Cause and Effect

In some cases, the relationship between two items seems clear and intuitive, like when there’s a sports game, people will order pizza even though it’s getting late. One member of the group swears that eating several slices before retiring for the evening gives him nightmares.

Now, the only way to answer this question "Does eating pizza right before going to sleep cause someone to have more nightmares?" is to design an experiment.

In one type of experimental design, a cause-and-effect relationship, a researcher can determine whether manipulating an independent variable—in this case, eating pizza before bedtime—causes a particular effect—changes in the dependent variable, the number of nightmares that occur throughout the night.

They could assign half of the participants to the experimental group, which is given the experimental manipulation—the assignment of eating three slices of pizza right before they go to bed—and the second half to the control group, which is instructed not to eat anything.

They can also take certain measures to control for confounding variables that may produce alternative explanations.

For example, by randomly assigning participants to different groups—using a probability-based method—the researcher can ensure that participants are equally matched on potential confounds, for instance, their history of and predisposition for nightmares, the time it takes them to fall asleep, and even their overall quality of sleep.

In addition, the researcher could conduct the experiment in a laboratory setting where even more factors can be controlled for, like the sleep environment and what they watch before sleeping.

In the end, the only way to establish causality between two variables is by running sound experiments!