Dear Dieter, 

Because testing with monofilaments and its resulting "score" have not a continuous nature, the descriptive and inferential statistics have to be done considering a categorical classification - variable with ordinal nature. The patient feels or do not feel A or B or C or D.... monofilament, then he is classified with categories mutually exclusive. In the case you described, if you have 4 monofilaments and he does not feel the last one, then he could be classified as a fifth category. I hope I have helped you!

Best Regards,

Isabel

Dear Dieter,

I completely agree with Isabel's answer, it depends on which kind of  descriptive statistics you are thinking about.. do you need to treat your 'data' as a value (phisical measurement)  or as a piece of information (patients assigned to categories)?

Dieter,

I would also agree with what has been said.  Because of the ordinal nature of the monofilaments, median scores are generally reported and you could still do this by creating a fifth category as Isabel stated.  You could also calculate median scores on those that felt one of the monofilaments and seperately report the % of patients that did not feel the strongest monofilament.  The best way to report the data will depend on your research/clinical question.  I hope this helps and best of luck

Erik

For patients with a missing response to the 6.65 filament are defined as absolute loss of protective superficial sensation and you do not need any bigger monofilament because the thus patients usually have  deep pressure sensation only (fifth category).

Dear Bel, Francesco, Erik & Lauren

Thank you all for providing your answers and offering your help. I agree that the median is probably the best way to report results in this case.

However, we used a set of 20 monofilaments and since they are calibrated to the level of force that they apply it was an attempt to quantify the sensory threshold. If I reduce this to (only) categorical data I am afraid that I might loose this quantitative information.

But we will try to proceed in the direction you all proposed!

Thanks again,

Dieter...