the special quality of the investigatorsubject relationship when it is just that special
quality that enables the investigator to recruit participants and establishes the fiduciary obligations of
the investigator to be trustworthy in relations with
them. If the rule that justifies scientific experimentation is "You shall know the truth, and the truth
167
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
shall set you free," then that rule applies also in the
conduct o f science.
Finally, violation o f trust is generally held to
be immoral, and even more so in a fiduciary relationship, especially when the subject is also a student
for whom the experimenter serves as a model. If the
student is beguiled, his or her trust has been misplaced. If the student is not beguiled, then deceptive
instructions do not undermine trust. Psychology
undergraduates in heavily experimental departments
expect rigged lotteries, deceptive instructions, and
the use of confederates. Those who adopt a gameset are not likely to be disillusioned because they do
not assume that the experimenter is trustworthy.
Under those conditions, subjects may be beguiled,
but they are not betrayed. The wrong done as well
as the harm done is trivial. But to the extent that
students routinely suspend belief, then experimental
control has not been achieved by deceptive instructions, and there are no benefits to weigh against the
costs. The costs include encouraging students to lie
in the interests of science and career advancement.
Even from the permissive stance of an actutilitarian cost-benefit analysis, it is the responsibility
of investigators who wish to use deceptive practices
that fail to conform to the informed consent provision o f the ethical guidelines to demonstrate first,
that the social and scientific benefits o f their proposed
research objectives are indeed sufficiently significant
to offset the costs to participants, the profession,
and society; second, that the research paradigm will
effectively accomplish those objectives; and third,
that those objectives cannot be accomplished equally
well by using nondeceptive research paradigms. We
proceed now to reexamine the costs and benefits o f
deception research.
I was harmed in an area of my thinking which was
central to my personal development at that time. Many
of us who volunteered for the experiment were hoping to
learn something about ourselves that would help us to
gauge our own strengths and weaknesses, and formulate
rules for living that took them into account. When,
instead, I learned that I did not have any trustworthy way
of knowing myself--or anything else--and hence could
have no confidence in any lifestyle I formed on the basis
of my knowledge, I was not only disappointed, but felt
that I had somehow been cheated into learning, not what
I needed to learn, but something which stymied my very
efforts to learn. I told literally no one about it for eight
years because of a vague feeling of shame over having let
myself be tricked and duped. It was only when I realized
that I was not peculiar but had, on the contrary, had a
typical experience that I first recounted it publicly. (Baumrind, 1978, pp. 22-23)
Anecdotal evidence such as this has been challenged as hearsay. A number o f studies have been
undertaken to establish whether subjects are harmed
by deception experiments. The results are equivocal.
However, most of these studies rely on self-report
rather than behavioral evidence. About 80% of subjects, when asked, say that they were glad to have
participated in the experiment. This is used illogically
to establish that subjects suffered no harm. Thus,
Milgram (1974, p. 195) justified his shocking procedure by citing results o f a follow-up questionnaire
in which 84% of subjects said they were glad to have
participated in the experiment. However, as Patten
(1977) pointed out, it is logically inconsistent for
Milgram to use the self-reported judgments of overly
acquiescent ("destructively obedient") subjects to
establish the ethical propriety o f his experiments.
Similarly, Marshall and Zimbardo's (1979) subjects
were chosen for their hypnotic suggestibility and
would be expected to defer compliantly to the
Costs of Deception Research
experimenter's expertise. After all, if self-reports
The costs o f deceptive research practices accrue to could be regarded as accurate measures o f the
the participants, to the scientific enterprise, and to impact o f experimental conditions, we could dispense
entirely with experimental manipulation and behavsociety.
ioral measures, substituting instead vivid descriptions
Harm Done to the Subject
o f environmental stimuli to which subjects would
A brief excerpt from an autobiographical account be instructed to report how they would act.
Self-report questionnaires used to assess partico f a former secretary who typed my earlier articles
illustrates the subtle but serious harm that can be ipants' reactions are tacked on as an afterthought
done to, subjects by undermining their trust in their and generally lack psychometric sophistication. Subown judgment and in fiduciaries as well as the jects' self-reported gladness to be stressed and dereluctance many have to admit, even to themselves, ceived may be explained by a variety of psychological
mechanisms in addition to deferential compliance
that they have been duped.
discussed above. These mechanisms include reducThis experiment [involving deceptive feedback about quality of performance relative to peers] confirmed my con- tion o f cognitive dissonance, identification with the
aggressor, and masochistic obedience. It takes wellviction that standards were completely arbitrary . . .
because the devastating blow was struck by a psychologist, trained clinical interviewers to uncover true feelings
whose competence to judge behavior I had never doubted of anger, shame, or altered self-image in participants
before. . . . It is not a matter of "belief" but of fact that who believe that what they say should conform with
I found the experience devastating.
their image of a "good subject." Ring, Wallston, and
168
February 1985 • American Psychologist
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Corey (1970), in their follow-up interview exploring
subjective reactions to a Milgram-type obedience
experiment, reported that many subjects stated that
they were experiencing difficulty in trusting adult
authorities. In a recently reported study of the effects
of debriefing (Smith & Richardson, 1983), about
20% of 464 introductory psychology undergraduates
reported experiencing harm. In the harm group,
61% had participated in a deception experiment as
compared to 38% in the no-harm group. Students
who had participated in deception experiments
tended to perceive psychologists as less trustworthy
than did nondeceived participants. Even subjects
who deny other harmful effects do report decreased
trust in social scientists following deception research.
For example, citing instances of experimental deception:
FiUenbaum (1966) found that deception led to increased
suspiciousness (even though subjects tended not to act on
their suspicions), and Keisner (197I) found that deceived
and debriefed subjects were "less inclined to trust experimenters to tell the truth" (p. 7). Other authors (Silverman,
Shulman, & Wiesenthal, 1970; Fine & Lindskold, 1971)
have noted that deception decreases compliance with
demand characteristics and increases negativisticbehavior.
(In Wahl, 1972, p. 12)
Decreased trust in fiduciaries then is a generally
acknowledged cost of deception itself. Even if we
choose to accept self-report data as veridical, 20%
of subjects report such harm and the proportion is
highest for deception research. From an ethical and
legal perspective, harm is done to each individual.
The harm the minority of subjects report they have
suffered is not nullified by the majority of subjects
who claim to have escaped unscathed, any more
than the harm done victims of drunk drivers can be
excused by the disproportionate number of pedestrians with sufficient alacrity to avoid being run
over by them.
From a rule-utilitarian perspective, the procedural issue concerns where the locus of control
should rightly reside. The generally accepted principle
of respect for self-determination dictates that the
locus of control should reside with each participant.
T h e subject, like the investigator, retains the right
to decide whether the likely benefits to self and
society outweigh the likely costs to self and society.
The investigator is not privileged to weigh the costs
to the subjects against the benefits to society. The
principle of informed consent allows the subject to
decide how to dispose of his or her person. The
subject acting as sovereign agent may freely agree to
incur risk, inconvenience, or pain. But a subject
whose consent has been obtained by deceitful and
fraudulent means has become an object for the
investigator to manipulate. A subject can only regain
February 1985 • American Psychologist
sovereignty by claiming to have been a subject all
along and not an object. Not surprisingly, subjects
tend to affirm their agency by denying that they
have allowed themselves to be treated as objects,
and when queried by an experimenter, most will say
that they were glad to have been subjects.
Harm Done to the Profession
The harm done by deception researchers accrues to
the profession and to the larger society as well as to
the individual. The scientific costs of deception in
research are considerable. These costs include (a)
exhausting the pool of naive subjects, (b) jeopardizing
community support for the research enterprise, and
(c) undermining the commitment to truth of the
researchers themselves.
The power of the scientific community is conferred by the larger community. Social support for
behavioral science research is jeopardized when investigators promote parochial values that conflict
with more universal principles of moral judgment
and moral conduct. The use of the pursuit of truth
to justify deceit risks the probable effect of undermining confidence in the scientific enterprise and in
the credibility of those who engage in it. As a result
of widespread use of deception, psychologists are
suspected of being tricksters. Suspicious subjects
may respond by role-playing the part they think the
investigator expects, doing what they think the investigator wants them to do (Orne, 1962), or pretending to be naive. The practice of deceiving participants and of justifying such deception undermines
the investigators' own integrity and commitment to
truth. Short-term gains are traded for the cumulative
costs of long-term deterioration of investigators'
ethical sensibilities and integrity and damage to their
credibility.
