File Exchange > DataAnalysis >    Post-hoc Analysis for Nonparametric Tests

Author:
OriginLab Technical Support
Date Added:
8/8/2017
Last Update:
9/19/2019
Downloads (90 Days):
792
Total Ratings:
6
File Size:
77 KB
Average Rating:
File Name:
npposthoc.opx
File Version:
1.1
Minimum Versions:
License:
Free
Summary:

Perform multiple comparisons for Kruskal-Wallis ANOVA, Moods Median Test and Friedman ANOVA.

Screen Shot and Video:
Description:

  • Purpose
    This app can be used to perform multiple comparisons to decide which two groups are significantly different from each other, provided that nonparametric tests indicate the populations are significantly different.
  • Installation
    Download the file "npposthoc.opx", and then drag-and-drop onto the Origin workspace. An icon will appear in the Apps gallery window.
  • Operation
  1. Click the icon in the Apps Gallery window to open the dialog.
  2. Select a test from Kruskal-Wallis ANOVA, Mood's Median Test and Friedman ANOVA.
  3. For Kruskal-Wallis ANOVA and Mood's Median Test, Dunn's Test is used as the default post-hoc analysis method. 
    For Friedman ANOVA, you can choose Dunn's Test or Wilcoxon-Nemenyi-McDonald-Thompson Test as the post-hoc analysis method.
  4. Specify Input Data Form and select data from worksheet.
    Specify Significance Level.
  5. Click OK.

Updates:

v1.1: Add paired comparison plot.

Reviews and Comments:
06/13/2019OriginLabHi ktrand, thank you for your suggestion. We will improve it in the next version.

06/12/2019ktrandWhew, so glad I found this! I was really struggling without a non-parametric equivalent of the Tukey Test. My only critique is that interpretation of the results would be much easier if there was a graphical output, as there is with the Tukey Test under the ANOVA menu, to show which pairwise comparisons are signficantly different.

04/15/2019mohenri123It works conveniently!

08/03/2018vladanton100Thank you! this is a very simple app that fills a major gap. Post-hoc tests for non-parametric ANOVA tests should be offered by Origin by default, but this is excellent for now.

01/17/2018123qwertz123Okay, please delete my first comment, now it works perfectly with my dataset!

01/17/2018123qwertz123Does not work for larger datasets (n~2500)