6.4- Interview with Dr. Bradley Efron

So welcome. In this next module, we’re going to have a conversation with Dr. Bradley Efron. Many of you out there may be familiar with some of his work on the bootstrap resampling technique and lot of you have probably run into that. Just wanted to say a few words about his many accomplishments. I won’t hit everything here, but he’s a member of the National Academy of Sciences, a member of the American Academy of Arts and Sciences. He was a MacArthur fellow, he won the National Medal of Science, he was the president of the American Statistical Association at one time. He in addition, to all of his scientific and compliments, is extremely well-spoken. I’ve interviewed him before for journalistic pieces. Yeah, and he is the founding editor and editor-in-chief of The Annals of Applied Statistics and so, we’re really privileged to have him here today to get some advice from the editing and publishing side of things which is. I’m glad you didn’t say the horse’s mouth. So i’m just going to start by asking, besides good science, which is obvious we need in our papers, what key elements are journal editors looking for when you get a paper? Well, now with me you’re talking about a statistics journal. Yes, yes. Which is a different… A little bit of a different piece which is why I’m glad to have you. Statistics papers have a philosophical side and a technical side, and the people tend to get overwhelmed with the technical side and forget that they’re trying to make a point about doing statistics at some level is going to apply to scientiests communicating with statisticians, or maybe I should say statisticians reading the article and communicating with scientists. So it’s an essay in communication and if it’s not enough to be published, you want to be read also. So if you want to be read, make the paper pleasant to read. So newspaper writers are very good at this. The first paragraph or the first line says what story is going to be about. The first paragraph says it again in a somewhat greater extension. The first section then says it all again, every more detail. And a good scientific paper should make it clear to the reasonably good reader what is going to be said. A paper that’s going to be read should have some element of surprise in it. That is, it shouldn’t be by the time you say what you’re going to do everybody knows exactly how this is going to come out. But you should make it easy for the reader to get into the subject matter. In particular, if you have an example that motivated you, put it up front. Don’t hide it in the back after the person who’s worked his way through a lot of lumbers and theorems and bad notation. So editors and referees have to read the paper and make it easier to read the paper vastly increases your chances of having a success. Great, great. At every level. That’s great to hear. That’s a lot of what we’ve been talking about in this course. And what do you think is the number one mistake that scientists, mathematicians, statisticians make when submitting a paper for publication? Well, I run an applied statistics journal right now and I turned down a certain number of papers the first minute because they aren’t about applications. People love to write about theory and methodology because that’s an easy thing to do, applications are harder. So for my journal a little mistake is to send it to the wrong journal. And if you want to make a point about something, choose the right journal to submit to. And a journal is a magazine and people get the magazine and read it for interest. So it should be interesting. Try to avoid heavy notation and things that slow readers down right at the beginning, lots of definitions are bad right at the beginning. You should clearly say what you’re going to do and it should be clear, the abstract and the first introduction are crucial. That’s what editors mainly look at usually is the beginning of the paper and then they farm it out to associate editor or referees. So pay a lot of attention to the very beginning. Great. And you’ve said a few things already, but what other tips can you give to authors to increase their chances of getting published? Good graphics helps a lot in my field and an attractive format. And I avoid in my own writing, which is far from perfect, I avoid masses of equations or masses of definitions. I don’t mind using bullet points to set things off quickly. You can easily kill yourself by messing around at the beginning before, some speakers in talks spend an awful lot of time at the beginning not getting anywhere and they’ve lost, the best hook you have is right at the beginning. And so, think about that. Style isn’t terribly necessary. Einstein said, I’ll leave style to my tailor. But a pleasant reading style helps a lot and clunky English hurts, every time you use you have a sentence that sort of hard to decode, you’ve hurt the reader and they’ll stop pretty soon. And the reader might be the referee. Good, yeah. And so I think you hit upon this a little bit already about writing style and one of that questions that frequently comes up in my course has come up in this course that I’m teaching now is students say to me well, you know that’s not the way the scientific literature is written and, you know, it’s written in this style. And if I don’t copy the very verbose style that’s there, I’m somehow not going to be a member of the club and I’m going to be rejected and I might dumb down my sciences and are very free to write in a more clear style. So can you help alleviate that fear a little bit. It’s hard to dumb down more than, most people don’t know as much about the thing you’re talking about as you do. So dumbing it down is a good thing. That’s what