
172 Data Analysis and Statistics for Geography, Environmental Science, and Engineering
Alternatively, at the Rcmdr you can use Graphs|Boxplot, pick moist variable and select iden-
tify and plot by groups you will get a second dialog window, select loc and then once you get
graphics can use mouse to identify outliers. Then, use Statistics|Non parametric test|Kruskal-
Wallis and set variable a factor in the dialog box (Figure 5.16).
5.6.13 anova nonparaMetric: frieDMan
Next, we will run the Friedman test on the same set of yield by treatment and variety. Recall that
the parametric indicated very low p-value for treatment.
> friedman.test(yield, treat, ...