I think there is a type of chi-square analysis that takes into account rank order levels of one of the categorical variables. Can anyone remember what this is called? We want to use it in an analysis in which the response variable is in abundance categories (e.g., 0, 1-10, 11-25..., etc.) and we want to compare two different treatments. For these data, we don't think it would be appropriate to just use the midpoint of the range.
Mark Dixon
