humans. We can exert nearly perfect control over
animals. We can control their diets, living condi-
tions, and even genetic background. We can be so
much more intrusive in the lives of animals. Because
animals generally have shorter life spans, phenom-
ena that may take years to study in humans can be
studied in a few months with, say, rats. Further,
there are a number of“naturally”occurring behav-
ior disorders (e.g., aggression, mood disorders,
hyperactivity, eating disorders) commonly seen in
veterinary practice that appear to be relevant to
the field of psychopathology (Aldelda & Joel,
2011). Some even argue for the relevance of animal
research to personality traits (e.g., Canli, 2006;
Pawlak, Ho, & Schwarting, 2008). But again,
exactly how similar is animal behavior to human
behavior? Perhaps similar enough in certain
instances, but in others, not at all.
In the final analysis, analog research is impor-
tant and can be quite enlightening. But we can
never afford to completely relinquish our skepti-
cism when we employ it.
A Closing Note. In closing this section on exper-
imental methods, it is important to note that not
everyone is enamored with these traditional
approaches. For example, many years ago, Cattell
(1965) critiqued the so-called bivariate research
strategy. This is a hallowed method, going all the
way back to Pavlov and Wundt, in which only two
variables are studied at once. The investigator
manipulates an independent variable and then
observes its effects on the dependent variable. For
example, the experimenter induces in participants
the belief that they lack personal control over the
onset of electric shock (independent variable). The
question is, what happens to the participant’s anxi-
ety level as measured by the galvanic skin response
(dependent variable)? If the focus is on anxiety, it
may become necessary to carry out thousands of
such bivariate studies to determine how people
become anxious. The experimenters must vary
measures of anxiety, the nature of the stimulus,
and the presence of preexisting personality traits
that may affect the nature of the participants’
responses. If experimenters vary one condition at
a time in study after study, they are left with a
piecemeal view of the human being. Putting the
results of all these bivariate studies together can be
worse than trying to put Humpty Dumpty together
again. And because the study is looking at anxiety
in isolation from other variables, such as compe-
tence and adjustment, the results offer no sense of
how these variables might affect anxiety.
As a consequence, some have advocated the
use of a multivariate strategy. Here, the experimen-
ters use a variety of measures on the same person
but do not exert much in the way of control. This
is a common approach in the study of child and
adolescent psychopathology. Researchers may use
questionnaire data, life records, observation, and
so on. Such data can be correlated and factor ana-
lyzed. Because the method can focus on naturally
occurring phenomena and can deal with numerous
variables simultaneously, many regard it as a super-
ior strategy. However, this method, like other cor-
relational approaches, also has its limitations (e.g.,
costs, time, and access to participants).Single-Case Designs
Single-case designsare an outgrowth of behavioral and
operant approaches. They bear similarities to both
experimental and case study methods. For example,
an experimenter measures a subject’s behavior under
several conditions and in this sense is employing a
method akin to experimental techniques. But the
focus is on the responses ofoneparticipant only.
Such research usually begins by establishing a base-
line. Here, a record is made of the participant’s
behavior prior to any intervention—for example,
the number of anxiety attacks per week. After a reli-
able baseline has been established, an intervention is
introduced. The effects of this intervention are then
determined by comparing the baseline level of
behavior with the postintervention level. Single-
case designs are often used to study the effectiveness
of a therapeutic method.
Single-case studies allow the experimenter to
establish cause–effect relationships. More than that,
they provide a method of studying clinical behavior
(especially therapy methods) that does not require theRESEARCH METHODS IN CLINICAL PSYCHOLOGY 113