A Data Science Central Community
© 2021 TechTarget, Inc.
Powered by
Badges | Report an Issue | Privacy Policy | Terms of Service
Most Popular Content on DSC
To not miss this type of content in the future, subscribe to our newsletter.
Other popular resources
Archives: 2008-2014 | 2015-2016 | 2017-2019 | Book 1 | Book 2 | More
Most popular articles
Comment Wall (27 comments)
You need to be a member of AnalyticBridge to add comments!
Join AnalyticBridge
Hi
It's my proud to join such a great group and meet professional ones.
Recently I have finished "The Role of Statistics in business and science" book, from Gerald J. Hahn and Necip Doganaksoy from Wiley series in probability and statistics.
This book gives me a lot of ideas for working with statistics. And as a student of statistics, it is really enjoyable and interesting.So I think it can be a good source for those who want know more about statistics roles in other science.
Thanks and Regards
Mohammad
Thanks Stephan! I look forward to learning from the knowledge and experience this group has to offer.
Chris
Regards,
Cliff
Good to hear that you're from Sweden, we're underrepresented here!
I currently don't have any collaborations with anyone in the US. Looking forward to online collaborations and networking within the AnalyticBridge network.
Tomas
No, I haven't written anything that has been published on the subject. But I just have a way of explaining things to people that they seem to understand. I'll give an example. Power. It confuses people to no end. Many people think power = type II error. But we know it is 1-beta. When I tell people about power I tell them to think of power as a flashlight, and that significance is hiding in the corner, but he is there. Power can be likened to how bright the flashlight is to see the significance that is truly in the room. Then when they get that idea going I can get a bit more technical...
Another thing is repeated testing and increased Type I error. I tell clients that if they do lots of tests on the same sample they inflate the family wise error rate...and their eyes glaze over, lol. So I can tell them something like, "If you look for a needle in the haystack long enough and hard enough you'll find it, but it is still just a tiny needle.
So, things like this. I keep attempting to start a newsletter for my clients...just so busy!! Maybe it will slow down this summer and I can put more things together.
Thanks for the welcome.
On your question about adaptive designs and the FDA, I've found in general the FDA is cautiously accepting of adaptive designs. (CDRH more warmly than CDER/CBER.) Trials using these designs need to have a lot of detail spelled out up front (e.g. alpha spending functions, expected sample size) as well as some assurance that the safety database will be large enough to accurately assess risk-benefit (a hard problem). In addition, some divisions (oncology, neurology) are more accepting than others. Bob O'Neill and Sue Jane Wang are quite familiar with the work and tend to show up to meetings with the less conventional approaches. On one occasion, I've seen the review team recommend an adaptive design over a more conventional design, but that's rare since that usually requires the medical reviewer or team leader to be familiar and comfortable with the methodology.
How do you like NC?
Thank you for adding me as a member of Statistical Consulting group.
Regards,
Medha.
View All Comments