Descriptive and causal studies answer fundamentally different kinds of questions. Descriptive studies are designed primarily to describe what is going on or what exists. Causal studies, which are also known as “experimental studies,” are designed to determine whether one or more variables causes or affects the value of other variables.

A causal study’s hypothesis is directional -- it does not simply claim that two or more variables are related, but predicts that one variable or set of variables, called “independent variables,” will affect another variable or set of variables, known as “dependent variables,” in a certain way. An example of a directional hypothesis would be, “I predict that increased levels of exercise will lead to weight loss.” A non-directional hypothesis, which would be suitable for a descriptive study, would simply predict that there exists some relationship between the variables “amount of exercise” and “weight loss.”

In a causal study, researchers manipulate the set of independent variables to determine their effect, if any, on dependent variables. Researchers in causal studies also typically make use of a “control” -- a case in which the independent variables have not been manipulated, to allow researchers to compare the effects of manipulating the independent variables to the effects of leaving them the same. A descriptive study does not typically involve variable manipulation or a control.

Descriptive studies make use of two primary sorts of data collection: cross-sectional studies and longitudinal studies. The cross-sectional study attempts to give a snapshot of data at a certain moment in time -- variables in a cross-sectional study are measured only once. The longitudinal study, on the other hand, involves a fixed, relatively stable sample measured repeatedly over time. In both cases, methods used might include mail, online or in-person surveys or interviews.

Case studies likewise make use of two primary sorts of data collection: laboratory experiments and field experiments. Laboratory experiments are conducted in artificial environments which allow researchers to carefully control exactly which variables are manipulated while keeping other factors constant. Field experiments are conducted “in the field,” in a natural or realistic environment. Field experiments allow researchers to test how their hypotheses apply to the “real world.” However, it is often impossible for researchers to control for all possible variables in field experiments, making it harder for researchers to say with confidence exactly what produced a given effect.