Independent and dependent variables[1]

To figure out which direction to percentage a table is not always straightforward. It rests on the distinction between the independent variable and the dependent variable, and invokes an implicit causal logic.

Statements about association are usually stated in terms of a relationship between an independent and a dependent variable. The idea is that one variable is the effect of another variable or, to say it another way, that one variable precedes and/or causes another.

The dependent variable is the variable to be explained (the ‘effect”).

The independent variable is the variable expected to account for (the “cause” of)             the dependent variable.

Cause-and-effect relationships are not always easy to infer in the social sciences. In order to establish that two variables are causally related, you need to meet three conditions.

(1)   The cause has to precede the effect in time. This is often a good clue about which is the independent and which is the dependent variable. For example, it is difficult to imagine childbirth preceding conception, or achieving a college degree before you get a high school diploma.

(2)   There has to be an empirical relationship between the cause and the effect; that is, an observable pattern of covariation between the variables.

(3)   The relationship cannot be explained by other factors. So, for example, there is an observable relationship between daily temperature and the incidence of drowning, but high temperatures do not CAUSE drowning. What do you think the missing variable is?

Because of the limitations in inferring cause-and-effect relationships in the social sciences, be cautious about using the terms cause and effect when examining relationships between variables. However, using the terms independent and dependent variables is still appropriate even when this relationship is not articulated in terms of direct cause and effect. Here are a few guidelines that may help you to identify the independent and dependent variables:

(1)    The dependent variable is always the property you are trying to explain; it is always the object of the research.

(2)    The independent variable usually occurs earlier in time than the dependent variable.

(3)    The independent variable is often seen influencing, directly or indirectly, the dependent variable.

[1] This discussion is taken from Chava Frankfort-Nachmias and Anna Leon-Guerrero. 2000. Social Statistics for a Diverse Society, 2nd ed. Thousand Oaks, CA: Sage Publications, pp. 9-10.