C. Hatton Humphrey wrote: >> The problem is that there are a couple bad apples who blow through the >> surveys rating someone either all 1s or all 5s, throwing off that person's >> ratings and effectively ruining the value of the performance review. >> The surveys are all anonymous, so I can't simply go to the person entering >> the survey and tell him/her to stop cheating. >> >> Can anyone think of a way to monitor and block the cheaters? > > At the school where I went we were told that all "straight-line" > surveys would not be counted. Additionally we had to indicate that we > had completed the survey by marking a "completed" circle (these were > paper surveys, scantron style). > > You're talking more about a judgement call than a logical one.
Not at all. Outlier detection is a branch of statistics that is a genuine science with strict rules. (But also something I do not have fond memories off.) To do it quick and dirty I would look along the lines of: - calculate the averags of all entries - calculate the standard deviation of all entries - calculate the mean of all entries - calculate the number of entries that deviate from the mean Then look for entries with a low standard deviation (indicating somebody used the same answer everywhere), entries with a high standard deviation (randm) or entries with a medium/high standard deviation but a low number of entries deviating from the mean (all 1 and a few 5 or vice-versa). Add additional checks to see if all answers occur at least once etc. and when calculating the results look into presenting it graphically with a dot for the average and a line indicating the relative size of the standard deviation. Most databases have statistical functions that can help. If you want to do it right, buy a book on statistics. Jochem ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~| Discover CFTicket - The leading ColdFusion Help Desk and Trouble Ticket application http://www.houseoffusion.com/banners/view.cfm?bannerid=48 Message: http://www.houseoffusion.com/lists.cfm/link=i:4:208325 Archives: http://www.houseoffusion.com/cf_lists/threads.cfm/4 Subscription: http://www.houseoffusion.com/lists.cfm/link=s:4 Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user=89.70.4 Donations & Support: http://www.houseoffusion.com/tiny.cfm/54
