received therapy and the former did not. Therefore,
the study is internally valid. But is it really? Experi-
ence has shown that subjects who go on a waiting
list do not always fail to receive help. Instead, they
often seek help from a doctor, a minister, wise
counsel from a friend, or some other type of sup-
port. Thus, any improvement shown by the ther-
apy group may be confounded by the informal help
received outside therapy by subjects on the waiting
list. Therefore, the fact that the therapy group
improved no more than did the waiting-list group
does not necessarily mean that the therapy inter-
vention was ineffective. It may only mean that
both groups received some form of intervention.
Another factor that detracts from internal valid-
ity involvesexpectations. When either the investiga-
tor or the participant expects a certain outcome,
that very outcome may be produced. It is not the
experimental manipulation that causes the out-
come; it is the expectations. This phenomenon is
called aplacebo effect. For example, people have been
known to behave in a drunken manner when they
believe that the situation calls for such behavior
even when they have not been drinking alcohol
but only thought they were. Patients have been
known to report that therapy has helped them
when objectively they did not seem any better.
Because the therapist has devoted so much time
to them, they feel they must be better!
In other cases, experimenters have been known
to unwittingly produce the very responses they
expected to get. Perhaps they acted in subtle ways
that encouraged their patients to behave in the
“proper”fashion. This is especially likely to happen
when the experimenter knows who are the experi-
mental subjects and who are the control subjects.
Clinicians have been known on occasion to
“unconsciously”give a patient a little extra time
on a test item simply because they expected the
patient to do well.
To avoid the effects of experimenter or
participant expectations, studies commonly use a
double-blind procedure. Here, neither participant nor
experimenter knows what treatment or procedure
is being used. For example, if an investigator is
interested in studying the effects of two drugs, theLaboratory experiments often have high internal validity.
copyright Robin Nelson/PhotoEditRESEARCH METHODS IN CLINICAL PSYCHOLOGY 111