InvestorsHub Logo
Followers 12
Posts 801
Boards Moderated 0
Alias Born 02/09/2013

Re: Protector post# 183895

Wednesday, 07/23/2014 7:28:22 AM

Wednesday, July 23, 2014 7:28:22 AM

Post# of 346043
Personally, I doubt that it would be stopped so soon. It is important to keep in mind that the more times that a data monitoring committee plans to look at the data to make stop/keep going decision, the higher the target p-value has to be to count as a success.

As an example, suppose that you are testing avastin against avastin (in other words, there is no difference between the control and experimental group). If your target is a p<.05 then every time you look at the data and run the numbers there is a 1 in 20 chance of finding that the experimental group is stat sig better than the control. So suppose you plan to look at the data 20 times while it is accruing, then you are virtually guaranteed to "discover" that Avasting is better than itself and that this superiority meets the threshold for stat sig.

This is why when conducting multiple interim looks, it is essential to use a more stringent threshold than .05 for statistical significance.

Volume:
Day Range:
Bid:
Ask:
Last Trade Time:
Total Trades:
  • 1D
  • 1M
  • 3M
  • 6M
  • 1Y
  • 5Y
Recent CDMO News