Attributes that make getting tenure easy: curiosity, effectiveness, and conscientiousness.

  Curiosity may be the most important characteristic of a successful academic. It is something that motivates us a lot in our earliest days, but sadly, many get over it. Nurture your curiosity for a successful academic career in both teaching and research. Look for puzzles and enigmas. Find new ways to teach and learn. Wonder about things that others may seem to think solved. Wander intellectually into new areas. Bring techniques from one area to another. Study a novel organism, or a known organism in a novel way. Explore.

I think that it is curiosity more than anything that is a key to a successful academic career. Some may argue that drive alone can make a career successful. But the trouble with drive is its close relationship to competitiveness. If this alone motivates you, your research will parallel that of others too closely as you try to compete with others. Your teaching may also suffer. Curiosity gives you the freedom to go on your own path and discover completely new things in all aspects of your career.

But curiosity alone is not enough. Alone, it might cause you to stop when you think you know the answer and move on. A successful academic needs to share her discoveries with others, typically by publishing, something often best done with a funded team of eager, curious investigators. The only way to lead such a group is to be effective. By this, I mean planning and publishing the fruits of curiosity. I mean running a well-functioning group. Some of these skills are like businesses, full of meetings, evaluations, goals, and plans. But they are driven by curiosity, not profit.

I suppose that curiosity and effectiveness will take you nearly all the way there. After all, if you are doing excellent curiosity-driven research with a well-led and productive team that publishes regularly, what else is there?

This is where conscientiousness comes in. It is for all the other stuff professors need to do. It is astonishing how many small but important tasks can fill your day, or disappoint others if you fail to do them. You need to be responsible to others from attending meetings, teaching and planning classes, writing letters of recommendation, reviewing papers, serving the department, and your professional societies, and mentoring. You need to make hotel reservations, reserve flights, pay dues, and many other such things if you aren’t lucky enough to have an assistant. You need a good system for keeping track of these things, for answering emails expeditiously, and not disappointing those who depend on you. You also need to figure out how to keep these tasks from consuming the part of your day that should be devoted to curiosity. I bet I have at least 5 of these kinds of tasks a day. I sometimes do them first to get that sense of accomplishment, but I know it isn’t enough.

Curiosity, effectiveness, and conscientiousness matter. Pay attention to these three, and the rest may well take care of itself.

Keep your research honest, unbiased, comprehensive, and blind

Science cannot advance on fraudulent publications, whether the problems are big or small. We all know the basics of honest research, but there are also things we need to be taught. These are based on understanding our inadvertent tendencies to accept data that support our hypotheses and view data that reject them as aberrant.

Did you see the Comment by Glenn Begley in the 23 May 2013 Nature entitled Six red flags for suspect work, pages 433-434? It is a followup on a couple of studies that show that most preclinical cancer papers in top journals could not be reproduced, even in the same lab (references in the paper cited above). The point of Begley’s commentary is that it is not so hard to identify the poorly done papers. Begley breaks it down to six thoughtful questions that are good to think about in planning your own research and in evaluating the research of others. Some of the points may be more relevant to certain kinds of lab projects, but most of them are generally true.

I’ll list them here. Just to be clear, I’ll be using Begley’s exact words for the six headings, then my own words to explain them in contexts I understand.

1. Were the experiments performed blinded? This means that the person gathering the data does not know which case is part of which treatment, so the person cannot inadvertently boost some scores and lower others. You may think you wouldn’t do this, but you would, all the psychology studies show, no matter how honest you want to be. So scramble those treatments and make sure the person doing the counting, measuring, or scoring does not know how any given score will impact the overall result.

2. Were basic experiments repeated? Of course there need to be replicates. There need to be as many kinds of replicates as you can think of, within feasibility. In the paper you should explain what replicates you did. Back when I was working on wasps, the replicates were simply different wasp nests treated the same, so there was a sample size. Without this, I could not do statistics. You might have replicates that are locations. If you have a high elevation and a low one, you need more than one pair, or the differences could be due to something else. Now I work on a micro-organism, I can do exact replicates of the exact same clones. A full replicate will come out of the freezer and be tested independently fully. In the lab we have started talking about replicates and duplicates to reflect different levels of independence. I can’t believe anything would be published in a high quality journal or anywhere without repeated experiments.

