4

I have two sets of spatial data which underwent a transformation. I'd like to compare the effect of the transformation on two sets of data to test the hypothesis that the transformation had a larger effect on one set of data. I planned to use a Procrustes test to evaluate the effect of the transformation. The program I'm using returns a disparity statistic.

Is there a way to compare the two disparity statistics?

kjetil b halvorsen
  • 63,378
  • 26
  • 142
  • 467
Justine
  • 41
  • 1
  • Are these two transformations on the same original "true" set of spatial coordinates? Another analysis may be bidimensional regression - see [Tobler, 1994](http://onlinelibrary.wiley.com/doi/10.1111/j.1538-4632.1994.tb00320.x/pdf). There you could compared the sets of coefficients. – Andy W Jun 17 '15 at 11:50

1 Answers1

1

What you could do is use a PROcrustean Randomization TEST (PROTEST), which makes use of permutations in order to assign a significance to the match of your in this case spatial datasets.

So what you could do is run two PROTEST-analyses, one with and one without transformation and see which scenario returns the highest statistical significance.

An alternative might be to use the Mantel-test, which already assigns a p-value to your match. Of course, the Mantel test also has its drawbacks (see for example here. )

Archie
  • 572
  • 6
  • 16