Harm Done to Society
The moral norm of reciprocity proscribing deceitful
social relations both acknowledges and places a
positive value on the fact that the elements of social
reality are reciprocally determined. The inherent
cost of behaving deceitfully in the research setting
is to undermine trust in expert authorities. If conduct
in the laboratory or natural setting cannot be isolated
from conduct in daily life, the implications are farreaching. In a popular article entitled "Snoopology,"
John Jung (1975) discussed some probable effects
of experimentation in real-life situations with persons
who did not know they were serving as experimental
subjects. These included increased self-consciousness
in public places, broadening of the aura of mistrust
and suspicion that pervades daily life, inconveniencing and irritating persons by contrived situations,
and desensitizing individuals to the needs of others
by "boy-who-cried-wolf" effects so that unusual
169
public events are suspected of being part of a
research project.
Truth telling and promise keeping serve the
function in social relations that physical laws do in
the natural world; these practices promote order and
regularity in social relations, without which intentional actions would be very nearly impossible. By
acting in accord with agreed-upon rules, keeping
promises, acting honorably, and following the rules
of the game, human beings construct for themselves
a coherent consistent environment in which purposive behavior becomes possible.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Benefits of Deception
Even if a simple cost-benefit calculus consistent with
act-utilitarianism is adopted and we agree to weigh
the costs to subjects against the benefits to humankind, we must still inquire as to what these benefits
might be. Consideration of the benefits of a proposed
investigation within the context of a cost-benefit
ethical analysis requires more stringent standards
for what constitutes scientifically and socially valid
research than in a purely empirical context. I will
argue that the scientific and social benefits of deception research cannot be established with sufficient
certitude to tip the scale in favor of procedures that
wrong subjects. The deception paradigms employed
by experimental social psychologists do not and
cannot deliver the reduction in ambiguity that could
justify what would otherwise be regarded as ethically
unacceptable research practices. Deceptive practices
do not succeed in accomplishing the scientific objectives that are used to justify such deception, any
better than methods that do not require deception.
If the phenomenon being studied is socially important
it can be studied in natural or clinical contexts that
do not require laboratory manipulation to produce.
If laboratory controls are required to create a counterfactual condition that exists nowhere, then in
order to claim benefits, it must first be shown that
the counterfactual conditions are possible to create
in situ and that the common good would be enhanced
by doing so.
The claim is made that deceptive manipulations
are required to create a psychological reality under
experimental conditions that permit valid inference.
This claim is based on two assumptions: (a) deceptive
instructions create a uniform psychological reality;
and (b) causal inference in the social sciences can
be achieved with a high level of certitude. But
neither of these assumptions has gone unchallenged
by critics of experimental social psychological
methods.
Deception does not create a uniform psychological reality when subjects in an experiment differ
in their level of naivet~ or in their responses to the
possibility of experimental manipulation. It is now
170
common knowledge among many kinds of prospective subjects that deception is employed routinely in
social psychological experiments. The tendency of
some subjects to assign idiosyncratic meaning rather
than to buy the experimenter's cover story, even
when it is true, defeats the purpose of deceptive
instructions, which is to control subject set. There
is evidence that investigators untrained in phenomenological assessment methodology will fail to detect
subject suspiciousness. Page (1973) has shown that
asking subjects fewer than four questions will classify
only 5% of subjects as suspicious, whereas extended
questionnaires will yield about 40% suspicious subjects, and the behaviors of suspicious subjects in the
experimental situation are generally found to differ
from those who are not. Referring to laboratory
research, Seeman (1969) concluded, "In view of the
frequency with which deception is used in research
we may soon be reaching a point where we no longer
have naive subjects, but only naive experimenters"
(p. 1026).
The traditional experimental social psychologist
justifies deception research on the logical positivist
presupposition that laboratory observations could
provide unassailable knowledge if only we were able
to produce a uniform psychological reality and do
away with error variance. The objective of the
traditional social psychology experiment is to enable
the experimenter to infer unambiguously the existence and direction of causal relations by ruling out
alternative causal explanations. Controls requiring
deceptive instructions are introduced with the implicit expectation that their use can provide such
unassailable knowledge. But the claim that observations can provide value-free, objective knowledge
has been challenged by philosophers and scientists
at least since Heisenberg's (1958) principle was
enunciated. From the perspective of their critics,
experimental controls distort by controlling the phenomena the investigator is attempting to explain.
The meaning subjects assign to a situation depends
upon the characteristics of that situation, as well as
upon subject characteristics. A psychological mechanism observed to operate under one set of experimental conditions often fails to replicate under a
somewhat different set of experimental conditions
because the meaning persons give a situation is
contextual and purposive and dependent upon factors
that the experimenter may not even be aware of,
such as the strangeness of the situation from the
perspective of the subject. Typically the conjunction
of events in a laboratory situation is atypical, and
the effect of constraining the options a subject has
is itself a factor that distorts the responses given and
the behavior observed. Whereas laboratory methods
construct situations and contexts for persons and
then assess how they respond to these extrinsicially
February 1985 • American Psychologist
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
constructed situations, persons in their natural settings typically construct or select their own social
worlds among the options available. Bronfenbrenner
(1977) called for an ecological perspective in developmental and social research precisely because he
disputed the possibility that subjects could assign
the same meaning to their behavior in the natural
setting as they did in the highly artificial laboratory
setting.
Thus, it can be argued that laboratory conditions
create the very ambiguity they are intended to
dispel. For example, Gardner (1978) could not
replicate Glass and Singer's (1972) findings of negative aftereffects of noise when subjects knew that
they had the option of discontinuing participation.
Assurance Of freedom to withdraw removed the
effect of the noise stressor. If Glass and Singer
intended to study the negative aftereffects of noise
per se, their experimental manipulation was inappropriate. If, on the other hand, they intended to
study the effects of inescapable noise, their experimental situation was highly relevant, but clearly
violated participants' right to withdraw. Similarly,
the experimental condition created by Milgram in
his studies of "destructive obedience" exemplifies a
highly ambiguous control that seriously compromises
the generalizability of his findings. Because subjects
were paid for their participation and recruited on
that basis, obedience in Milgram's setting for some
subjects might have reflected a sense of fair play and
employee loyalty rather than, as for other subjects,
shocking obedience. Moreover, Milgram's experimental directives were incongruous and bizarre, thus
confusing and distressing the subject. Furthermore,
the experimenter's orders were legitimized by the
laboratory setting, thus permitting subjects to resolve
their sense of incongruity by trusting the good will
of the investigator toward both subject and confederate. Far from illuminating real life, as he claimed,
Milgram in fact appeared to have constructed a set
of conditions so internally inconsistent that they
could not occur in real life. His application of his
results to destructive obedience in military settings
or Nazi Germany (Milgram, 1974) is metaphoric
rather than scientific.
Not only was the situation artificial in Milgram's
experiment, but all the necessary data were provided
by the graduate student confederates who demonstrated conclusively by obeying Milgram's instructions to inflict suffering upon the subjects that
normal, well-intentioned people will hurt others who
are innocent. The confederates justified their actions
on the same bases as the subjects justified theirs,
that they were inflicting no real harm. Because
subjects' motives could not unambiguously be called
destructive obedience and the behavior of his graduate student confederates could, Milgram's deceitful
February 1985 ° American Psychologist
manipulation was neither necessary nor could it
permit valid inference to the real-life situations to
which Milgram generalized his results.
Defenders of laboratory manipulations have
attempted to rebut the criticism that laboratory
experiments lack external validity and therefore do
not produce knowledge of benefit to society that
could justify misinforming subjects. Berkowitz and
Donnerstein (1982) argued that laboratory experiments are oriented mainly toward testing causal
hypotheses concerning mediational processes and
are not carried out to determine the probability that
a certain event will occur in a particular population.
They claimed that, in theoretically oriented investigations, the specific manipulations and measures are
merely arbitrary operational definitions of general
theoretical constructs and the subject sample is
merely an arbitrary group from the general universe
of all humans to which the hypothesis is assumed
to apply.
