Ranking question

I have data from a student survey in which they ranked a list of potential difficulties. There were given 13 choices, and had to rank them 1-13, with an option to not rank something if it didn't affect them. The data I have is how many times a phrase got each rank (e.g. "Trouble communicating" got ranked #1 three times, #2 once...) and the percentage each phrase got each rank.
What is the proper way to compile the data into an overall ranking, and what is the proper way to report this on a table? Some non-statistics sources said to get a raw score for each rank by multiplying the weighted rank by the number of times it was given that rank (e.g. #1 three times means 13*3) and adding up the values.
What is the "proper" research way of doing this? Is there a "proper" way?
I'm going to be presenting these results at a poster session at a research conference, so I'd like to make sure I'm on point.

P.S.- If it helps, I use Stata and SPSS



TS Contributor
Since this is ordinal data I'd use the median ranks. Perhaps with
interquartile range, in order to indicate degree of agreement.

The multiplication algorithm makes the heroic assumption that
the differences between ranks are evenly spaced.

With kind regards

Last edited: