Am 29.11.2017 um 22:56 schrieb Sven Schreiber:
> Am 29.11.2017 um 08:42 schrieb Sven Schreiber:
>> Am 29.11.2017 um 01:34 schrieb Allin Cottrell:


>>> We're looking for a z-statistic, and that's NA for the given sample 
>>> size. If anyone wishes to pursue tiny-sample statistics in this area, 
>>> that's fine by me.

> z = (wp < wm) ? wp : wm;
> }
> 
> where z later gets stored and returned as the test statistic.

Although I have to admit it's not good to return a totally different 
type of test statistic under the same accessor, just because N changes 
from, say, 9 to 8. Hm, not sure what the best way out of this is.

thanks,
sven

Reply via email to