Descriptive statistics and graphing

This learning resources explains descriptive statistics and graphical techniques for the first phase of dealing with a dataset (after data screening).

In order to describe and understand a variable, descriptive statistics and graphing are recommended. Univariate descriptive statistics and graphs should also be used prior to proceeding with bivariate and multivariate analyses.

The main descriptive statistics of interest are the sample size, central tendency (e.g., mean, mode, median), variation (e.g., variance, standard deviation, range, percentiles), shape of response distributions (e.g., modality, skewness, kurtosis), and also frequencies and percentages.

Levels of measurement is key to determining what kinds of descriptive statistics and graphical techniques are appropriate for a given variable. Commonly used univariate graphs include bar charts and histograms, but also includes boxplots and stem and leaf plots.

Topics

 * Descriptive statistics
 * Graphing
 * Levels of measurement

Readings

 * Presenting data (Statistics Glossary)

Lecture slides

 * Descriptives and graphing (James Neill, slideshare)