BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//talks.cam.ac.uk//v3//EN
BEGIN:VTIMEZONE
TZID:Europe/London
BEGIN:DAYLIGHT
TZOFFSETFROM:+0000
TZOFFSETTO:+0100
TZNAME:BST
DTSTART:19700329T010000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0100
TZOFFSETTO:+0000
TZNAME:GMT
DTSTART:19701025T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
CATEGORIES:Statistics
SUMMARY:Simultaneous Directional Inference - Ruth Heller (
Tel-Aviv University)
DTSTART;TZID=Europe/London:20230602T140000
DTEND;TZID=Europe/London:20230602T150000
UID:TALK199495AThttp://talks.cam.ac.uk
URL:http://talks.cam.ac.uk/talk/index/199495
DESCRIPTION:Joint work with Aldo Solari\n\nWe consider the pro
blem of inference on the signs of n>1 parameters.
Within a simultaneous inference framework\, we aim
to: identify as many of the signs of the individu
al parameters as possible\; provide confidence bou
nds on the number of positive (or negative) parame
ters on subsets of interest. Our suggestion is as
follows: start by using the data to select the dir
ection of the hypothesis test for each parameter\;
then\, adjust the one-sided p-values for the sele
ction\, and use them for simultaneous inference on
the selected n one-sided hypotheses.\n\nThe adjus
tment is straightforward assuming that the one-sid
ed p-values are conditionally valid and mutually i
ndependent. Such assumptions are commonly satisfie
d in a meta-analysis\, and we can apply our approa
ch following a test of the global null hypothesis
that all parameters are zero\, or of the hypothesi
s of no qualitative interaction. We consider the u
se of two multiple testing principles: closed test
ing and partitioning. The novel procedure based on
partitioning is more powerful\, but slightly less
informative: it only infers on positive and non-p
ositive signs. The procedure takes at most a polyn
omial time\, and we show its usefulness on a subgr
oup analysis of a medical intervention\, and on a
meta-analysis of an educational intervention. \n\n
The relevant paper is arXiv:2301.01653
LOCATION:MR11/B1.39\, Centre for Mathematical Sciences
CONTACT:Qingyuan Zhao
END:VEVENT
END:VCALENDAR