Social Research Methods: Qualitative and Quantitative Approaches

THEORY AND RESEARCH

To show causality, an association does not have

to be perfect (i.e., every time one variable is present,

the other is also). In the example involving exam

grades and days of the week, there is an association if

on ten Fridays I got seven As, two Bs, and one C,

whereas my exam grades on ten Mondays were six

Ds, two Cs, and two Bs. An association exists, but the

days of the week and the exam grades are not perfectly

associated. The race and income-level association

shown in Figure 2 is also an imperfect association.

3.Eliminating alternativesmeans that we must

show that the effect is due to the causal variable, not

to something else. It is also called no spuriousness

because an apparent causal relationship that is actu-

ally due to an alternative but unrecognized cause is

called a spurious relationship. While we can observe

temporal order and associations, we cannot empiri-

cally eliminate all logical alternatives. Eliminating

possible alternatives is an ideal. This means we can

demonstrate this only indirectly or rule out the more

obvious alternative explanations. In an experiment,

we build controls into the study design itself to elim-

inate alternative causes and isolate the experimental

situation from the influence of all variables except

the main causal variable. Nonexperimental research

eliminates alternatives by identifying possible alter-

native causes and measuring them. This is common

in survey research. Once we measure potential alter-

natives, we use statistical techniques to learn whether

the causal variable or something else operates on the

effect variable.

4.Specifying the mechanism in a causal rela-

tionshipmeans that when we create a causal expla-

nation, we must have more than two variables that

are correlated, which is “a satisfactory explanation

requires that we also specify the social ‘cogs and

wheels’” (Hedstrom and Swedberg, 1998:7). We go

beyond saying that an independent and dependent

variable are linked, as if the connection were through

a “black box” of unknown processes. A full causal

explanation identifies a causal relationship and

specifies a causal mechanism.

Let us say we find a strong association between

a person’s social class and her health. We may state

our “theory” as high-class people live longer and

get sick less often than low-class people. However,

it is not enough to say that a person’s social class

causes health outcomes. We must also explain why

Causal mechanism The part of a causal explanation

that specifies the process by which the primary inde-

pendent variable(s) influence the primary dependent

variable(s).

Lower Income Upper Income

FIGURE 2 Association of Income and Race