Extraneous Variables, Confounding Variables, and Internal
Validity
For a research study to have internal validity, there must be one, and only
one, explanation for the research results. If a study includes a confounding
variable, then there is an alternative explanation and the internal validity is
threatened. Therefore, the key to achieving internal validity is to ensure that
no extraneous variable is allowed to become a confounding variable. Because
every research study involves thousands of extraneous variables,
avoiding a confounding variable can be quite a task. Fortunately, however,
confounding variables can be classified in a few general categories that make
it somewhat easier to monitor them and keep them out of a research study.
Before we examine the different categories of confounding variables, we
look more closely at the general structure of a research study for which internal
validity is a concern.
When the goal of a research study is to explain the relationship between
two variables, it is common practice to use one of the variables to create different
treatment conditions and then measure the second variable to obtain a
set of scores within each condition. For example, Hallam, Price, and Katsarou
(2002) conducted a second study in which they created three background
conditions by playing pleasant, calming music in one room, unpleasant,
aggressive music in one room, and no music in one room. The researchers then
measured problem-solving performance (variable #2) for a group of students
in each of the three rooms. Because they found differences in the problem solving
scores from one room to another, the researchers successfully demonstrated
that problem solving depends on background music; that is, there is a
relationship between the two variables. The general structure of this study is
shown in Figure 6.3.