We have now come to our central thesis: The meaning the
subjects assign to the situation they are in and the behavior
they are carrying out plays a greater part in determining
the generalizabilityof an experiment's outcome than does
the sample's demographic representativenessor the setting's
surface realism. (Berkowitz & Donnerstein, 1982, p. 249)
But if the specific operations were interchangeable
as Berkowitz and Donnerstein claimed, the context
and subject populations could be altered without
changing the results. However, failure to replicate
social psychological findings when probed by a critical investigator is more the rule than the exception
in social psychology, and the failure to replicate can
seldom be attributed unambiguously to a controlled
change in experimental conditions. Results do not
survive even minor changes in the experimental
conditions, such as notifying subjects that they may
withdraw from an experiment. Therefore, Berkowitz
and Donnerstein were incorrect in claiming that the
specific operations typically employed in social psychological experiments are interchangeable. In the
event that the variables that are untied and independently manipulated in the laboratory setting are
necessarily or typically confounded in the natural
setting, conditions in the laboratory cannot or will
not be replicable. Psychological processes do not
occur in a psychosocial vacuum. When the task,
variables, and setting can have no real-world counterparts, the processes dissected in the laboratory
also cannot operate in the real world. In that case,
deceptive research practices cannot be justified by
their benefits to science and society.
Furthermore, deceptive practices cannot be justified unless they result in findings that are controversial because the benefit to society of noncontroversial, that is, trivial, findings is minimal. Berkowitz
171
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
and Donnerstein markedly attenuated the importance
of causal hypotheses by claiming that experimenters
ask only, c a n "alterations in Variable X lead to
changes in Variable Y?" This is generally a trivial
question because we almost always know prior to
conducting the experiment that the answer is yes.
Generally, the phenomenon has already been shown
to occur in real life. For example, did Zimbardo
and his colleagues (1981) really need to induce
partial deafness through posthypnotic suggestion to
confirm what is a generally acknowledged clinical
observation that unexplained deafness in older people
induces suspiciousness?
The mechanistic model of development implied
by experimental social psychological procedures is
not really applicable to social psychological phenomena. If the Cartesian, mechanistic world view is
insufficient to explain the physical world, then it is
certainly not adequate to explain biosocial and
psychosocial phenomena. If physical reality is subject
to indeterminacy introduced by the observer, as
Heisenberg (1958) assured us it is, how much more
true is this of human behavior? The level of complexity of social phenomena and the implausibility
of treating human beings as interchangeable or even
as identical with themselves over time sharply limit
the level of certitude that can accompany any empirical generaliTation in the sociobehavioral sciences.
A little ingenuity may well yield fruitful alternatives to deception (see Geller, 1982, for a systematic
review of such alternatives). If a phenomenon is
socially significant it can frequently be observed in
situ, making experimental manipulation unnecessary,
Alternatively, experimenters could act as subjects in
their own experiments and employ introspection.
Aggression can certainly be studied in situ. Aggression researchers can study acts intended to harm
others in such naturalistic situations as organized
sports, which vary along relevant parameters such
as rules and normative expectations. Investigators
interested in studying how people justify intentionally
causing others to suffer in real life have the option
of introspective examination of their own behavior
as participants in the research process. Thus, in lieu
of the familiar teacher-learner aggression paradigm
(Berkowitz & Geen, 1966) in which a confederate
makes a series of preplanned errors on a wordassociation task and subjects deliver shocks to the
confederate, investigators and their confederates
could introspect. It turns out that subjects believe
that they are benefiting the learner and therefore are
not behaving aggressively in the sense of intending
to inflict harm (Kane, Joseph, & Tedeschi, 1976).
Experimenters and their confederates argue just as
research subjects do that their motives in inflicting
emotionally painful discomfort are altruistic or justified by role expectations, even though an objective
172
observer might regard the behavior of confederates
and subjects as equally aggressive. For example,
Milgram justified his deceptive procedure by suggesting that many subjects were grateful for the
insight into their own destructively obedient tendencies that the experiment and debriefing afforded.
Investigators could also study retaliative aggression
without using deception by introducing certain nondeceptive experimental conditions in which they
acted as real, rather than confederate, subjects. Immediately after the usual perfunctory debriefing in
deception research, subjects would be instructed to
demonstrate behaviorally how they feel about their
participation by delivering 0 to 20 mild but genuine
electric shocks to the forearm of the experimenter
who designed the research. In addition to providing
data on subjects' response to deception and some
data on retaliative aggression, this aversive reinforcement coda might have the added advantage of
rendering superfluous any need for extrinsically
imposed codes mandating ethical practices in the
conduct of research with human participants.
The suggestion that psychologists serve as their
own subjects and introspect was made before me by
the investigators whose work Marshall and Zimbardo
(1979) failed to replicate:
In these days of ethical guidelines and human subjects
committees, this may very well be the end of the matter,
for it is unlikely that anyone will do experiments such as
ours or Marshall and Zimbardo's for quite a while, if ever
again. On the particular issue at stake, however, this is
probably of little moment, for this is one issue on which
the readers can serve as their own subjects. If they will do
a thorough introspective job after convincing a physician
to inject them with .5 cc of a 1:1000 solution of epinephrine, they can decide which of us is right. (Schachter &
Singer, 1979, p. 995)
Debriefing
Effective debriefing does not nullify the wrong done
participants by deceiving them and may not even
repair their damaged self-image or ability to trust
adult authorities. Subjects did, after all, commit acts
that they believed at the time could be harmful to
others, and they were in fact, entrapped into committing those acts by an authority whom they had
reason to trust. And if the participants (subjects and
confederates) are students, they have, in fact, been
provided with a model of behavior in which scientific
ends are used to justify deceitful means.
However, if an investigator does elect to use
deception, he or she must include an effective debriefing procedure in order to reduce the long-range
costs of deception and offer partial reparation to
subjects. Sieber (1983) offered carefully considered
recommendations for debriefing when deception is
used. She argued convincingly that deceptive debriefFebruary 1985 • American Psychologist
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
ing is especially u n e t h i c a l a n d t h a t debriefing in • of deception. IRB: A Review of Human Subjects Research, •(6),
1-4.
d e c e p t i o n research s h o u l d b e u n d e r t a k e n o n l y b y a
Berkowitz, L., & Donnerstein, E. (1982). External validity is more
skilled a n d s y m p a t h e t i c professional, o r it m a y d o
than skin deep: Some answers to criticisms of laboratory
m o r e h a r m t h a n good. F o r e x a m p l e , Mills (1976)
experiments. American Psychologist, 37, 245-257.
p r e s e n t e d in detail a debriefing s c e n a r i o t h a t he Berkowitz, L., & Geen, R. G. (1966). Film violence and the cue
properties of available targets. Journal of Personality and Social
developed over 20 years o f debriefing, a n d t h a t
Psychology, 3, 525-530.
provides the p a r t i c i p a n t with a n e d u c a t i o n a l expeBlaekstone, W. (1941). Commentaries on the taws of England.
rience as well as a t r u t h f u l a c c o u n t o f the experiWashington, DC: The Washington Law Book Co. (Original
m e n t ' s a c t u a l nature. T h e e x p e r i m e n t is e x p l a i n e d
work published 1765-1769)
very g r a d u a l l y a n d every p o i n t reviewed until the Bronfenbrenner, U. (1977). Toward an experimental ecology of
human development. American Psychologist, 32, 513-531.
subject understands. Subjects a r e given t i m e to reDresser, R. S. (1981). Deception research and the HHS final
organize their p e r c e p t i o n s o f t h e e x p e r i m e n t a n d
regulations. IRB: A Review of Human Subjects Research, 3(4),
their responses t o it, f r o m possible h u m i l i a t i o n a n d
3-4.
d i s c o m f o r t to self-acceptance and, it is to be hoped, Fillenbaum, S. (1966). Prior deception and subsequent experimental
performance: The "faithful" subject. Journal of Personality and
s y m p a t h e t i c u n d e r s t a n d i n g o f the researcher's perSocial Psychology, 4, 532-537.
spective. Subjects a r e offered a g e n u i n e o p p o r t u n i t y
Fine, R. H., & Lindskold, S. (1971). Subject's experimental
to w i t h d r a w their d a t a after having received a full
history and subject-based artifact. Proceedings of the Annual
e x p l a n a t i o n o f the p u r p o s e s o f t h e e x p e r i m e n t .
