NONMEM Users Network Archive

Hosted by Cognigen

post p < .05 era

From: Kimko, Holly [JRDUS] <HKimko>
Date: Mon, 11 Apr 2016 17:54:36 +0000


On March 7, 2016, the American Statistical Association announced a statemen=
t regarding p-value (click "Statement<
/10.1080/00031305.2016.1154108>"). I welcome this statement, especially bec=
ause of some clinical teams that design PoC studies to focus mainly on gene=
rating a "good p-value" for Go/No-go decisions rather than "learning". In m=
y opinion, a p-value that follows a hypothesis test could be useful but, wi=
th the same budget, a PoC study should be informative to design the next st=
udy for confirming efficacy (and/or safety) with a good p-value for a clear=
er regulatory decision.

This is a seminal paper that emphasizes the importance of clinical pharmaco=
logy approach over empirical p-value driven drug development approach:

      Learning versus confirming in clinical drug development.
      Sheiner LB<
Author%5D&cauthor=true&cauthor_uid=9084453>. Clin Pharmacol Ther.<http:=
//> 1997 Mar;61(3):27=

A recent development in considering "totality" of data in designing and/or =
approving a pediatric drug with less emphasis on p-value (my interpretation=
) can be found in:

      (Draft) Reflection paper on extrapolation of efficacy and safety in p=
aediatric medicine development (April 1, 2016) Click here<

I look forward to living in the post p< .05 era in drug development!

Kind regards,
Holly Kimko
Clinical Pharmacology & Pharmacometrics
Janssen Research & Development, LLC

Monday, March 07, 2016 11:15 AM

Today, the American Statistical Association Board of Directors issued a sta=
tement on p-values and statistical significance. We intend the statement, d=
eveloped over many months in consultation with a large panel of experts, to=
 draw renewed and vigorous attention to changing research practices that ha=
ve contributed to a reproducibility crisis in science.
"Widespread use of 'statistical significance' (generally interpreted as 'p =
< 0.05') as a license for making a claim of a scientific finding (or implie=
d truth) leads to considerable distortion of the scientific process," says =
the ASA statement (in part). By putting the authority of the world's larges=
t community of statisticians behind such a statement, we seek to begin a br=
oad-based discussion of how to more effectively and appropriately use stati=
stical methods as part of the scientific reasoning process.
In short, we envision a new era, in which the broad scientific community re=
cognizes what statisticians have been advocating for many years. In this "p=
ost p < .05 era," the full power of statistical argumentation in all its nu=
ance will be brought to bear to advance science, rather than making decisio=
ns simply by reducing complex models and methods to a single number and its=
 relationship to an arbitrary threshold. This new era would be marked by ra=
dical change to how editorial decisions are made regarding what is publisha=
ble, removing the temptation to inappropriately hunt for statistical signif=
icance as a justification for publication. In such an era, every aspect of =
the investigative process would have its appropriate weight in the ultimate=
 decision about the value of a research contribution.
Is such an era beyond reach? We think not, but we need your help in making =
sure this opportunity is not lost.
The statement<
154108> is available freely online to all at The American Statistician Late=
st Articles website<
urnalCode=utas20>. You'll find an introduction that describes the reasons=
 for developing the statement and the process by which it was developed. Yo=
u'll also find a rich set of discussion papers commenting on various aspect=
s of the statement and related matters.
This is the first time the ASA has spoken so publicly about a fundamental p=
art of statistical theory and practice. We urge you to share this statement=
 with appropriate colleagues and spread the word via social media. We also =
urge you to share your comments about the statement with the ASA Community =
via ASA Connect<>. Of course, you are more =
than welcome to email your comments directly to us at ron
On behalf of the ASA Board of Directors, thank you!

Jessica Utts
American Statistical Association

Ron Wasserstein
Executive Director
American Statistical Association

Received on Mon Apr 11 2016 - 13:54:36 EDT

The NONMEM Users Network is maintained by ICON plc. Requests to subscribe to the network should be sent to:

Once subscribed, you may contribute to the discussion by emailing: