I have a survey questionnaire data. For each question people can answer one of the five options, [Not Important at all (1) to Very important (5)] In my table I have proportions of people who chose each of these answers to each of the questions. Now I want to check whether these proportions (i.e. proportion of people who chose any of the options for each question) are statistically same or different. Can someone please help me which test can I use for this. 
$\begingroup$
$\endgroup$
4
-
$\begingroup$ What exactly is your research question? Whether the proportions are different within the questions? Whether the proportions are different within the scores? Whether the proportions are different considering all the cells? Something else? $\endgroup$– user2974951Commented Feb 20, 2020 at 8:11
-
$\begingroup$ I want to check whether the proportions are different within the question. $\endgroup$– arshadCommented Feb 20, 2020 at 8:17
-
$\begingroup$ Do you also have frequency data for the cells (not proportions)? $\endgroup$– user2974951Commented Feb 20, 2020 at 8:21
-
$\begingroup$ Yes, I have the frequency as well $\endgroup$– arshadCommented Feb 20, 2020 at 8:29
Add a comment
|
1 Answer
$\begingroup$
$\endgroup$
If testing differences in frequencies (not proportions) is of interest to you then you can use a Chi square test (or Fisher test if the assumptions are not met) to test for different frequencies (different distribution of values).
Since you will be conducting many tests you will need to correct the obtained p-values by some procedure such as Bonferroni-Holm correction.
answered Feb 20, 2020 at 9:34