It’s been quite some time since I’ve come across Orac (several years ago I came across him in the context of criticism of homeopathy, I think). Good to see that he is still around, and writing as clearly and acerbically as ever. 
Not so good to see another prominent scientist has disappeared down the Covid-19 rabbit-hole.
Also an always-timely reminder that being able to calculate some measure or index does not guarantee that is necessarily meaningful. As a quant-geek, it’s something that I have to take a (metaphoric) baseball-bat to my head to drum that point in again every now and then. 
For the record,
| Scientist | K-index |
|---|---|
| Joshua Swamidass (me) | 2.30 |
| Nathan Lents @nlents | 3.91 |
| Pierre F. Baldi (my advisor) | 0.00 |
I think lower is better? I suppose none of us are Kardasians, as the cutoff is (apparently) 5.0.
For reference…
K\text{-index} = f / {43.3 c^{0.32}}
Where c is the number of citations and f is the number of twitter followers of a scientist.
So if you have no followers and don’t tweet, shouldn’t your K-index be undefined, not zero? Or did you reverse the numerator and denominator?
Ah, I see what happened: you reversed f and c.
| Scientist | K-index |
|---|---|
| Me | 0.10881 |
But I only have 66 Twitter followers and an inflated number of citations.
Thanks for the catch. Fixed it.
It’s linear in f? That just seems wrong, as it would be far easier to generate a million followers than a million citations. I would have thought it would be log(f) or similar.
But then, as Orac pointed out, it is (or is supposed to be) a satirical index.
Those of us who have at least one cited paper, but no Twitter account have an infinite K index, right? That makes infinitely more seriously scientific that the rest of you!
No, you just end up with a K-index of 0.
But what if you have 0 cited papers? Undefined.
And since lower is better, having a K-index of zero makes you the most sciencey you can possibly be.
Still, the best strategy for those already with twitter followers is to avoid publishing any papers that could be cited. That way they will never have a K-index over 5.
If you have 1 citation and x Twitter follower, Josh’s formula gives 1/(43.3 x^0.32). Which approaches infinity as x —> 0. I think high, not low is more seriously sciency. Right?
I think you have that backwards. The more citations, the lower the K-index, isn’t it?
He originally had the variables switched. I thought that had been fixed.
OK, I see, f and c had been reversed in the formula I used. Those of us with no Twitter account but some citations are the proud owners of a K-index of zero, the ultimate serious-sciency achievement.
I think you are misusing the statistic outside its originally intended and (lackadaisically) validated domain…
The statistic is not something to be taken seriously.
Here I am asserting that someone with any old citation, even just one, but no Twitter account, is a world-class scientist. And you folks don’t get it that I have “tongue in cheek”. Hmm
This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.