Convention of the American Psychological Association, 6, 289290.
Moreover, b y a d d i n g to the investigators" e m o t i o n a l
a n d fiscal costs, p a i n s t a k i n g a n d effective debriefing Frankena, W. (1963). Ethics. Englewood Cliffs, NJ: Prentice-Hall.
Gardner, G. T. (1978). Effects of federal human subjects regulations
p r o c e d u r e s i n t r o d u c e a noncoercive b u t persuasive
on data obtained in environmental stressor research. Journal
d e t e r r e n t to investigators w h o are c o n t e m p l a t i n g
of Personality and Social Psychology, 36, 628-634.
deception research.
Geller, D. M. (1982). Alternatives to deception: Why, what, and
how? In J. E. Sieber (Ed.), The ethics of social research: Surveys
and experiments (pp. 40-55). New York: Springer~Veflag.
Glass, D. C., & Singer, J. E. (1972). Urban stress: Experiments
~NC'ES
on noise and social stressors. New York: Academic Press.
American PsychologicalAssociation, Committee for the Protection Goldstein, R. (1981). On deceptive rejoinders about deceptive
research: A reply to Baron. IRB: A Review of Human Subjects
of Human Participants in Research. (1973). Ethical principles
Research. 3(8), 5-6.
in the conduct of research with human participants. Washington,
Heisenberg, W. K. (1958). Physics and philosophy. New York:
DC: Author.
Harper & Row.
American PsychologicalAssociation, Committee for the Protection
of Human Participants in Research. (1982). Ethical principles Holder, A. R. (1982). Do researchers and subjects have a fiduciary
relationship? IRB: A Review of Human Subjects Research, 4(1),
in the conduct of research with human participants (2nd ed.).
6-7.
Washington, DC: Author.
Baron, R. A. (1981). The "Costs of deception" revisited: An Jung, J. (1975). Snoopology. Human Behavior, 4(10), 56-59.
openly optimistic rejoinder. IRB: A Review of Human Subjects Kane, T. R., Joseph, J. M., & Tedeschi, J. T. (1976). Person
perception and the Berkowitz paradigm for the study of a~res,
Research, 3(1), 8-10.
sion. Journal of Personality and Social Psychology, 33, 663Baumrind, D. (1964). Some thoughts on ethics of research: After
673.
reading Milgram's "Behavioral study of obedience" American
Keisner, R. (197 t). Debriefing and responsiveness to overt experiPsychologist, 19, 421-423.
menter expectancy cues. Unpublished manuscript, Long Island
Baumrind, D. (1971). Principles of ethical conduct in the treatment
University, Long Island, IVY.
of subjects: Reaction to the draft report of the Committee on
Ethical Standards in Psychological Research. American Psy- Marshall, G. D., & Zimbardo, P. G. (1979). Affectiveconsequences
of inadequately explained physiological arousal. Journal of
chologist, 26, 887-896.
Personality and Social Psychology, 37, 970-988.
Baumrind, D. (1972). Reactions to the May 1972 draft report of
the Ad Hoe Committee on Ethical Standards in Psychological McNamara, J. R., & Woods, K. M. (1977). Ethical considerations
in psychological research: A comparative review. Behavior
Research. American Psychologist, 27, 1082-1086.
Therapy, 8, 703-708.
Baumrind, D. (1975a). It neither is nor ought to be: A reply to
Wallwork. In E. C. Kennedy (Ed.), Human rights and psycho- Milgram, S. (1964). Issues in the study of obedience: A reply to
Baumrind. American Psychologist, 19, 848-852.
logical research: A debate on psychology and ethics (pp. 83Milgram, S. (1974). Obedience to authority. New York: Harper & •
102). New York: Thomas Y. Croweli.
Row.
Baumrind, D. (1975b). Metaethical and normative considerations
governing the treatment of human subjects in the behavioral Mills, J. (1976). A procedure for explaining experiments involving
deception. Personality and Social Psychology Bulletin, 2, 3-13.
sciences. In E. (2. Kennedy (Ed.), Human rights and psychological
research: A debate on psychology and ethics (pp. 37-68). New McNamara, J. R., & Woods, K. M. (1977). Ethical considerations
in psychological research: A comparative review. Beh~ior
York: Thomas Y. Crowell.
Therapy, & 703-708.
Baumrind, D. (1978). Nature and definition of informed consent
in research involving deception. In The Belmont Report: Ethical Milgram, S. (1964). Issues in the study of obedience: a reply to
Baumrind. American Psychologist, 19, 848-852.
principles and guidelines for the protection of human subjects
of research (DHEW Publication No. (OS) 78-0014, 23-1-23- Milgram, S. (1974). Obedience to authority, New York: Harper
and Row.
71). Washington, DC: The National Commission for the Protection of Human Subjects of Biomedical and Behavioral Re- Orne, M. T. (1962). On the social psychology of the psychological
experiment: With particular reference to demand characteristics
search.
and their implications. American Psychologist, 17, 776-783.
Baumrind, D. (1979). IRBs and social science research: The costs
F e b r u a r y 1985 • A m e r i c a n Psychologist
173
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Page, M. M. (1973). On detecting demand awareness by postexperimental questionnaire. Journal of Social Psychology, 91,
305-323.
Patten, S. C. (1977). Milgram's shocking experiments. Philosophy,
52, 425-440.
Ring, K., Wallston, K., & Corey, M. (1970). Mode of debriefing
as a factor affecting subjective reaction to a Milgram-type
obedience experiment: An ethical inquiry. Representative Research in Social Psychology, 1, 67-88.
Schaehter, S., & Singer J. (1979). Comments on the Maslach and
Marshall-Zimbardo experiments. American Psychologist, 37,
989-995.
Seeman, J. (1969). Deception in psychological research. American
Psychologist, 24, 1025-1028.
Shanab, M. E., & Yahya, K. A. (1977). A behavioral study of
obedience in children. Journal of Personality and Social Psychology, 35, 530-536.
Sieber, J. E. (1983). Deception in social research III: The nature
and limits of debriefing. IRB: A Review of Human Subjects
Research, 5(3), 1-4.
Silverman, I., Shulman, A. D., & Wiesenthal, D. L. (1970). Effects
of deceiving and debriefing psychological subjects on performance
174
in later experiments. Journal of Personality and Social Psychology, 14, 203-212.
Smith, S. S., & Richardson, D. (1983). Amelioration of deception
and harm in psychological research: The important role of
debriefing. Journal of Personality and Social Psychology, 44,
1075-1082.
Waddington, C. H. (1960). The ethical animal. Chicago: University
of Chicago Press.
Wahl, J. M. (1972, April). The utility of deception: An empirical
analysis. In Symposium on Ethical Issues in the Experimental
Manipulation of Human Beings, Western Psychological Association, Portland, Oregon.
Wallwork, E. (1975). In defense of substantive rights: A reply to
Baumrind. In E. C. Kennedy (Ed.), Human rights andpsychological research (pp. 103-125). New York: Thomas Y. Crowell.
White, L. A. (1979). Erotica and aggression: The influence of
sexual arousal, positive affect, and negative affect on aggressive
behavior. Journal of Personidity and Social Psychology, 37,
591-601.
Zimbardo, P. G., Andersen, S. M., & Kabat, L. G. (1981, June).
Induced hearing deficit generates experimental paranoia. Science,
212, 1529-1531.
F e b r u a r y 1985 • A m e r i c a n Psychologist
534987
research-article2014
JIVXXX10.1177/0886260514534987Journal of Interpersonal ViolenceCheit
Article
Research Ethics and Case
Studies in Psychology:
A Commentary on Taus
v. Loftus
Journal of Interpersonal Violence
2014, Vol. 29(18) 3290–3307
© The Author(s) 2014
Reprints and permissions:
sagepub.com/journalsPermissions.nav
DOI: 10.1177/0886260514534987
jiv.sagepub.com
Ross E. Cheit1
Abstract
Loftus and Guyer have been criticized for the methods they employed in
investigating an anonymous case study published by Corwin and Olafson.
This article examines the ethical dimensions of their investigation. Loftus
and Guyer have offered three defenses for their actions. All three of those
defenses lack merit. Their investigation did not constitute oral history
because it failed to comport with the basic requirements of that practice.
