Excerpts from conversations with 3 PHD students about what they want, what they would like to see.
I recently asked some friends of mine who were PHD students what they wanted to see more of in their field. Is there anything that they needed/wanted? What would make their lives easier? Student 1 (AKA LF)
Ok, as for the bigger question. I’m not sure I’m a great person to answer this yet. All of my research is so nascent that I have no desire to have people read anything yet on a larger stage. I could imagine, maybe, when a project is pretty far along but not yet published, maybe wanting to have an outlet for people to see it. A lot of students, especially those on the job market for professor jobs, have personal websites where they could post papers at their discretion.I wonder if a social media site where students can friend other students, and kinda make a mini personal website with abstracts of their research projects and maybe drafts of working papers would catch on. I think it would be crucial to have a lot of controls on who can see what, bc you may not always want people to see stuff that’s not ready for publication. I could see potential for that maybe tipping. Message boards where you could crowd source questions about the field (or just see who’s going to conferences, find people to room with, etc), and a good search engine to see what’s out there by topic would be nice features in theory.
From her PHD friends on the topic of finding PHD papers when doing your own PHD research, and how journals are currently the best gatekeeper of quality, and problems with journals.
On the topic of cost of discovering new great PHD papers and the cost of just reading them:The approach is, you learn who the good people are in any given area, and basically you just read their papers. There are enough good people in any subfield, past and present, that you could read papers only from good people all your life and never run out of more things to read. So the obvious problem is that there must be plenty of good papers by unknown people, which will just never get read, hence people will just keep wastefully rediscovering the same things until someone good happens to rediscover it, and then it becomes known. But there’s no obvious solution. (The problem of how to get exposure for yourself is the other side of the same coin).The current best solution is journals – each journal has a quality level, journal referees review papers, and if a journal accepts a paper then you can guess about how good it is – so like that unknowns get a chance and gradually become knowns, but journals are noisy, and they’re very slow, so they’re pretty good for catching up on what was happening five years ago but not so useful for what’s going on now.
A challenge for trying to come up with a better solution – these two things are true:1) The cost of reliably telling a good paper from a mediocre one is very high. To approximately calibrate this cost, think about how time consuming the referee process is, and still the noise to signal ratio is pretty high (a lot of papers in top journals are actually mediocre). I’d guess, not counting the authors’ time which is much more on top of this, there’s conservatively maybe $10k worth of labor for each accepted paper in a top journal (based on 3% acceptance rate of QJE, 3-4 hrs/paper average referee + editor time – more for many papers, much less for obviously terrible ones).2) The cost of just not bothering to read any paper that you suspect is mediocre, is very low (since there’s effectively an unlimited supply of good papers, and nobody else is reading the mediocre stuff anyway).
On MOOCs and their shortcomings:
Hmmm as for the big question about research display and distribution …I would have to think more and perhaps discuss at lunch to percolate more ideas . . but my first thought is I knew MOOCs won’t replace in person education . . humans just don’t have the resolve and attention span to commit to that way of absorbing information.
Lastly, on lit review:
One frustrating thing about doing a lit review is that I never feel complete. i only think i know a topic if i am more embedded in the actual social network of people that work on the topic (george and authenticity). . otherwise i have no idea if the google scholar generated articles related to authenticity are complete/useless/pioneering etc. it takes awhile to get a sense of the lay of the land, to develop taste.
That’s about it for now. I really enjoy listening to PHD students who are in pretty much most regards, extremely insightful, great writers, and clear thinkers. I hope to learn more about what they think can be improved in the field and what technology they wished existed. I would like to learn more and perhaps help fill those gaps in the future.
Written on August 18, 2014