I would be grateful for some advice / references to resolve a dispute between myself and colleagues. The data consists of series of consultants assessing a patient on various critera and scoring the results as 1,2,3 or 4 (4 being "better" than 1). The current analysis involves taking the means of these scores to give an average patient response which is then subject to further analysis such as anovar. This is justified on the grounds that everybody analyses this type of data this way. I am unhappy about this and would rather use the median or mode as summary statistics for further analysis. Many thanks, Gill. _________________________________________________________________ Want to block unwanted pop-ups? Download the free MSN Toolbar now! http://toolbar.msn.co.uk/