Variables: Independent, Dependent, Intervening As in all scientific research, the goal of research in sociology and criminal justice is to establish causal relationships between variables. In these endeavors, one variable is hypothesized to be the cause and the second variable is the effect. The term for the cause is the INDEPENDENT variable; the term for the effect is the DEPENDENT variable. [Actually, this statement is an oversimplification of the definitions of these variables, as well as the complexities of their relationships. However, for the purposes of this course, we will treat these variables as they are presented here.] Once we have a basic understanding of independent and dependent variables, a question arises: in a research hypothesis, how do we tell which is the independent variable and which is the dependent variable. Actually, there is a relatively simple way to distinguish them: The independent variable occurs prior in time to the dependent variable. Consider this example. Hypothesis: The average annual income of men is greater than the average income of women. It should be clear that the 2 variables in this hypothesis are gender (men and women) and average annual income. Which of these 2 variables comes first in time? Does a person earn an annual income and then become a man or woman? Or is a person a man or woman first and then earn an annual income? It should be obvious that gender precedes average annual income; hence gender is the independent variable and average annual income is the dependent variable. Note that, in this example, the independent variable- gender- comes first in time, but not first in the sentence. The dependent variable- average annual income- is presented first and is followed by the independent variable. Not infrequently, the sentence is presented with the independent variable first, followed by the dependent variable. The important point is that the independent variable is not necessarily the first one stated in the hypothesis, but it is the first one in time. Though not typically stated, many social science hypotheses begin with an implied expression: “Other things being equal”. In the current example, we could re-phrase the hypothesis as: “Other things being equal, the average annual income of men is greater than the average income of women”. Stated this way, the hypothesis acknowledges that there may be additional variables which might have an effect on the proposed relationship between the independent and dependent variables. These additional variables are often termed intervening variables because they come between the independent and dependent variables and may modify the relationship between them. In our hypothesized association between gender and average annual income, here are 3 possible intervening variables. 1) Job category: Many women are (still) employed in clerical positions, while men are more often in managerial positions; 2) Education: Male employees may have higher levels of education than females; 3) Seniority: Women may have had to remain out of the work force to raise children, thus delaying their entry into employment; men are less likely to have to do this. These are not the only intervening variables which might effect the relationship between gender and average annual income, but they do point to the necessity to acknowledge them and incorporate some measure(s) of them in the research design. Here is how we might diagram the relationships between these variables in our hypothesis. Independent Variable Intervening Variable(s) Dependent Variable Male Job Category- Clerical Average Annual Income Female Job Category- Clerical Average Annual Income Male Job Category- Managerial Average Annual Income Female Job Category- Managerial Average Annual Income Male Education- College Degree Average Annual Income Female Education- College Degree Average Annual Income Male Education- Less than College Average Annual Income Female Education- Less than College Average Annual Income Male Seniority- More than x Years Average Annual Income Female Seniority- More than x Years Average Annual Income Male Seniority- Less than x Years Average Annual Income Female Seniority- Less than x Years Average Annual Income The first 4 rows of this table allow us to compare the average annual incomes of males and females in clerical and managerial job categories. If male workers in both job categories earn higher average annual incomes than female workers, this confirms the original hypothesis. The next 4 rows compare the average annual incomes of male and female employees who are college graduates or not college grads. If male workers at both educational levels earn higher average annual incomes than female workers, this confirms the original hypothesis. The last 4 rows compare the average annual incomes of male and female employees who have x years of work experience (x is a number which can be derived from a literature review) and those who have less experience. If male workers at both seniority levels earn higher average incomes than females at the same levels, this confirms the original hypothesis. It should be pointed out that, if any of these comparisons show that there is no difference in average annual incomes between the genders, or that female workers earn higher average annual incomes than their male counterparts, further research is necessary to clarify the relationships between these variables.