ABSTRACT
If we want to describe, explain, explore, or predict some phenomenon, we must first be sure that the variables in the study are actually variables. Imagine an extreme situation in which the only people to have completed your survey were all women. The sex variable is now a constant and you could no longer use sex as a variable in explaining or predicting any other variables in the study. Hence, before any further data analysis is accomplished, it is important to do some univariate (one variable at a time) analysis and look at every item in the study to get a sense of the variability of responses. This is necessary in order to decide whether the variables can be used for additional statistical analysis.
