CHAPTER 2 – Understanding and Testing Theories of Deviance
If deviance is a destination, many paths lead to it, some more heavily used
One path is marked by psychopathy (some people are pyromaniacs).
Another is fraud (a person burning property to claim insurance money). Yet
another path is organized crime (using arson in extortion rackets). Jilted
lovers or mischievous children may set fires.
Theories can be classified in many ways. For our purposes, one of the most
important distinctions is among theories that are mainly empathetic,
scientific, or ideological. All social theories have some empathetic, scientific,
and ideological content. What matters here is the degree of emphasis placed
on these features.
When we hear of a woman who has killed her husband, we may be quite
satisfied with the explanation that he was abusive to her or threatened her
children. The empathetic explanation, because it includes much of the
subjective (inner feeling) as well as the objective (viewed from the outside)
reality of the event being explained, often seems more complete, and thus
more satisfying, than an exclusively objective explanation.
Empathy, however, has limits. Most of us cannot imagine any circumstances
in which we would, if not already infected, seek out a partner with AIDS for
unprotected sex, yet this kind of “bug chasing” does occur.
Similarly, we cannot easily empathize with serial killing, or someone who
eats people he has invited for dinner, and yet that happens too.
We sometimes enclose the word “scientific” in quotation marks to indicate
that, for studies of deviance, science is an ideal. It is not usually an
achieved reality, because deviance is very difficult to observe and study
using the tools of the scientific method, and because human beings are
simply much more complex than rocks and stars.
We use the term “positivism” to indicate work that may be extreme in its
refusal to give any importance to human feelings, understandings, and
Ideological explanations are based on systems of ideas that are held as
irrefutable doctrine. They are not just tentatively proposed, in the manner of
scientific hypotheses, but rather are often a matter of passionate belief.
Ideologies are almost impervious to new information or alternative
interpretations. In deviance studies, ideological explanations are found most openly in the
writings of prescientific religious authorities who demanded obedience to the
dogmas of the Church and in the work of some radical conflict theorists who
believe just as tenaciously in the tenets of Marxism or one of its competitors.
In an Inductive Approach we look at many specific cases and then make
generalizations about them.
In deductive approaches, we derive specific expectations from general
rules that have been suggested either by previous research or by the logical
relations of an existing theory.
A sensitizing concept symbolizes aspects of reality that we particularly want
to think about.
A variable is a concept that can be operationalized by being counted or
The relationship between variables is usually expressed as a tentative
statement or hypothesis. It asserts that particular variables are regularly
related to each other in a specific way whenever specified conditions occur.
The classical experimental design has evolved as the ideal type of
research in science. It is the norm against which most research designs are
evaluated. Very few, if any, studies in the field of deviance fully meet all of
its criteria, but it remains an important model for our work.
The correlation coefficient is a statistical measure that tells us to what
extent, and in what ways, two variables are related to each other. Its size
helps us to decide whether a difference is statistically significant.
Causality is a complex subject in both philosophy and science. Causality is
considered strongly supported when three conditions are met. First, when
one variable changes, the others change in predictable ways. Second, we
must show that the variable deemed to be the independent (causal) variable
occurs, or changes, before the variable that is deemed to be caused
undergoes change. Third we must be able to rule out other variables that
emerge as candidates for causal status.
A Spurious correlation is one that is high but not causal.
A major goal of research is to find data that are meaningful and can be
generalized beyond one study. Reliable measures are consistent measures that presumably reflect a
regularity that occurs in nature.
To find out whether something is a reliable finding, we have to do further
research under the same methodological conditions as existed in the earlier
study. Achieving the same results will give us test-retest reliability.
Validity raises the question “Are we measuring what we think we are
measuring?” Measures can be accurate and consistent but irrelevant.
Self-report studies of workplace deviance can produce what appear to be
well-grounded statistics. These studies show much more deviance than is
reported through official channels.
Social statistics have the following components:
1. Real figure. The real figure represents the actual number of people
who engage in a particular kind of deviance, or the total number of
deviant acts that occur in a specified time and place.
2. Error component. The error component is caused by over or
3. Random errors. Random errors are generally accidental and self-