An interesting post from the Science Careers blog:
"Numerous sources claim that Stephen Hawking once said that someone had
told him that every equation he put in one of his books would reduce
sales by half. Apparently, that's true of biology papers as well.
According to a study by Tim W. Fawcett and Andrew D. Higginson, scheduled to come out today in the Proceedings of the National Academy of Sciences
(PNAS), including lots of equations in a biology paper reduces its
influence, with the most math-heavy papers receiving 50% fewer
citations, on average, than other papers.
Strangely, while the
article is about biologists--the article is titled "Heavy use of
equations impedes communication among biologists"--the EurekAlert press release
makes it sound as if this phenomenon applies to science articles in
general. But one presumes that you would not find the same bias in, say,
a theoretical physics journal.
So what should biology
researchers do? Avoiding equations in science--even biology--probably
isn't a good idea. Assuming the authors have drawn the correct
conclusion--that is, that math-heavy biology papers aren't inherently
less important than math-light ones--it probably makes sense to put your
equations in an appendix, where, the article's authors say, they did
not affect citation rates.
Once the article is live it will appear here."
I don't know how biology works but I would imagine in economics there is some non-linearity to a trend like this (highly equation-dense papers are less popular, but so are those that have little equations... empirical papers would obviously have a somewhat different pattern, probably peaking out earlier).
I would guess this is just a proxy for something else. Of course no one likes slogging through equations. Even if you are comfortable with them, you still have to put a lot of effort into reading that kind of paper. I would guess that papers chock full of equations are by authors who have not made much of an effort to distill the point of it all.
I was once discussing math in journal articles with Richard Freeman and he said that when he's working on a paper if it's got lots and lots of math it means he doesn't have a real intuitive grasp of his argument yet. As he thinks through what's really essential and gets a better sense for what the math is telling him, he's able to pare it down and frame it better. So if it's a more manageable, clean model it's also probably something he is explaining really well in the text. That makes for a more influential and more citable paper.
Of course this is often in the eyes of the beholder. What some people call "math-heavy" will not seem that way to others.
I have not had the pleasure of writing a really math-heavy paper yet. I've used some pretty fancy empirical techniques, but the fancy math really comes in theoretical papers, which I did not do at the Urban Institute and have not tried to tackle on my own. I'm making a determined effort to get some more mathematical papers under my belt during the doctoral program. In my career I doubt I'll do much of that - I'll continue to be an empiricist. But it seems like a good thing to try out.
Turning a probability vector into a state
7 hours ago