Because quantitative methodology requires measurement, the concepts being investigated need to be defined in a way that can be measured. Leadership style, reading comprehension, or learning style are concepts but they cannot be measured as such. Types of leadership style, reading comprehension scores, or learning style preferences can be measured. They are variables (concepts that can vary).
Quantitative research involves many kinds of variables. There are four main types:
Each is discussed below.
Independent variables (IV) are those that are suspected of being the cause in a causal relationship. If you are asking a cause and effect question, your IV will be the variable (or variables if more than one) that you suspect causes the effect.
There are two sorts of IV, active independent variables and attribute independent variables:
Independent variables are frequently called different things depending on the nature of the research question. In predictive questions where a variable is thought to predict another but it is not yet appropriate to ask whether it causes the other, the IV is usually called a predictor variable rather than an independent variable.
Dependent variables are those that are changed in some way by the independent variables. If you ask, "Does A cause [or predict or influence or affect, and so on] B?," then B is the dependent variable (DV).
In questions where full causation is not assumed, such as a predictive question, the dependent variables are usually called outcome variables, and the independent variables are usually called the predictor variables.
In some studies, some characteristic of the participants must be measured for some reason, but that characteristic is not the IV or the DV. In this case, these are called sample variables. For example, suppose you are investigating whether amount of time on task affects level of comprehension in visual learners. In order to obtain a sample of visual learners, a learning style preferences inventory will be given. So the presence or absence of visual learning style preference will be a sample variable. That score is not used as an IV or a DV, but simply to get the appropriate people into the sample.
When there is no measure of a characteristic of the participants, the characteristic is called a "sample characteristic." When the characteristic must be measured, it is called a "sample variable."
Extraneous variables are not of interest to the study but may influence the dependent variable. For this reason, most quantitative studies attempt to control extraneous variables. The literature should inform you what extraneous variables to account for. For example, in the study of third graders' reading scores, such variables as noise levels in the testing room, the size or lighting or temperature of the room, and whether the children had had a good breakfast, all might be extraneous variables.
There is a special class of extraneous variables called confounding variables. These are variables that can cause the effect we are looking for if they are not controlled for, resulting in a false finding that the IV is effective when it is not. In a study of changes in skill levels in a group of workers after a training program, if the follow-up measure is taken relatively late after the training, the simple effect of practicing the skills might explain improved scores, and the training might be mistakenly thought to be successful when it was not.
There are many details about variables not covered in this handout. Please consult any text on research methods for a more comprehensive review.
Doc. reference: phd_t2_soe_u04s2_h03_quantvar.html