Their investigation did not constitute ethical journalism because of the
unjustified use of anonymous sources and the clear violation of basic fairness.
Their investigation did not constitute justified medical research because of
a failure to analyze or weigh the harms against the benefits. Their methods
also violated ethical principles for psychologists, including the rule against
activities that could reasonably be expected to impair the psychologist’s
objectivity. This case demonstrates that there is no ethical way to investigate
a clinical case, without the patient’s approval, that is both comprehensive
enough to provide strong scholarship and yet respectful enough of privacy
and medical confidentiality to honor important professional norms.
Keywords
Taus v. Loftus, case studies, ethics, confidentiality, objectivity
1Brown
University, Providence, RI, USA
Corresponding Author:
Ross E. Cheit, A. Alfred Taubman Center for Public Policy and American Institutions, Brown
University, Campus Box 1977, Providence, RI 02912, USA.
Email: rc@brown.edu
Cheit
3291
Nicole Taus feels violated. She was the subject of a published case study
involving memory and child abuse, and she agreed to allow a videotape of
two interviews of her—one conducted in 1984 when she was 6 years and the
other in 1995 when she was 17 years—to be shown to educational audiences
without any identifying features. Corwin and Olafson (1997) presented the
case in a professional journal, where five others were invited to provide commentary. Identified in print only as Jane Doe, Taus’s case became part of a
lively debate over recovered memory, largely because of those two interviews. The following year, however, Taus found out that a private investigator was making inquiries about her in her hometown. She later learned that
the investigator had been hired by a prominent psychology professor,
Elizabeth Loftus, who also befriended Nicole Taus’s estranged mother—the
woman who had been found to have physically and sexually abused Nicole
as a child. Loftus eventually wrote a two-part article in a popular magazine
with Melvin Guyer, a psychology professor at the University of Michigan,
which was titled “Who Abused Jane Doe?” and argued that the mother’s side
of the story called Corwin and Olafson’s interpretation of the case into doubt.
They asserted that Jane Doe probably had not been abused at all, and if she
was, it was probably by her now-deceased father. Loftus and Guyer (2002a)
did not use Taus’s actual name, but it is not difficult to understand out why
she experienced both the article and Loftus and Guyer’s investigative methods as invasive, harmful, and unfair.
The complexity of this case poses major challenges to anyone seeking to
analyze it. One of those challenges is disentangling ethical concerns from
regulatory and constitutional ones. All three exist in this case, but because
there have been formal proceedings in the regulatory and legal contexts,
those two frameworks have come to define the case. The regulatory issues
involve the relevance and application of federal rules about human subjects
research, which are overseen by institutional review boards (IRBs). IRBs
might see themselves as guardians of ethics; their existence is justified in
ethical terms. But the justification for IRBs is hotly contested and some have
argued that they are unconstitutional infringements on free speech
(Hamburger, 2005; Tierney & Corwin, 2007). There are also conflicting
accounts of exactly what transpired at the IRB at Loftus and Guyer’s respective institutions. Loftus appears to have avoided sanctions in connection
with this case on the grounds that she was not conducting “research” as that
term is defined by relevant regulations. Such a decision would not involve
“exoneration,” as it has often been characterized, so much as a lack of jurisdiction. It has been reported that Loftus was instructed not to contact Nicole
Taus’s mother again without permission from the IRB and told “she should
take an ethics class” (Shea, 2003). Moreover, Tavris (2002) reported that
3292
Journal of Interpersonal Violence 29(18)
two of the three IRB committee members at the University of Washington
recommended that Loftus be reprimanded but that a dean overturned that
recommendation. What happened at the University of Michigan is also
unclear because the relevant documents that would clarify that issue are confidential. Although some of those documents have been quoted selectively
by Tavris (2002), none have been released by the parties who claim to have
been treated unfairly, and a faculty member at Michigan who was involved
in the case has taken issue with how Tavris and Guyer have characterized the
matter (Berent, 2002).
The constitutional issue in the case is whether the First Amendment protects Loftus and others against civil damages related to invasion of privacy,
which Taus alleged in a lawsuit that reached the California Supreme Court
(and made her real name public for the first time) before it was settled. The
Supreme Court’s decision that the First Amendment trumped all but one of
the torts alleged by Taus says nothing about the ethics of any of those actions
(Taus v. Loftus, 2007). The First Amendment protects a vast array of objectionable and unethical behavior. It even protects dubious and biased
research—against civil claims for monetary damages. That does not mean
that we should celebrate such work. Moreover, it is only because of how the
court interpreted the application of California’s anti-SLAPP suit provisions1
that Taus was put in the bizarre position of having one of her causes of action
sustained by the California Supreme Court while being told at the same time
that she had to pay several hundred thousand dollars in attorneys fees to the
defendants for the ones that were dismissed. Under those conditions, Taus
could not afford to keep litigating. Accordingly, she settled the single remaining cause of action, which, if litigated, would have provided direct testimony
about the alleged trickery that was used to obtain information.
Two general questions motivate the analysis that follows. First, when, if
ever, is it acceptable for an academic psychologist to pierce the veil of confidentiality in an anonymous case in the psychology literature for the purposes
of research? Second, if such an inquiry is justified, what are the ethical issues
involved in such research? Those questions will be considered after clarifying the behavior at issue and considering the justifications that Loftus and
Guyer offered in their article for those actions.
Clarifying the Behavior
Loftus and Guyer (2002b) argue in “Who Abused Jane Doe? The Hazards of
the Single Case History, Part II” that critics of their investigation were challenging their right to “track down information, reassess the evidence and
claims, and come to a different conclusion than Corwin’s” (p. 39). But
Cheit
3293
nothing that has been written about this case challenges the reassessment of
evidence and claims, or the right of others to “come to a different conclusion.” The original publication in which Jane Doe’s case appeared provided
the opportunity for various commentators to do just that: reassess the evidence and claims and reach their own conclusions. The concerns raised by
Nicole Taus are all related to “tracking down information.” That innocuous
phrase stands in for activities that were not fully disclosed by Loftus and
Guyer in their article. Instead, they were, at times, more poetic than specific
in describing the nature of their inquiry. By their account, they “set out on an
odyssey to learn more about the case” (Loftus & Guyer, 2002a, p. 29).
What was involved in that “odyssey”? The authors were fairly specific
about the first step: “After a long and tedious search of the social security
death records and newspaper obituaries, we found out who [Nicole’s father]
was” (Loftus & Guyer, 2002a, p. 29). Beyond that, however, these scientists
were surprisingly vague about their methods. As Loftus and Guyer (2002b)
put it, “We tracked down many documents pertinent to [Nicole’s] case and
met a few individuals who knew her” (p. 40). What was actually involved in
“tracking down” those documents and how they came to have “met” a “few
individuals” was not spelled out. Those actions are at the heart of any ethical
analysis of the research methods in this case.
Two actions are particularly noteworthy in this case. First, Loftus engaged
the services of two private investigators to figure out Jane Doe’s identity, to
obtain court records and other documents, and to locate key people in Jane
Doe’s family. How, exactly, the investigators accomplished those tasks was
contested in the civil suit, and it was never resolved through a full adjudication. The California Court of Appeals thought there was sufficient reason to
think that some of the court records that Loftus obtained might have been
obtained through trickery that they would have allowed the suit to proceed
(Taus v. Loftus, 2005). Even the California Supreme Court, which dismissed
those counts, ruled that the allegations concerning misrepresentations by
Loftus herself were not immunized by the First Amendment against the possibility of civil damages (Taus v. Loftus, 2007).
Loftus used private investigators for more than just identifying Nicole
Taus and locating documents pertaining to her legal status as a child. She also
used one of them as an intermediary who actually conducted an unspecified
number of interviews. For scientists, of course, it is a well-accepted norm that
assistants should always be appropriately credited for their work. In this case,
however, the assistant’s name and occupation was not disclosed by the
authors.
Second, Loftus did more than investigate this family. She intervened in a
personal manner, insinuating herself into the lives of people around Nicole
3294
Journal of Interpersonal Violence 29(18)
Taus, particularly her estranged mother, who had been adjudicated to have
been abusive years ago. Loftus has since acknowledged that she befriended
the woman and that she secretly hoped she could reunite Nicole Taus and her
mother.
