I read the article by David Vaux yesterday, and being a biostatistician, I
could not agree more with his views. Although his comments refered mainly
to small sample size studies, large sample size studies are even more
problematic due to other statistically trickier issues such as batch
effects and confounding factors. Frankly, the whole field is an absolute
mess and it is not suprising to me hat so much of modern biology is
irreproducible, and therefore it is questionable if such
research represents genuine science. I am more inclined to think that it is
a waste of money since a lot of published "results" are wrong, or are
overinterpreted, eventually misleading other scientists. The problem is not
only that biologists don't have the statistical background, but more often
than not, they just don't have the aptitude to learn (I recall from my
school days that those doing biology were, generally speaking, rather poor
at Maths and Physics, and even hated these subjects...). Most
worryingly, biologists sometimes don't even listen to the statisticians
(without sounding paranoid here, but I have detected apathy, fear and
prejudice in biologists when confronted by statisticians). Finally,
budgetary restrictions, and the pressures of grant funding means that
biologists often knowingly turn a blind eye to statistically fragile
results or simply abuse what seem to be positive but statistically
insignificant results. As long as they get their paper into Nature, that is
all that matters to them, and determining whether the result underpinning
their work is statistically sound or not is not of their main concern. Sad
but true.
*From:* William Stanbury <williamstanbury@GMAIL.COM>
*To:* allstat@JISCMAIL.AC.UK
*Sent:* Friday, 4 January 2013, 1:12
*Subject:* Re: Comment in Nature on discipline of statistics (Vaux 13 Dec
2012)
Interesting situ: Vaux is a francophone name, do you know his University
and possibly nationality or background? As a native speaker having worked
several years with francophone Belgian statisticians, I understand the
issues here. Thanks.
On 3 January 2013 12:24, Allan Reese (Cefas) <allan.reese@cefas.co.uk>wrote:
Nature (Vol 492 p180) contains a personal comment article by a professor
of cell biology. He condemns reporting of "P values for single
`representative' experiments. ... Because science represents the knowledge
gained from repeated observations or experiments, these have to be
performed more than once - or must use multiple independent samples - for
us to have confidence that the results are not just a fluke, a coincidence
or a mistake." ****
** **
There are follow-up comments at
http://www.nature.com/nature/journal/v492/n7428/full/492180a.html that
appear to come from other scientists, not statisticians.****
** **
I have sympathy with Vaux's views but think they are badly expressed and
muddled as a critique of the way statistics is used in science. One
specific point is that he confuses the statistical analysis with the
reporting ("I see figures with error bars that do not say what they
describe.") I have personally exchanged emails with Nature on the subject
of poorly designed graphs, and was put off on the basis that they had a
comment article in press. Hence I don't want to continue an individual
campaign against the magazine, but maybe Allstat members would be
interested to read Vaux's article and respond. Maybe editors should ask a
qualified statistician to referee statistics. ****
** **
Just to rattle your cage, "[Experimental biologists] don't all need to
understand complex statistics, or hire professional statisticians, but
there would be fewer sloppy papers if every author, reviewer and editor
understood statistical concepts such as standard deviation .... [boxed
Glossary] Standard deviation: The typical difference between each value and
the mean value." ****
** **
Allan Reese****
(personal view, not on behalf of Cefas)****
This email and any attachments are intended for the named recipient only.
Its unauthorised use, distribution, disclosure, storage or copying is not
permitted. If you have received it in error, please destroy all copies and
notify the sender. In messages of a non-business nature, the views and
opinions expressed are the author's own and do not necessarily reflect
those of Cefas. Communications on Cefas' computer systems may be monitored
and/or recorded to secure the effective operation of the system and for
other lawful purposes.
You may leave the list at any time by sending the command
SIGNOFF allstat
to listserv@jiscmail.ac.uk, leaving the subject line blank.
You may leave the list at any time by sending the command
SIGNOFF allstat
to listserv@jiscmail.ac.uk, leaving the subject line blank.
You may leave the list at any time by sending the command
SIGNOFF allstat
to listserv@jiscmail.ac.uk, leaving the subject line blank.