SMS scnews item created by Sean Gardiner at Wed 24 Sep 2014 0556
Type: Seminar
Distribution: World
Expiry: 26 Sep 2014
Calendar1: 25 Sep 2014 1300-1400
CalLoc1: New Law 104
Auth: seangard@cpe-124-189-70-169.wkcz1.ken.bigpond.net.au (sgar9702) in SMS-WASM

SUMS: Wormell -- Minding your p’s and q’s: adventures in statistical inference

This week the SUMS talk is being given by undergraduate John Wormell.  

Abstract: Hypothesis testing using p values is the standard method of determining
statistical significance in science.  By themselves, p values are not well suited for
multiple hypothesis testing.  They are inflexible and somewhat opaque, and the standard
correction is often too conservative for many purposes.  The false discovery rate and
its associated q values provide an adaptive and surprisingly informative alternative
method of analysing the significance of data.  Methods based on the false discovery rate
have been widely used, especially in genomics, in the last twenty years.  This talk
presents some ideas from statistical inference for non-statisticians with some fun
examples.