Three Proffered Defenses
Were the actions in this case justified in ethical terms? Loftus and Guyer have
offered three different defenses of their actions, each of which is examined
below.
Oral History
Through their lawyers, Loftus and Guyer argued at the California Court of
Appeals that they were doing nothing more than “oral history interviewing”
(Opening Brief of Appellants, Taus v. Loftus, 2004, p. 37). Oral history is an
accepted practice and nobody questions the ethics of doing oral histories in
general. But the claim that Loftus and Guyer were engaged in oral history
cannot be squared with the Principles and Best Practices for Oral History,
which clearly express that oral history
begins with an audio or video recording of a first person account made by an
interviewer with an interviewee (also referred to as narrator), both of whom
have the conscious intention of creating a permanent record to contribute to an
understanding of the past. A verbal document, the oral history, results from this
process and is preserved and made available in different forms to other users,
researchers, and the public. (Oral History Association, 2009)
It is not clear which of the interviews conducted by Loftus and her private
investigator were taped in the first instance, and how many of those recordings have been preserved. Carol Tavris has written that the interview with the
foster mother was not taped (Tavris, 2002). Moreover, Loftus and Guyer have
not made any of these interviews available to “to other users, researchers, and
the public.” To the contrary, they did not even include these interviews in the
references section of their article, leaving the reader without any of the markers that come with a standard academic citation.
Journalism
Loftus and Guyer have also invoked journalism as a defense. In an appellate
brief, they argued that “their actions are no different from those of reporters
Cheit
3295
who use routine reporting techniques to discover information” (Opening
Brief of Appellants 2004, p. 34). Similarly, Carol Tavris (2002) has cast the
case in terms of Loftus and Guyer’s “right to do what good reporters do every
day” (p. 41). But these claims are impossible to square with various statements of journalistic ethics. There are two significant ways in which the
Loftus and Guyer’s article violated basic principles of journalistic ethics.
Unjustified use of anonymous sources. Journalistic codes of ethics are clear that
anonymous sources are to be disfavored. “We should identify sources as
completely as possible,” states the “Los Angeles Times Ethics Guidelines”
(2011). Those guidelines state further that “An unnamed source should have
a compelling reason for insisting on anonymity, such as fear of retaliation,
and we should state those reasons when they are relevant.” The Guidelines
also say, “The reporter and editor must be satisfied that the source has a sound
factual basis for his or her assertions. Some sources quoted anonymously
might tend to exaggerate or overreach precisely because they will not be
named.”
Loftus and Guyer violated these norms repeatedly by providing anonymity to three different professionals in their article without any statement about
the reasons why. Given the obvious concern that anonymous sources might
have an axe to grind, one wonders why Loftus and Guyer did not identify Dr.
S., the clinical psychologist who they assert was the most important person in
the case. Without identifying information, the reader has no way to assess
whether this psychologist might have an established bias or some history of
professional misconduct that would speak to their qualifications and credibility. Loftus and Guyer also did not identify the former emergency room nurse.
In neither case did the authors provide a reason for not disclosing the identity
of these professionals.
The most glaring violation of journalistic norms about anonymous sources
involves the private investigator whom Loftus hired. He appears to have
played a critical role in the most important interview that the authors conducted—the one with the mother, who had been found to have abused Nicole
in the 1980s. According to Loftus and Guyer (2002a),
When he explained why he was there, Mom welcomed him sobbing her way
through his interview, saying, “I never thought this day would come.” (p. 30)
Loftus and Guyer refer to him as “our assistant.” It was revealed through
Taus v. Loftus that the assistant was actually a private investigator named
Harvey Shapiro. There are two reasons that Shapiro should have been identified by name. First, that is the only way to allow the reader to make any
3296
Journal of Interpersonal Violence 29(18)
independent assessment of the source. As the New York Times’ Confidential
News Source Policy explains,
When we use such sources, we accept an obligation not only to convince a
reader of their reliability but also to convey what we can learn of their
motivation—as much as we can supply to let a reader know whether the
sources have a clear point of view on the issue under discussion. (New York
Times, 2012)
Harvey Shapiro’s business—which can be found through www.theinnocenceteam.com, formerly www.accused.com—specializes in representing
people charged with sex crimes. His business plan involves “a clear point of
view” about sex crimes—he takes the defendant’s view.
The other reason that Harvey Shapiro should have been identified is that he
conducted perhaps the most important interview in their investigation—the
first interview with the mother. He paved the way for the later interview by
Loftus and Guyer. This process has not been scrutinized. Why would a statement by Shapiro that simply explained “why he was there” cause the mother
to sob with joy and relief? It must have been something quite sympathetic to
the woman he had not yet interviewed. This is not the only fishy thing about
the role of this private investigator in this case.2 Shapiro’s role is also central
to the one cause of action that was left in place by the Supreme Court because,
according to Cantrell’s declaration, he told her that Loftus worked with
Corwin. Cantrell’s declaration concerning the misrepresentations that induced
her to allow Loftus to interview her in the first place was held sufficient to
state a prima facie case for invasion of privacy, sufficient to defeat the antiSLAPP motion. Loftus flatly denied the allegations of misrepresentation, but
that is what would have been tried if the case had not settled.
Violation of basic fairness. Journalistic codes of ethics are also clear that it
violates basic tenets of fairness to write a story about someone without providing them an opportunity to respond. For example, the “Los Angeles Times
Ethics Guidelines” (2011) state as follows:
People who will be shown in an adverse light in an article must be given a
meaningful opportunity to defend themselves. This means making a good-faith
effort to give the subject of allegations or criticism sufficient time and
information to respond substantively.
The Associated Press Statement of News Values and Principles also states
this obligation in mandatory terms: “Whenever we portray someone in
Cheit
3297
a negative light, we must make a real effort to obtain a response from that
person” (Associated Press, n.d., para. 9). There is no question that Loftus cast
Nicole Taus in an adverse light, particularly by mentioning juvenile court
proceedings that occurred years after the contested events in 1984.
Loftus and Guyer stated two reasons for not contacting Nicole Taus. One
of those explanations contradicts the fundamental basis of journalism: seeking as much information as possible and keeping an open mind. As the “Los
Angeles Times Ethics Guidelines” (2011) puts it, “Reporters should try genuinely to understand all points of view.” Further, “We seek out intelligent,
articulate views from all perspectives” (“Los Angeles Times Ethics
Guidelines,” 2011, Fairness section, para. 2). In dramatic contrast to that conception, Loftus and Guyer (2002b) assert that “Jane’s own account at this
point might well not shed additional light” (p. 40). Professional journalists
would never spurn an interview with a key subject because they guessed that
the person “might well not shed additional light.”
Their second reason for not providing Nicole Taus with a voice is couched
in terms of ethics. As Loftus and Guyer (2002b) put it, “We worried that such
contact might be upsetting to Jane” (p. 41). Avoiding the harm that could occur
by intruding into the life of someone who thought her case was anonymous
might well be sufficient reason for deciding against embarking on the project.
But Loftus and Guyer did not state their concern about possible “upset” in
terms of deciding whether or not to investigate her life. They were already
planning to publish an article filled with private information about her. Their
question was apparently whether or not to give Nicole Taus the opportunity to
respond, and whether the harm that they were already willing to cause would
be made even worse if they contacted her. For reasons they did not articulate,
they made the judgment that Taus would somehow be less upset if they did not
provide her with a meaningful opportunity to speak than if they did. They did
not, in other words, respect her privacy or her autonomy.
An Analogy to Medical Research
Loftus and Guyer also reported in their article that they consulted with
Thomas McCormick, a physician with expertise in medical ethics. One wonders why two prominent psychologists would not seek the advice of someone
in psychology. Whatever the reason, the medical ethicist they consulted
apparently offered
. . . a hypothetical situation in which a professional has published a case history
claiming that he cured cancer using marijuana leaves and Crisco. Oncologists
would naturally have many questions about this case study: Did it really work?
