Total Pageviews

Sunday, April 19, 2020

The H-index has “Severe unintended negative consequences”

SUMMARY:
The H-index is the 
most widely used metric 
in academia for measuring 
the productivity and impact 
of researchers. 

But ...
The H-index encourages
researchers to focus
on popular research 
subjects, and to agree
with the consensus.

The index discourages
Innovative thinking 
that deviates from the 
mainstream consensus.

But ...
The history of progress
in science is filled with
individuals, or small teams,
refuting consensus beliefs.

That's why 
the H-index
is anti-science.



DETAILS:
Jorge Hirsch 
proposed the H-index 
as an objective measure 
of scientific achievement 
in 2005, specifically for 
theoretical physics.

Hirsch was a physicist 
at the University of 
California, San Diego. 

The metric looks at 
the number of papers 
a researcher publishes, 
and how many 
citations they 
receive. 

It is a popular tool 
for assessing job 
candidates and 
grant applicants.

It is also one of the 
most contentious topics 
scientists discuss, 
as Hirsch wrote in the 
Physics and Society 
newsletter in January.

“About half 
the scientific 
community loves 
the H-index 
and half hates it,” 
wrote Hirsch. 

“The H-index 
of the scientist itself 
is a great predictor 
of whether s/he 
belongs to the first 
or the second group.”

Hirsch still believes
that the H-index 
is a decent measure, 
but he knows it can 
“fail spectacularly and 
have severe unintended 
negative consequences.”

It can deter researchers 
from innovative thinking. 

A student working 
under a professor 
with a high H-index, 
for example, may be 
reluctant to question 
concepts they are 
being taught.

The quest for a high 
H-index encourages 
researchers to choose 
‘hot’ research topics, 
that are more likely 
to gain attention.

On such a subject, 
such as climate change, 
they will be tempted 
to publish one paper 
after another, to boost 
their H-index. 

“It’s a little too sensitive 
to what’s popular and 
fashionable in science,” 
says Hirsch. 

Hirsch points out that 
the metric doesn’t pick up 
on research that deviates 
from the mainstream 
consensus.

“If you write a paper 
that that’s not 
generally accepted, 
it’s an uphill battle 
to get people 
to consider it,” 
says Hirsch. 

“But just because 
something 
is accepted, 
it doesn’t mean 
that it’s right.”

Hirsch writes 
that his own papers, 
on the theory of 
hole superconductivity 
are “far more important 
than any other work 
I have done that has 
a lot of citations”
but their significance 
cannot by gauged 
by using his 
H-index alone.

“If we believe 
citations and 
H-indices, 
by all counts 
my contributions 
to the 
understanding of 
superconductivity 
are insignificant,” 
writes Hirsch. 

“Therefore, 
I have to conclude, 
much to my regret,
that the H-index 
fails in this case.”

“If you make decisions 
just based on someone’s 
H-index, you can end up 
hiring the wrong person, 
or denying a grant 
to someone who is 
much more likely to do 
something important. 

It has to be used carefully.”,
said Hirsch.