I have conducted a survey with multiple 5 point Likert scale questions ranging from 1 = very unimportant to 5 = very important.
I used frequency distributions, and stacked bar charts to show which percentage of respondents rated the items as either important or very important. There are some great differences between the importance placed on the different items: e.g. one item was only rated (very) important by 6 % of the respondents while another one was rated (very) important by 70 % of the respondents.
To me, this difference seems very significant.
Is there a statistical test that I can run to determine whether the difference between those two t-point likert scales items is statistically significant?
