Skip to content

Create fake 2-sample Likert data, and analyze with some tests.

License

Notifications You must be signed in to change notification settings

zimolzak/fun-with-likert

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

43 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Fun with Likert

Make some fake unpaired Likert data, plot it, and analyze it with some tests. This pretends that you have two groups (A and B), with 60 people in each group, and your survey has 7 questions, each scored on a 0 to 4 Likert scale.

Results

Below is a table of p values, comparing group A to group B, for each of the questions. The following statistical tests (columns in the table) are tried:

  1. Wilcoxon rank sum, a.k.a. Mann-Whitney
  2. T-test (unpaired)
  3. Chi-squared test for trend in proportions (Cochran–Armitage test for trend)
  4. Pearson's chi-squared test ("plain old" chi-squared that doesn't know ordinal from categorical)

Some authors say that the t-test might be more appropriate than it seems. See Norman and Sullivan.

  question wilcoxon_p    ttest_p     catt_p      chi_p
         1 0.00008602 0.00001416 0.00002332 0.00001884
         2 0.00000000 0.00000000 0.00000000 0.00000003
         3 0.00000000 0.00000000 0.00000001 0.00000028
         4 0.00090077 0.00037737 0.00045219 0.00241743
         5 0.00327405 0.00119031 0.00131765 0.01453107
         6 0.61915745 0.66912503 0.66595722 0.95185781
         7 0.07541019 0.09881224 0.09722275 0.10866653

Plot of fake survey data (one question per row):

plot of likert data

About

Create fake 2-sample Likert data, and analyze with some tests.

Topics

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages