Variables, Independent

Independent variables influence or cause variation in another variable (referred to as the dependent variable). Independent variables are isolated, manipulated, and controlled by a researcher to see how the manipulation affects the dependent variable. Independent variables themselves are not affected or changed by other variables; their existence is not dependent on other variables in the proposed relationship. Also referred to as predictor variable, the independent variable is an essential component of quantitative research designs, as it is the catalyst for the measurable changes in a research study. This entry describes qualities of independent–dependent variable relationships, provides basic examples, and lists common statistical approaches for variable relationship analysis.

To test study predictions, a researcher looks to see if a dependent variable changes based on the presence or absence of an independent variable. Most quantitative research studies seek to determine if an independent variable influences, either directly or indirectly, the dependent variable. Independent variables often occur earlier in time than dependent variables and are generally represented graphically on the x-axis. The researcher’s goal is to accurately predict how the dependent variable will change in the presence of the independent variable. To do so, the researcher must find a significant relationship between independent and dependent variables, represented in the form of correlation (connection) or cause–effect claims.

Commonly cited examples of independent variables include age, sex, race, education, and income; these things do not change. For instance, a researcher might look to see if age influences the extent to which an individual uses social media applications; in this case, the independent variable of age is proposed to have an effect on the dependent variable of social media use. Age is isolated by the researcher as a predictor of social media use behaviors, which may vary. A second example could use sex as a predictor variable; a researcher could isolate sex as an independent variable to determine the difference between levels of self-disclosure in men and women. It is important to remember that a dependent variable in one study can serve as an independent variable in other research studies. For example, a researcher could consider how self-disclosure influences feelings of closeness in a relationship. In this case, self-disclosure is the independent variable that predicts how close individuals feel in a relationship. Research design determines which variables are independent and dependent.

To measure relationships between independent and dependent variables, researchers must determine appropriate statistical tests. Researchers [Page 1846]looking for relationships between one independent and one dependent variable, such as the age and social media example, have a number of options for testing the significance of the relationship. Some research designs are more complex and feature one or more independent or dependent variables. For example, some research designs call for a measure of significant differences between the effect of two categorical independent variables and one categorical dependent variable, so a researcher might use a t-test. Researchers often use analysis of variance (ANOVA) procedures, such as one-way ANOVA, to analyze one categorical independent variable among three or more groups with one continuous dependent variable. Research designs with two or more categorical independent variables are analyzed with a factorial ANOVA, and to determine how one or more independent variables influence one or more dependent variables, multivariate ANOVA (MANOVA) is utilized. Finally, a researcher looking to estimate how much of the variance in a dependent variable is accounted for by variance in an independent variable would use multiple regression. There are numerous statistical tests a researcher can use to determine the relationship between independent and dependent variables; the appropriate test depends on the specifics of the research design.

...