An **error plot** is appropriate for this sort of data, and is used in the example report. It shows the mean and variance of the samples being compared, and is a clear way of demonstrating the degree of overlap between samples.

*Examples of error plots comparing two samples with differing degrees of overlap*

As part of the check that data satisfy the parametric criteria, it may be appropriate to plot out the data for each sample as a **histogram** to assess the closeness of the distribution to a normal distribution. This would not apply to small samples.

*Example histogram*