Univariate Data vs Bivariate Data
Both terms below come up in the same part of the course, and students mix them up. Here is each one defined on its own, side by side, so you can see where they part company.
Univariate data
Variables and data types
Univariate data records one variable per observational unit, so the analysis describes a single distribution rather than a relationship.
Univariate data is one column of values, one per individual, which is why the questions you ask of it are shape, center, spread, and unusual values. For example, the five quiz scores 6, 7, 8, 9, and 10 are univariate: the mean (x-bar) is , and there is no second variable to relate the scores to. Dotplots, histograms, stemplots, and boxplots are all univariate displays. Record a second measurement on the same individuals and you have bivariate data, where questions about relationships become possible.
Bivariate data
Variables and data types
Bivariate data records two variables on each observational unit, so you can study the relationship between them and not only each one alone.
Bivariate data is a set of ordered pairs collected from the same individuals, which is what makes a scatterplot possible. For example, hours studied and points scored for three students, , , and , fall exactly on the line , since every extra 2 hours adds 4 points. Both variables can be categorical instead, in which case the pairs are summarized in a two-way table rather than plotted. The whole point of bivariate data is association: you are asking how one variable moves with the other.