you’re supposed to do especially at the beginning of the paper. You’re not going to impress anybody with fancy technical material. Everybody’s seen that or maybe three people in the world will be impressed, but you’re not trying to get to those three people usually. One advice I’d have is go read the greats. People like Neyman in our field or telling wonderful writers. And you can see how good they are at getting to the point and and not jumping around at it and not trying to be fancy. So don’t be fancy, dumbing down is not such a bad idea. And what advice would you have for a first time, a lot of the classes in their first paper submissions? I’m always grateful when I get a submission that says this is my first paper. I’m a graduate student, this is my early try. I give special attention to those papers and I think most referees do. And so it’s okay to call it out when you’re submitting the paper. Yes, yes. Definitely. And that that excuses certain mistakes, you can easily believe that a first time author might put in too much detail or too much stuff that their thesis advisor thought was interesting or something like that. So I’m definitely easier on first time authors, we want them you know fresh. The trouble with most journals is they’re dull and you want a fresh point of view and you want fresh writing and you want fresh ideas most of all of course. Yeah. Good. And do you have any tips for let’s say that that paper comes in and you reject it but you give the author the chance to resubmit with major revisions. Are there some tips at that in the process that you can give my class? Well, most journals I think have a category of we have various categories, it is withdrawn. If I don’t think the paper is appropriate, I’ll just withdraw at very first minute. Then there is rejected, flat out, but rejected with re-submission is a dangerous one for both the referee and the author. And one thing to do is not resubmit to go someplace else. But if you do resubmit and you can usually, if you press hard enough you very often can get published again. And there’s the usual thing about paying attention to what the suggestions were and stuff like that. Making the papers shorter is a good idea when you resubmit, shorter and clearer. But choosing another journal was not a bad strategy. I’d say recent re-submissions get through less often than first submissions. That is, they’re very often unsuccessful. They take an awful lot of energy from the authors. Yeah. Good, good. And can you give some words of encouragement for any scientists who might have gotten their first paper rejection. Well, my papers still get rejected. And I found over the years that the papers of mine that get rejected fall into two classes, ones where I was much too enthusiastic about the material and the other my best papers and my bootstrap paper, my best paper ever got rejected. And there I did persevere. And so, the system is far from perfect and very often fresh ideas arouse the ire of referees who are tied to the old ideas. It’s the editor’s job to try and spot such cases and persevere. There are too many papers, so people are always looking for reasons to reject papers. And I’m always worried especially if I see a paper that arouses a lot of hostility in the review, that’s sometimes a clue, it’s in my second class of reading papers that are annoying the readers because of fresh ideas. So anyway, don’t take it too seriously. Papers get rejected all the time, and doesn’t mean that you had a bad idea. Yeah. Good, good. And the publication process itself is undergoing a lot of changes right now. What kinds of changes do you envision are going to be happening in the next decade? That’s a question I never can answer in that everybody says that there aren’t going to be any print journalists, but I think there will be print journals, same way there’s still movies in the same way there is still radio. I think the print journals will get more like real magazines like New York or something like that. That is something that where the bundling effect of receiving a bunch of stuff in the mail and then having it be labeled is very interesting is important. And, of course, there’ll be a lot more electronic publication and maybe that maybe there won’t be journals, maybe everybody will publish themselves. If somebody has to do quality checks and somebody has to evaluate papers to say this is worth reading and this isn’t, that’s what the journals do. I don’t know what’s going to happen about electronic versus print. Good. And if there is one thing that you could change it out of the application process, what would it be? I wish people would submit fewer papers. I think people write far too many papers. It doesn’t help you get promoted to have a very long list of papers, at least, not in any place decent. What gets you promoted if that’s what the worry is is having an idea that impresses people. So fewer and better. Gauss’s motto in Latin, which I can’t remember is fewer but right. And that’s a really good motto if you can stick to it. Now, I understand that when you have to have some papers say you’re going up for tenure, you have to have some papers. They have to have made an impression on people. And so, people who publish too much wear out the audience in my opinion. So that would be one that I could change about it is less. Less is more. And we’re talking a lot about that in writing right now. And is there anything else that you would, any other advice that you would offer for the class? Yeah. Take the writing and thinking very seriously, don’t take the submission and acceptance, rejection part too seriously because it’s pretty random. Good. Great. Thank you so much for being with us today. Good. I enjoyed this. Thank you.