hiren379 ★ India, 2013-05-16 10:25 (4378 d 21:19 ago) Posting: # 10589 Views: 3,157 |
|
Hello All, Candaian BE Guideline states that "No more than 5% of the subjects may be considered to be outliers" Now my questions are: 1) How to handle a situation where >5% subjects are detected to be outlier? 2) I have mentioned in protocol that `No more than 5% of the subjects shall be considered as outliers`. Now this leaves the reader of my protocol with an obvious doubt that what to do if >5% subjects are detected as outliers????? I need to clarify this too in the protocol right to avoid any biasness in data analysis?!!!!!! How to handle this? what to mention in protocol? ![]() ![]() ![]() Regards, Hiren |