Researchers can also compare people from derived groups, which means
that the researcher puts participants into groups based on how they rank
on certain variables of interest. For instance, Groomes and Leahy [5] used
a measurement tool to find out what kind of coping strategies
rehabilitation clients used most often. They then put them into three
groups (problem-, emotion-, or avoidance-oriented) to see how accepting
they were of their disability. When doing causal-comparative studies, the
researcher needs to think about the kinds of variables that will be used in
the study first. Like experimental designs, causal-comparative studies use
independent variables that are nominal or categorical. Causal-
comparative studies also usually use dependent variables that are
continuous. Categorical variables are those in which the participants are
put into one of two or more groups that cannot overlap. When you
measure a continuous variable, you look at how much or how high it is.
Higher scores mean that the variable is higher. A causal-comparative
design can use one or more independent variables, but the most
important thing about causal-comparative research is that the
independent variables are (a) categorical and (b) not changed in an
experiment. Researchers often look at how an experimental treatment
affects people of different races by combining independent variables that
are manipulated in the lab with those that aren't. This kind of study would
use both causal-comparative and experimental methods. 3. What makes
causal-comparative designs different from experimental designs When
researchers do experimental studies, they choose people to take part and
then randomly put them into groups. In a simple two-group experiment,
one group gets treatment and the other group doesn't. The same set of
procedures is used on both groups, except that the control group does not
get the experimental treatment. This step is taken to make sure that any
outside factors that could affect the expected results are kept under
control. The researcher then looks at the dependent variable, which is an
outcome variable of interest, to see how the treatment affected that
variable in the two groups. In some studies, though, participants may
already be grouped by variables.