BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:Statistical Significance and p-values
DTSTART;VALUE=DATE-TIME:20220203T130000Z
DTEND;VALUE=DATE-TIME:20220203T143000Z
DTSTAMP;VALUE=DATE-TIME:20220817T105529Z
UID:indico-contribution-22-244@conferences.enbis.org
DESCRIPTION:Speakers: Bernard Francq (GSK)\, Daniël Lakens (Eindhoven Uni
versity of Technology\, Netherlands)\, Ron Kenett (KPA Group and Samuel Ne
aman Institute\, Technion\, Israel)\, Stephen Senn (Statistical Consultant
\, Edinburgh\, Scotland\, United Kingdom)\nThis meeting is organized to pr
esent and discuss the issues listed in the title. It consists of 3 shorts
presentations and discussions of recognized experts. The objective is to b
oth\, provide an introduction and a review of a topic with current signifi
cant impact of the role of statistics in healthcare and beyond.\n\n**Trend
s towards significance**\nStephen Senn\n\nThere are many valid criticisms
of P-values but the criticism that they are largely responsible for the re
producibility crisis has been accepted rather lightly in some quarters. Wh
atever the inferential statistic that is used\, it is quite illogical to a
ssume that as the sample size increases it will tend to show more evidence
against the null hypothesis. This applies to Bayesian posterior probabili
ties as much as it does to P-values. In the context of P-values it can be
referred to as the trend towards significance fallacy but more generally\,
for reasons I shall explain\, it could be referred to as the anticipated
evidence fallacy.\nThe anticipated evidence fallacy is itself an example o
f the overstated evidence fallacy. I shall also discuss this fallacy and o
ther relevant matters affecting reproducible science including the problem
of false negatives.\n\n**p-value\, s-value\, B-value\, D-value\, … what
else?\nTolerance intervals: Beyond the t-test and p-values**\nBernard G F
rancq\, Ron Kenett\n\nThe statistical significance is often based on confi
dence intervals (or credible intervals in Bayesian analysis) and p-values\
, the reporting of which is requested by most top-level medical journals.
However\, in recent years there have been ongoing debates on their usefuln
ess\, leading to a ‘significance crisis’ in science.\nIn this talk\, s
ome alternative solutions proposed in the literature like the s-value\, D-
value or B-value (namely\, the probability that a patient under treatment
A ends up with a better clinical outcome compared to another patient under
treatment B) will be reviewed. We'll show that\, in medical research\, th
e treatment successes on the patient level can be elaborated using the con
cept of individual success probability (ISP) which generalizes the B-value
. The ISP allows a more pragmatic interpretation under both\, frequentist
and Bayesian\, paradigms. \nThe relationships between p-value\, ISP and to
lerance intervals will be discussed and illustrated with toy examples (1 s
ample t-test\, paired t-test\, cross-over trials\, 2 samples t-test) and r
eal-world data sets.\n\nhttps://conferences.enbis.org/event/22/contributio
ns/244/
LOCATION:
URL:https://conferences.enbis.org/event/22/contributions/244/
END:VEVENT
END:VCALENDAR