Sometimes, the relationship we originally see between two variables is spurious: It is actually the result of a third variable that is associated with both our dependent and independent variables. If the third variable influences both our independent and dependent variables, it is called an antecedent variable; if our independent variable influences the third variable, which in turn influences our dependent variable, it is called an intervening variable. We need to control for any antecedent or intervening variables if we want to get an accurate picture of the influence that our independent variable has on our dependent variable. Sometimes, controlling for the third variable will leave the relationship unchanged; sometimes, it will disappear; and sometimes, it will emerge. One way to control for a third variable is with a three-way contingency table.
After reading this chapter, you should:
- Understand how measures of association can be inaccurate if we do not control for relevant factors
- Be able to identify the various ways control variables can affect a relationship
- Know how to use a three-way contingency table to control for a third variable
- Be able to create a professional looking three-way contingency table in Word
- Be able to produce a three-way contingency table with SPSS
- Know how to interpret a three-way contingency table