What is the right statistic to use to estimate variance in rank orders?

I want to display the results of a rank order survey question, and identify items for which there was significant disagreement (i.e. some people ranking an item high and some ranking it low). Do I just take the standard deviation of the weighted scores, or should I do something else?

What is the right statistic to use to estimate variance in rank orders?
If I understand what you are trying to do you have one question and want to see if there was a significant disagreement between groups of people, i.e., some responded to this question negatively and some responded to this question positivity. If you have any demographic data I would use this is stratify the data and then look for the factors that might influence the response. This would be done using multiple regression methods and mapping the demographic variables to dummy variables and the response to the questions as well. There are some catches to this, but for the most part it works well.





this this isn't what you are looking for you might want to look at factor plots and main effects plots.





email me if you want to discuss this further. More specific information about what you are trying to do will help me suggest methods for analysis.