Statistical Advisor, How To Compare Variances in Two Groups
Use one of the following:
BASIC STATISTICS: This chapter describes the so-called t-test for independent samples, comparing the means in two groups. The test assumes that the data in the two groups are normally distributed. If the resultant t value is statistically significant, then one can conclude that the means in the two groups are different (that is, in the two populations from which the observations where sampled). Most software results for the t-test will also include the F-test for the comparison of the variances in the two groups; if statistically significant, one can conclude that the variances (variability) in the two groups are different. Some software packages also include a probability calculator that allows you to compare a single mean against any hypothesized value.
NONPARAMETRICS & DISTRIBUTIONS: This chapter describes numerous so-called nonparametric tests for comparing groups. These tests should be used if one is not sure of the distribution of the variables used in the comparison. For example, the data may actually consist of rankings rather than precise measurements.
GRAPHICAL ANALYTIC TECHNIQUES: Graphical analytic techniques include histograms, line graphs, scatterplots, etc. by groups. These graphs offer a wide variety of methods to visualize differences between groups.