176 CHAPTER 5
Placebo effect
A positive effect of a medically inert
substance or procedure.Attrition
The reduction in the number of participants
during a research study.and hearing loud clicking noises (like those produced during TMS when a magnetic
fi eld is discharged) can affect symptoms, whether or not magnetic pulses are actu-
ally sent to the brain (Loo, 2004; Nahas et al., 2003).
To discover whether a biomedical treatment is effective, researchers give a group
of participants a placebo, an inert substance or a procedure that itself has no direct
medical value. Often, a placebo is simply a sugar pill. A positive effect of such a
medically inert substance or procedure is called a placebo effect. If patients who are
given a sugar pill show the same improvement as patients who are given the pill with
active ingredients, then a researcher can infer that the medication itself is not effec-
tive and that its apparent benefi t is a result of the placebo effect.
The placebo effect can be strengthened or weakened by the outward qualities
of the placebo and how it is administered: Taking more placebo pills generally has
a greater effect; capsules do a better job than pills; and injections do better than
capsules. The placebo can have an even greater effect when the person dispensing it
shows interest in the patient’s problems and is sympathetic and friendly (Shapiro &
Morris, 1978). And placebos that patients are told cost more are more effective
than are “less expensive” ones (Waber et al., 2008). If a study is not double-blind,
an investigator’s high expectations of a treatment’s success also can cause a positive
placebo effect (Shapiro, 1964). (In double-blind studies on treatment, the mental
health clinician administering the treatment—as well as the patient—is blind to the
patient’s group assignment.
Thus, when investigators test the effects of any new medicine or other biomedi-
cal treatment, they invariably use some type of placebo treatment as a comparison.
They may also include additional comparison groups, in which patient receive more
established treatments, and researchers will likely use a double-blind design. Not all
researchers are convinced, however, that patients and clinicians are truly “blind” in
double-blind studies. In double-blind studies of medications for depression and anx-
iety, for instance, both patients and (“blind”) clinicians may correctly guess whether
patients are receiving a placebo or active medication based on the presence of side
effects and changing symptoms (Margraf et al., 1991; White et al., 1992).
Some researchers have suggested that in order to evaluate the results of double-
blind studies of medication, investigators should determine whether patients and
their clinicians were actually blind to the treatment that the patients received
(Piasecki et al., 2002); if not, the investigators should not consider the study to
have been truly double-blind. Furthermore, some researchers also advocate using an
active placebo—a placebo that mimics the side effects of an active medication but
does not directly affect the psychological symptoms of the disorder being treated
(Kirsch, Scoboria, & Moore, 2002; Moncrieff, 2001). For instance, when studying
the effects of a tricyclic antidepressant, an investigator might use as an active pla-
cebo an antihistamine—an over-the-counter cold medicine that has some side effects
that are similar to those of tricyclics.
The fact that people who suffer from certain disorders, including depression,
sometimes improve following a placebo treatment (Kirsch, Scoboria, & Moore,
2002) does not mean that their problem was imaginary, or that they should throw
away their medication. Rather, it seems that—for some people and for some
disorders—the hope and positive expectations that go along with taking a medica-
tion (or undergoing a procedure) allow the body to mobilize its own resources to
function better (Kirsch & Lynn, 1999; Scott et al., 2008).Dropouts
On average, more than half of those who begin a treatment that is part of a research
study do not complete the treatment—they drop out of the study (Kazdin, 1994).
The reduction in the number of research participants during a study is called
attrition. When attrition is different for different groups in a research study, re-
searchers can’t easily draw defi nitive conclusions about the treatment given to one
group (Kendall, Holmbeck, & Verduin, 2004; Lambert & Ogles, 2004). Suppose,
for example, that a group receiving medication (and experiencing its side effects)The color of a placebo pill can infl uence how
effective it is. Red, pink, and yellow placebo
pills work best when people believe them to
be “ stimulants,” whereas blue placebo pills
work best when people believe them to be
“tranquilizers” (Buckalew & Ross, 1981).Nathan Griffi
th/Corbis