Total Pageviews

Monday, October 7, 2019

We need a lot more honesty in science

Scientists want to 
get published.

The probability 
of getting published
increases a lot 
if your study 
or experiment 
has a positive result,
or at least 
you claim it does.

But, in reality, most
technological advances 
are built on learning 
from repeated failures.

Failed prototypes.

Thomas Edison was the 
'king of repeated failures',
when inventing a light 
bulb filament that worked.

Today's world
of science 
usually ignores 
negative results. 

Roughly nine out of ten
published studies claim
to have produced 
positive results.

When negative results 
are not published in 
well known journals, 
other scientists
can't learn 
from them, 
and may end up 
repeating failed 
experiments.

The pressure
to publish 
leads scientists 
to spin results 
in a better light, 
or they may even 
commit fraud and 
manipulate data.

Up to half of science
experiment results 
can not be replicated,
suggesting there is 
a huge problem.

When funding agencies 
reward researchers 
who publish positive 
results, they'll get what
they pay for !


But science would
move forward faster
when good scientists 
who happen to get
negative results, 
are just as likely 
to be published
as good scientists
who get positive
results.

Academic conferences 
should promote 
honest discussions 
of failed experiments.

Funding agencies 
should support 
good scientists, 
who produce sound
results, even if they
happen to be negative.

A science experiment
does NOT become 
important ONLY
if there is a 
positive result.

Negative results, 
or inconclusive results
can be just as important.