Univariate statistics are measures that refer to a single characteristic. These primarily include measures that characterise the central position of a distribution (mean values) and describe the dispersion of the data.
Univariate analysis refers to any situation in which the scores to be analyzed (whether via descriptive, inferential, or both method families) are collected on a single dependent variable.
The specific analytic method depends on: (a) the specific research question(s) to be answered; (b) how scores are quantified on the lone DV; (c) any requisite distributional assumptions (for inferential methods); and (d) how the data were collected.
Mutlivariate analysis refers to instances in which scores to be analyzed are collected on more than one dependent variable.