0 Replies Latest reply on Aug 13, 2019 5:53 AM by Dror Katz

    How to generate significance tests on combinations of groups?

    Dror Katz



      For concepts comparison, I have data sample from a survey, for only 1 scaled question in this case, answers could be 1 to 5, question is being asked for each concept (A-E).

      I would like to create significance tests for each answer (1 to 5) and for each combination from pairs of concepts.

      Below is a table summarised the proportion of respondents answered specific answer (1-5) for each concept:


      ie, 18% answered 1 for concept A, while 19.78% answered 2 for concept A.

      the comparison is between concepts and what I would like to achieve is for each answer, a significant test between all pairs (A vs B, A vs C, A vs D, A vs E, B vs C etc.),

      The results should follow this example:

      Let's say for answer 1 -  C is significantly higher than A and B, then the result should the tuple (A,B) to be listed under or by side the value of the cell for answer 1 and column C (20.12%), as below:



      So I need to create a new column with results for each pair and answer. Any idea how can I implement it in Tableau? I looked to use tabpy but didn't managed to solve the combinations challenge.

      workbook attached.

      (I'm not looking for which significant test to run, just how it can be done with multiple combinations of groups to compare.)


      Thanks in advance,