Are there other statistical tests available to compare paired sets of data e.g. {1, 2, 2, 3, 4} v's {3, 3, 3, 2, 5} for similarity, aside from the paired Wilcoxon and paired Student t-test?
Asked
Active
Viewed 47 times
0
-
There is discussion of several alternative tests in the answers to [A principled method for choosing between t test or non-parametric e.g. Wilcoxon in small samples](http://stats.stackexchange.com/questions/121852/a-principled-method-for-choosing-between-t-test-or-non-parametric-e-g-wilcoxon) – Silverfish Feb 10 '15 at 10:23
-
(Though several of the suggestions in that thread are for unpaired data rather than paired.) – Silverfish Feb 10 '15 at 10:30
-
How do you want to compare the data? The two you mention evaluate the medians and means (respectively) of the populations from which the data differences are presumed drawn, but there are infinitely many other properties on which a comparison could be based--and therefore infinitely many possible statistical tests. – whuber Feb 10 '15 at 14:19
-
I am looking to evaluate the similarity in frequency between model A and model B output data streams. To date, I have used paired t-tests, to determine if data (errors made by a robot arm) come from different populations and so to estimate how similar they are (i.e. if p > 0.05, they are seen as statistically similar). – alan Feb 11 '15 at 08:31