To better answer such questions, I usually recommend that step 1 in an analysis to do an analysis of relative warps (PCA). A real outlier (caused, for example, digitizing landmarks out of order on a specimen) should then stick out in the scatter plot. If one specimen was reflected then the point will stick way out. If the plot just looks like a noisy scatter then you have to ask whether that is “real” (that specimen really does have a different shape or could be a small problem in digitizing).
---------------------- F. James Rohlf, Distinguishedl Professor Emeritus, Stony Brook University The much revised 4th editions of Biometry and Statistical Tables are now available: http://www.whfreeman.com/Catalog/product/biometry-fourthedition-sokal http://www.whfreeman.com/Catalog/product/statisticaltables-fourthedition-rohlf Please consider the environment before printing this email From: Zakia [mailto:[email protected]] Sent: Wednesday, May 27, 2015 3:23 PM To: [email protected] Subject: [MORPHMET] Need help,Outlier analysis in MorphoJ Hi All ! I am a novice in geometric morphometrics methods and I need some aid if it possible. I study morphology of leaves in one specie and eventually her hybrid. 30 trees was selected in two sites and 20 leaves per tree was harvested. the landmarks was digitised with tpsDig2 with repetition. I analysed my data with tpsSmall, the results show of correlation = 0.999 and SM root error = 0.0007. When I analysed data in MorphoJ, after procrust superposition, in “Find Outliers” analysis shown on the left, the maximum was 75.32 and the minimum 6.98. I don’t know what the significance? does that mean there is error in the digitization? or there is high variability? should I remove the great values? thanks for your aid C. Zakia -- MORPHMET may be accessed via its webpage at http://www.morphometrics.org To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. -- MORPHMET may be accessed via its webpage at http://www.morphometrics.org To unsubscribe from this group and stop receiving emails from it, send an email to [email protected].
