There is a minor issue on the Correlations dialog. Under "Method" the names are given the actual names of the method (Pearson, Kendall, Spearman) rather than what the methods are that they do - as we have done elsewhere in R-Instat
Should this be changed to be more consistent with the software? If so, what should we change these to?
Pearson assumes parametric data so perhaps should be based around that.
And from what I can see, Kendall is for continous and ordinal variables, and Spearmann is for ratio, interval, and ordinal variables.
There is a minor issue on the Correlations dialog. Under "Method" the names are given the actual names of the method (Pearson, Kendall, Spearman) rather than what the methods are that they do - as we have done elsewhere in R-Instat
Should this be changed to be more consistent with the software? If so, what should we change these to?
Pearson assumes parametric data so perhaps should be based around that. And from what I can see, Kendall is for continous and ordinal variables, and Spearmann is for ratio, interval, and ordinal variables.