3298
Journal of Interpersonal Violence 29(18)
If the patient seems to be in better health after the “treatment,” did he or she
really have cancer in the first place? Would it be ethical for a physician to talk
to the “case history” and to examine the original doctor’s data? McCormick
thinks so, and so do we. (Loftus & Guyer, 2002b, p. 40)
Presumably, the reader is supposed to find that hypothetical example to be
so close to the case at hand that the medical ethicist’s view would carry over
as well. But there are many obvious differences between that hypothetical
example and the case at hand. First, the hypothetical example is about allegedly saving someone’s life with a treatment. In other words, the implications
of the underlying facts of the case could have life-or-death consequences to
many other people. That is not true of the case at hand. The case presented by
Corwin and Olafson does not counsel any kind of treatment or standard of
care for anyone. Accordingly, there was all the more reason to protect the
subject’s privacy.
Second, the question in the hypothetical example is phrased, awkwardly,
in terms of whether it is ethical “for a physician to talk to the ‘case history,’”
which presumably includes talking to the subject of the case history. Yet, that
is precisely what Loftus and Guyer did not do in this case. They talked to
people around the subject of the case study, including people not mentioned
in any way in the (anonymous) case. Third, the hypothetical does not involve
a 11-year delay between the underlying events and the later “investigation.”
It would be more relevant to ask whether it would be ethical for a physician
to use a private investigator to track down family members who were never
mentioned in a published but anonymous case study and contact them out of
the blue with questions about intimate events that occurred more than 10
years ago. My guess is that no ethics expert in medicine or psychology would
provide a simple “yes” in response.
Loftus and Guyer’s own ethics consultant is clear that “respect for autonomy of the patient” and “avoiding harm” are the first two principles in cases
across a range of issues in health care ethics (McCormick, 2010, p. 1188).
But Loftus and Guyer ignored Nicole Taus’s autonomy in deciding she would
be less “upset” if they wrote this article without talking to her. As for avoiding harm, Loftus and Guyer (2002b) mention the concept only once in their
article, when they state that the kind of inquiry they conducted is justified “as
long as this can be accomplished without undue harm” (p. 40). But the authors
did not follow that critical sentence up with any description or analysis of the
harms in this case. We are left unsure whether the authors recognize that there
were harms, and if they did, why they came to the conclusion that the harms
were not “undue.” Nicole Taus’s essay in this issue certainly makes it clear
that there were harms. Those were short term and immediate. There are also
Cheit
3299
important long-term harms to consider if this kind of invasion into an anonymous case is considered acceptable by the profession.
Loftus and Guyer’s view seems to be that their work is so important that
the harms—which are left unacknowledged—are definitely worth it. Of
course, deciding whether the benefits of one’s own work are worth the harm
to others is tricky business. The researcher is likely to undervalue the harms
and overvalue the benefits of his or her own work. And that appears to be
exactly what Loftus and Guyer did. First, they did not take into account the
possible harms in this case. Second, their account of the benefits gives far
greater importance to this case than it had in the real world. This was hardly
the first corroborated case of recovered memory. The Recovered Memory
Project at Brown University contains an archive of over 100 corroborated
cases of recovered memory (www.recoveredmemory.org).
Nicole Taus’s case had particular power because it involved videotape, but
the idea that the case was being offered as definitive proof of recovered memory is contradicted by the title of the article, “Videotaped Discovery of a
Reportedly Unrecallable Memory of Child Sexual Abuse.” Acknowledging
that the discovery was “reportedly [emphasis added] unrecallable” highlights
one of the inherent limits of any case study of this nature (Corwin & Olafson,
1997). The content of Corwin and Olafson’s presentation also contradicts any
notion that they offered the case as some kind of gold standard proof of
repressed memory. Corwin and Olafson (1997) left open the following questions: “Was Jane Doe’s memory truly unavailable, or was it just that she had
never specifically tried to recall sexual abuse?” (pp. 110-111). Moreover, the
case was published in a professional journal that included one commentary
that took a completely different view of the case (Neisser, 1997). From the
moment of original publication, then, this case was presented in qualified
terms. That is appropriate to the presentation of case studies. These facts
contradict the claim that this case was held out as some kind of ultimate proof
about recovered memory.
Respecting Confidentiality and Privacy
Returning to the two larger questions that motivated this comment: When is
it acceptable for a research psychologist to pierce the veil of confidentiality
in an anonymous case in the published literature for the purposes of research?
Second, if such an inquiry is justified, what are the ethical issues involved in
such research?
Loftus and Guyer (2002a) point out various instances in which case studies have later been seen in a very different light as a result of some kind of
investigation independent of the author of the original case study. Most
3300
Journal of Interpersonal Violence 29(18)
famously, Freud’s case studies have been reanalyzed in various ways.
Similarly, Cornelia’s account of Sybil has been challenged by later analyses.
Also, John Money’s case study of a male infant who received sex reassignment surgery was used for years as evidence that sexual identity was social,
not biological. But Loftus and Guyer described these cases with a passive
voice that obscures how they later came to light. For example, they describe
the dramatic follow-up on Money’s case as follows: “Subsequent investigation revealed that the particular boy, David Reimer, never adjusted well and
reverted to life as a male” (Loftus & Guyer, 2002a, p. 25). But this case was
not unmasked by someone employing a private investigator and investigating
the man’s life without his approval. The investigation was conducted through,
and with the approval of, his treating psychiatrist. The more famous cases
(Freud’s patients, Wilbur’s “Sybil”) involve archives of clinical papers, not
explorations by private investigators. None of the cases that they cited as
precedent for their work involved the kind of invasion of privacy that existed
in Nicole Taus’s case.
Official Cases Versus Clinical Cases
In assessing the question about piercing the veil of confidentiality, there is a
major distinction between “cases” in official government forums and “cases”
of a purely clinical nature. Obviously, some cases, including Nicole Taus’s,
are in both categories. But the distinction is nevertheless useful for the position that I propose: It is never acceptable for a psychologist to seek to identify
and research the subject of an anonymous clinical case. If clinical materials
have been archived, then those materials are obviously appropriate to examine. But seeking to identify a clinical patient to independently evaluate a published clinical case report is overly intrusive. Consider Kaplan and
Manicavasagar (2001) who presented three case studies in support of “false
memory syndrome.” One involved a 40-year-old woman, the third of five
children, who suffered from panic attacks and, through therapy, eventually
came to believe that she had been sexually molested by her father. The woman
eventually was referred to one of the authors (V. M.), who reported that the
patient’s belief that she had been sexually abused “progressively withered
away” (Kaplan & Manicavasagar, 2001, p. 345). The authors cite “the closeness of the family,” as attested to by the woman’s sister, as one of the reasons
it is unlikely that the molestations occurred (Kaplan & Manicavasagar, 2001,
p. 345) This case involves the same controversial subjects as Nicole Taus’s
case. But that alone would not, in my view, justify engaging a private investigator to try to identify the woman and then track down other members of her
family to investigate the details contained in the authors’ report.
Cheit
3301
There are two basic reasons in support of this position. First, the expectation of privacy is strongest when one’s case has never been in any kind of
official forum. The mental health treatment relationship is built on confidentiality. That value is undercut if third-party research psychologists can ignore
what first-party treating psychologists are duty-bound to protect. This standard asks more of psychologists than it does of journalists. A tabloid newspaper might be well within the protections of the First Amendment if it staked
out a therapist’s office and published stories that identified and embarrassed
patients. But those stories would not have the voice of authority that comes
with being a member of a profession unless they quoted a psychologist. In
this way, the ethical psychologist is held to a higher standard than the National
Inquirer. And that is one of the reasons why the word of the professional
psychologist carries more weight.
The implication of this distinction between clinical and official cases is
that archival research about cases that occurred in official proceedings is
acceptable even if those proceedings were confidential, and even if, as in this
case, the published case study is from a clinical setting. There were official
records in this case from divorce proceedings and child protective service
(CPS) investigations. Many of those documents were either sealed or marked
confidential. But those designations should not be used to determine the ethical boundaries of scholarly inquiry. First, official designations can be overly
broad. Second, what matters most is whether the purpose of the confidentiality is honored by the researcher. The primary reason for confidentiality in
cases involving child sexual abuse is to protect the identities of children from
public disclosure. It is ethically acceptable, then, to research a confidential
case of that nature so long as the researcher also protects the identity of the
child. To their credit, Loftus and Guyer did that in this case.
But the ethical obligations concerning confidential information go further.
The researcher who accesses confidential court documents has an ethical
obligation to limit any published information to relevant facts. In the case at
hand, the potentially relevant documents would be those related to the years
when the abuse allegations were evaluated and litigated. But Loftus and
Guyer went much further. They used confidential information from Nicole
Taus’s juvenile court records in an apparent effort to impugn Taus’s character.
