element that you predict is the cause of the change, the ingredient that is added to set
things in motion: the lit stove in the example above. The dependent variable is the
one that changes, the variable whose change “depends” on the introduction of the
independent variable: the coffee in the pot.
These are the key types of variables. But there are others. There are extraneous
variables, which may influence the outcome of an experiment but are not actually of
interest to the researcher. Extraneous variables might include the material the cof-
feepot is made of and whether your stove uses gas or electricity. (These might influ-
ence the speed of the boiling, or how high the temperature is, but they’re not what
you are interested in.) And there are confounding variablesthat may be affecting the
results of the study but for which you haven’t adequately accounted. Again, in the
example above, the intelligence of the researcher to correctly sort the pots might
confound, or complicate, the result.
Sociologists rarely conduct experiments: It’s too hard to change the independent
variable. Say you want to know if children of divorced parents are more likely to
become juvenile delinquents. You can hardly divide children into two groups and force
the parents of the first to divorce and the second to stay together.
Instead of experiments, sociologists are likely to engage in the following types of
research:
■Observation. Observing people in their natural habitat, joining their clubs, going
to their churches, getting jobs in their offices. This is usually called “participant
observation.”
■Interviews. Asking a small group of people open-ended questions, such as, “Can
you describe your last road rage experience?”
■Surveys. Asking a lot of people closed-ended questions, such as, “How many
times have you got angry in traffic in the last month?”
■Content analysis. Analyzing artifacts (books, movies, TV programs, magazine
articles, and so on) instead of people.What about going to the library and looking things up in books? Isn’t that doing
research? Sociologists would call that an incomplete literature review. A real literature
reviewneedn’t perform any original or new research, but it must carefully examine
all available research already done on a topic or at least a systematic sample of that
research, through a specific critical and theoretical lens.
Let’s look at each of these methods in a bit more detail.Observational Methods
In all observational studies, we directly observe the behavior we are studying. We can
do this in a laboratory, conducting an experiment, or we can do it in the place where
it more “naturally” occurs. When we observe phenomena, we do more than just
watch—we watch scientifically, testing hypotheses against evidence.Experiments.Anexperimentis a controlled form of observation in which one
manipulates independent variables to observe their effects on a dependent variable.
To make an experiment valid, one typically uses two groups of people. One is the
experimental group, and they are the group that will have the change introduced to
see what happens. The other is the control group, and they will not experience the
manipulation of the variable.
A control group enables us to compare the outcomes of the experiment to deter-
mine if the changes in the independent variable had any effects on the dependent vari-
able. It is therefore very important that the experimental group and the control group112 CHAPTER 4HOW DO WE KNOW WHAT WE KNOW? THE METHODS OF THE SOCIOLOGIST