2

I have a interesting problem, where patients are categorized in three groups, namely Severe, Moderate, and Mild and they are measured over two time slices. Each particular cell represents proportion of patients with each category.

            | Severe | Moderate | Mild |
------------+--------+----------+------|
Timepoint 1 |   18.8 |      4.3 |  9.7 |
------------+--------+----------+------|
Timepoint 2 |   23.8 |      4.5 |  4.3 |

Now I want to draw a conclusion if the proportions in Timepoint 1 are significantly different from proportions in Timepoint 2.

Any suggestions are welcome.

Andrej
  • 2,131
  • 2
  • 18
  • 26
  • 1
    Possibly relevant: https://stats.stackexchange.com/questions/205712/chi-squared-for-multiple-samples & https://stats.stackexchange.com/questions/158654/possibility-to-run-multiple-chi-square-tests-on-2x5-table – mkt Jul 18 '17 at 08:41
  • A nice and ticklish question. I think you may try chi square test. –  Jul 23 '17 at 11:40

0 Answers0