- There is no statistical test that validates the evaluation results for both user study and quantitative comparisons. Most comparisons are just comparing averages, which can be misleading. Use proper statistics (e.g., t-test and ANOVA) to confirm the differences are meaningful. Presenting the distributions, such as the violin plots in Fig. 10, is a great way to show the results as well. Fig 8, 11, and 12 could be improved with violin plots or histograms.