[Coral-List] A student's guide to the h-index

[Dennis Hubbard]

This has been an interesting thread, so I went back and looked at some of
the discussions about its use in cycling - where it was apparently created.
For those who have not tracked this down, "h" appears to be a number that
states "some rider" with a given h-score has ridden that number of rides of
that length. By way of example, a rider who has ridden 70 rides of 70 miles
would have an "h-score" of 70 - certainly better than a rider who has
ridden 2 rides of 2 miles. The problem is that, because the number is
non-dimensional, a rider who has ridden 70 rides of 70 miles has the same
score as a rider who has ridden 70 rides of 70 km (example in Google). And,
by extension, a rider who has ridden 70 rides of 70 cm will have the same
score as one who has ridden 70 rides of 70 km. If I apply this analogy
correctly to our discussion, a rider who rides 2 rides of 2500 km will
receive no merit badge and, therefore has no standing compared to the one
who rode 70 rides of 70 m.

More to the point, consider the poor graduate student who has published
only their second paper, but it is cited by 25,000 readers. (S)he either
doesn't qualify in this ranking... or would receive a maximum score of 2
(i.e., the same as published only two papers with two readers each). At the
other extreme, some prolific author (but whose research was so egregious
that their 2000 papers were typically cited in cited in 2000 rebuttals
would be an "h-god". If I am correct, Mike's statement, "Don't we conclude
from your two points that the h-index is a big big problem?" ranks near
infinity when it comes to an h-ranking?

Dennis

On Thu, Dec 31, 2020 at 10:15 AM David Blakeway via Coral-List <
coral-list at coral.aoml.noaa.gov> wrote:

> Mike and all,
> I'd like to learn more about how academia is done. As an outsider, no doubt
> my understanding is very incomplete.
>
> Perhaps I am focusing excessively on the h-index. And for sure that's
> partly because my own h-index, if I had one, would be so tiny I'd be
> ashamed to display it in public (but I don't want one so that's ok).
>
> BUT, reflecting on your two earlier points, which you said are more
> threatening to scholarship than the h-index:
> Your first point indicates that bureaucracy is likely to misuse the
> h-index, therefore isn't the h-index a big problem?
> Your second point I understood to mean gaming the system to artificially
> inflate authorship. Doesn't that equate to gaming the h-index, therefore
> isn't the h-index a big problem?
>
> Don't we conclude from your two points that the h-index is a big big
> problem?
>
> On Thu, Dec 31, 2020 at 4:09 AM Risk, Michael <riskmj at mcmaster.ca> wrote:
>
> > Hi David.
> >
> > Don’t mean to be a bubble-burster…I am just a cynical old guy living in
> > the forest.
> >
> > There are tons of problems with science, and only some of them are
> > self-inflicted. My pet peeve is…did you know that “manager” is now the
> > world’s most common profession? And the very idea that science can be
> > “managed” is repulsive. Doesn’t stop them trying.
> >
> > Way back when,. Canada used to pride itself on our granting system. NSERC
> > grants were restricted to 4 pages, and were reviewed by a committee of
> > peers. The camel got its nose under the tent flap. First, a government
> rep
> > sat in…next, the length limit was raised…then more managers...20 years
> down
> > the road, we are not much better than the soul-crushing NSF system. And I
> > don’t think we should attribute any higher motives to these managers than
> > we would to any other group. They will protect themselves.
> >
> > When we get involved with grading or ranking our colleagues, we really
> > need objective criteria, or seemingly objective. Otherwise, it’s
> > opinions-and they are like noses, everybody has one.
> >
> > In short, I see h-indices as one tool, to be used intelligently. Much
> > better, of course, is to go to the library and READ the papers on which
> > decisions will be based. I will spare you the story of the guy at our
> place
> > who MADE UP a buncha papers on the assumption no one would check. (I
> wanted
> > to fire his butt-the Dean said the paperwork would kill us and besides 4
> > more years and he’d retire.)
> >
> > #2 is a direct result, not so much of h-scores, but of the importance we
> > place on authorships. That’s as it should be, and that’s why this recent
> > hydra-like trend is so dismaying. All I will say is-have some
> self-respect.
> > Don’t do this. Because, as you say, it’s obvious what’s going on.
> >
> > Mike
> >
> > On Dec 30, 2020, at 2:14 PM, David Blakeway <
> > fathom5marineresearch at gmail.com> wrote:
> >
> > Mike,
> > Thank you! (for bursting my bubble :) I have this delusion I see the
> > problem, but it's only 'a' problem.
> > Regarding your point 1, nasty me thinks 'yeah, managers, my second
> > least-favourite species after marketers' Nice me thinks 'they are
> > well-meaning, just trying to advance the department/school/university
> > according to current best practice'
> > To point 2 it's 'come on, stop this joke, we all see what you're doing'
> > and 'worldwide collaboration is exactly what we need right now'
> > What do you mean when you say you 'despair of solutions'?
> > And isn't your point 2 a direct consequence of the h-index?
> >
> >
> >
> _______________________________________________
> Coral-List mailing list
> Coral-List at coral.aoml.noaa.gov
> https://coral.aoml.noaa.gov/mailman/listinfo/coral-list



-- 
Dennis Hubbard - Emeritus Professor: Dept of Geology-Oberlin College
Oberlin OH 44074
(440) 935-4014

* "When you get on the wrong train.... every stop is the wrong stop"*
 Benjamin Stein: "*Ludes, A Ballad of the Drug and the Dream*"