That had nothing to do with arguments about who abused Jane Doe years
earlier. Lacking sufficient relevance to the issues in the case, publishing
information that impugns someone’s character but does not actually speak to
the issues in the case is unethical. The same applies to Loftus and Guyer’s
decision to inform their readers that the stepmother, who rejects their interpretation of this case completely, had once been arrested for vandalism,
although the charges were dropped (Loftus & Guyer, 2002a).
3302
Journal of Interpersonal Violence 29(18)
Archival Research Versus Interviewing
While the acceptable scope of potential research in the realm of official proceedings is quite permissive, it is similarly restrictive on the issue of interviewing parties to a case that was presented anonymously. There is a world of
difference between examining all of the documentary evidence available in
the case and tracking down people years later and talking to them about those
events. One major difference involves the nature of the intrusion into private
lives that have a reasonable expectation of privacy—that is, research subjects
who have been involved in research where their identity was protected well.
Writing an article based on original archival research, which was part of
Loftus and Guyer’s effort, is not significantly different from writing one
based solely on the materials that were available at the time. In the case of
original research, the research subject would know that the researchers had
figured out his or her identity in order to conduct the research. But when the
research is restricted to archival research, the research process itself is completely separate from the life of the research subject. Interviews are entirely
different. Interviewing people years after the fact about family dynamics in a
dysfunctional family is bound to have a serious effect on contemporaneous
interpersonal relationships. Psychologists are supposed to share the same
imperative that doctors have to avoid doing harm above all else. That would
seem to counsel against intruding in families to conduct interviews about
sexual abuse allegations years after they have been resolved in the courts.
This distinction preserves the important ability to reassess cases in the
literature. Reassessment can be accomplished with the case material alone. A
good example is the commentary on Corwin’s and Olafson’s original presentation of the case. Reassessment might even include original research into
additional documentary evidence from the case. But there is a difference
between reassessing and relitigating a case. What Loftus and Guyer did went
beyond reassessing existing evidence. They conducted interviews that have
not been made available to others and then relied on them to reinterpret the
case. By all appearances, their effort was closer to advocacy work for the
defense.
The nature of Loftus’s activity in this case went far beyond interviewing
and advocating. Loftus admitted “befriending” Nicole Taus’s biological
mother (Kelleher, 2003). She did not disclose that fact in the Skeptical
Inquirer article about the case, let alone address the conflict between her
friendship and her capacity and willingness to research the case in an objective fashion. That concern is magnified by the fact that this was not just a
simple friendship; it was a friendship with an agenda. Loftus told Kelleher
(2003) that “she was motivated in large part by a desire to unite the mother
Cheit
3303
and daughter.” Whether a research psychologist without any clinical experience or training should ever intervene in a family with such intentions is a
question worth considering. For the purposes of this commentary, however,
the more important question is why Loftus did not disclose this friendship in
her article about the case. The woman she “befriended” is the most important
character in the claim that Corwin and Olafson’s account of the case is wrong.
As Loftus’s interpretation might be the product of her friendship and other
ulterior motives, this “evidence” is questionable at best. This blurring of professional and personal lines raises ethical issues involving neutrality and
objectivity. Psychologists are supposed to maintain professional distance
from their clients and their research subjects. This protects against results that
are influenced by personal considerations. Objectivity may be lost if psychologists have social relationships with clients or research subjects. The relevance to this case has been lost in all the attention to the First Amendment,
which, of course, places no restrictions on psychologists getting too close to
their subjects. But the American Psychological Association Ethical Principles
of Psychologists and Code of Conduct does. The American Psychological
Association (APA, 2010) says one should refrain from activities that “could
reasonably be expected to impair the psychologist’s objectivity” (Standard 3,
Human Relations section, para. 1).
What About Medical Records?
Loftus and Guyer’s ethical framework apparently included special considerations for confidential medical records. They state at one point, while discussing the allegations concerning burned feet, that “of course, we did not get
[Nicole’s] medical records” (Loftus & Guyer, 2002b, p. 39). Loftus and
Guyer do not explain why their position is so obvious. If the reason is that the
documents are confidential by law, then why did not the authors apply the
same logic to juvenile court records and CPS records? If the reason is that
medical documents are somehow more confidential than other documents,
then why did not that logic apply to the letter they quote from “Dr. S”? Loftus
and Guyer (2002b) also wrote that “we learned from other sources that Jane
had a fungal condition that could have been responsible for injuring her feet”
(p. 38). That seems like an invasion of her medical privacy, but because they
did not specify anything about the nature of these “other sources,” it is impossible to know whether to give these statements any credence. That kind of
loose attribution of factual claims would not be tolerated in a peer-reviewed
scientific journal. Moreover, it would not be necessary if the inquiry was
restricted to documented court records. Moving beyond court records is
bound to raise this problem in virtually any clinical case in psychology, as
3304
Journal of Interpersonal Violence 29(18)
clinical records have the same strong confidentiality provisions as medical
records. There is no ethical way to study a clinical case without the patient’s
approval that is both comprehensive enough to provide strong scholarship
and yet respectful enough of privacy and medical confidentiality to honor the
laws and related professional norms.
Conclusion
The novelty of what Loftus and Guyer did in this case has been lost in the
broader framing of the entire case in terms of the First Amendment. Loftus
has claimed that this case is about her “right to speak out on matters of grave
importance” (Loftus, 2003). But none of the criticisms of her behavior in this
case involve “speaking out.” They involve invasions of privacy and ethical
lapses. The question of whether the conduct is proper in an ethical context is
not answered by the fact that it may not be actionable in a legal context. I
have not been able to find any examples in any academic publications where
a professor engaged private investigators to ascertain the identity of someone
from an anonymous clinical case report.3 The novelty of the actions in this
case goes much further, including using the investigator to track down members of that person’s family and question them about highly sensitive events
from many years ago, all without the consent or involvement of the research
subject. Those facts contradict Loftus’s protestation that she has been unfairly
criticized for engaging in “a reasonable quest for information on a controversial subject” (Loftus & Geis, 2009, p. 161).
What are the implications of this case for the publication of case studies in
psychology and medicine? If what Loftus and Guyer did is considered acceptable, then individuals involved in potential case studies will have to be
warned that their family and friends could get contacted out of the blue
decades later if someone decides to devote the energy to ascertain their identity and probe into their private life. It is difficult to imagine why anyone
would ever agree to participate in an anonymous case study if this is one of
the risks that needs to be disclosed.
Loftus and Guyer purported to be doing a case study of a case study. But
the contrast between Corwin and Olafson’s case study and Loftus and Guyer’s
case study of a case study involves three important distinctions that have not
previously been highlighted: (1) Corwin published in a professional journal
that invited a range of commentary that included Ulrich Neisser, who challenged the accuracy of the memory and offered an alternative hypothesis.
Loftus and Guyer published in a popular magazine that did not invite commentary from anyone who might challenge the authors’ interpretation and
that published claims that were so poorly sourced that they would never pass
Cheit
3305
muster in a scholarly journal. (2) Corwin qualified the case that he presented
in several ways that were appropriate to a case study, but that also made his
claims less than definitive. Loftus and Guyer also stated her conclusions in
qualified terms, but Loftus has since made hyperbolic claims to journalists
that contradicts those qualifications, stating that the mother had been “railroaded”4 (Grossman, 2003). (3) Corwin was quite transparent about his process, making the tapes available to other scholars. Loftus and Guyer were not
transparent about their use of a private investigator, nor have they been transparent with any of the fruits of those efforts. Their article does not live up to
its stated promise of producing “much valuable information that should assist
scholars in making their own decisions” (Loftus & Guyer, 2002a, p.29).
Indeed, it does not answer the question of “who abused Jane Doe” as much as
it raises the question of who abused Nicole Taus.
The meaning of case studies in psychology can and should be subject to
scientific debate. Such debate should reflect the kind of analysis reflected in
comments published along with Corwin and Olafson’s original publication. It
can even involve additional investigation of documents that can be obtained
with trickery or violations of law. But it crosses an ethical line to employ a
private investigator to identify and interview family members connected to
an anonymous case study. It also violates the profe...
Purchase answer to see full
attachment