An important step in designing all quantitative research projects is defining or identifying the variables that will be manipulated, measured, described, or controlled. Although qualitative researchers do not define variables to the same extent that quantitative researchers do, they still must outline what kinds of phenomena they are studying. The major types of variables, or phenomena of interest, are described briefly here, with common examples from service-learning research provided. These are presented in terms of labels from the quantitative research approach, but the qualitative tradition includes analogous examples.
- Independent Variable (IV): A variable that is selected or controlled by the researcher, to determine its relationship to the observed outcome of the research—also called explanatory, predictor, or manipulated variable. A common example is whether or not a course section involves service-learning pedagogy. The nature of what is varied should be carefully described so that the attributes of the different interventions or experiences are clear.
- Dependent Variable (DV): The variable being measured as an outcome—also called outcome, response, criterion, or explained variable. Many examples of dependent variables (variables of interest) are presented in Table 2.
- Intervening (Mediating) Variables: a hypothetical concept that attempts to explain the relationship between the independent and dependent variables (Baron & Kenny, 1986). Mediating variables, also called process variables, explore why the independent variable is linked to the dependent variable. For example, this might be a concept such as altruism or social responsibility that is presumed to explain why a service-learning course influenced subsequent volunteer behavior. There are statistical methods for evaluating the role of a mediating variable (see David A. Kenny's page on mediation: davidakenny.net/cm/mediate.htm).
- Moderator Variable: A variable that is related to the direction or strength of the relationship between the independent and dependent variables (Baron & Kenny, 1986). A moderator variable may be qualitative (such as student gender, type of community organization, or type of college) or quantitative (e.g., number of service visits). In addition, it may be related to the strength or the direction of a correlation, or it may interact with the independent variable and the dependent variable. In either case, a moderator variable describes an "it depends" relationship (e.g., the strength of the correlation between two variables depends on the past volunteer experience of the student). Generally, moderator variables are variables that exist prior to data collection, as opposed to mediating variables that are assumed to occur during the phenomena being studied.