I have a small study of 21 respondents, who rated some items from 0-5, with 0 being unimportant and 5 being very important. I know beforehand that the items are related (they are a published list of items from a document); and I do not want measures which show how related these items are. The list of items is fixed.
Is there some non-parametric way of showing some form of agreement in the ratings between the respondents? I've seen a study which uses Kendall W, but there are a lot of ties and perhaps this is not the most useful way.
Just to clarify, I'm looking for some form of agrement measure for each item within the data, not specifically one item.