3. Were all the results presented? One of the thing we talk about a lot in the lab is that you can’t edit out a troubling datapoint, unless you do tests both ways and point out a reason to remove it, like the incubator broke or something. You can’t edit out bands on a gel. Of course you can’t differentially enhance areas of the gel picture. I imagine you might do your analyses on all the gels but just show a representative one. These days I suppose you could put all the images in the supplement. If your data are more ecological or evolutionary, you might have tons of kinds of analyses you did, but don’t present. I’ll have to think about what to do here. How do you distill hundreds of hours of videotape of wasp behavior, for example, to answer a question and still show all the results? Think about what this means for your field. Be uniform and as comprehensive as possible.

4. Were there positive and negative controls? This is much more of a lab experiment thing. If you are doing PCR you should have a lane with no template DNA in it, and a lane with a template you are confident will work, a negative and positive control. But with field experiments the exact nature of controls gets more complicated. As much as possible, have positive and negative controls. Think hard about controls, about random design using blocks so conditions you haven’t thought of don’t impact your outcomes. I wrote earlier about a student who worked in a plastics lab after working in our lab. When I asked her if there was any carry over, she said that she had a much better understanding of thinking hard about appropriate controls than her fellow students had.

5. Were reagents validated? Begley’s description of this one talks about whether antibodies bind to the right thing or not. You should always be sure your reagents work. We have had lots of experiments fail because of bad reagents. I think this one really is tied up with the one before, that there should be appropriate controls that would detect poor reagents.

6. Were statistical tests appropriate? In my lab we spend a ton of time worrying about statistics. We generally use R so what students learn they can take with them anywhere. We worry about independence, about appropriate degrees of freedom, about whether the data are normal and so can use parametric statistics. We get it that you can’t pick and choose what data go into the statistics. We know you can’t end the experiment just because you now have significance (p hacking, written about earlier). We often have nested data. Everyone should take a statistics course or take several and be careful. There is no such thing as curves that are so obviously different that statistics are not needed. IN general ecologists and evolutionary biologists are probably good at this one overall.

This is Begley’s list.  There are certainly other points that are important. So, think about all of these as you design your study and see if when you read a paper you can determine whether the paper you are reading succeeds or fails on the Begley scale.

Check list for a professor’s retirement symposium

You will do few things more important in your academic career than organize a symposium for your retiring professor, or at least that is how I think of it. The point of this entry is to make it easier for others to organize similar meetings. Think of it as getting a hundred thousand dollar grant to have all your favorite people come to a symposium. All you have to do is organize and get the university to throw in a little more for the events. The most wonderful people will pay their own way to come. I did this for Alan Templeton this week and will share here how it is done.

1. Get a list of people who should be invited. This will be the honoree’s students, postdocs, anyone on this list from him. It could include family, undergrads, and collaborators. There is no way you can put this list together yourself, so get your person to identify the people. Get up to date emails for all of them. Ask them if there are people missing from the list.

2. Choose a date well in advance. I think about 10 months in advance or more is about right. That way people will have their next year’s meeting comittments all set. Choose a few dates that are not graduation weekend or anything like that, according to what works for your honoree. Then have the potential attendees let you know. For us, what worked was Sunday evening 2 June, with a symposium on Monday 3 June.

3. Reserve a block of hotel rooms in a nice hotel, preferably walking distance away. We got a great price on a block of rooms at the Moonrise on the Delmar Loop, near to places of nostalgic importance for many.

4. Choose a location. Ideally this should be at the university the students attended, where, with luck, the professor still works. That maximizes the sentimentality and increases attendance, but I have been at symposia elsewhere.

5. Set up a mailing list. This should always contain everyone, even those that say they can’t come. We got some last minute additions from faraway places like Australia. Just keep everyone in the loop. Don’t do blind CC so everyone can see everyone else on the list and get excited to forge new ties and renew old ones.

6. Plan the event and reserve the venues. We decided to begin on Sunday evening with an informal dinner at our home, then have an all day symposium on Monday, ending with a dinner on campus that evening. We reserved the room on campus for the symposium, an auditorium that seated 300 with room outside for posters and lunch. We also reserved the dinner location and contacted potential caterers for that dinner. We planned the menus.

7. Figure out who will come, though this will be a fluid list. Keep in touch with everyone and remind them every month or so about the event. In particular remind them before the hotel releases the rooms.

8. Buy a guest book. People should have something to sign and write something. Make sure the out of town people can email you words to be put in the guest book. Bug people all day to sign the guest book. I got one with lined paper and a nice cover at the campus bookstore.

9. Decide exactly what the event will be like. I met with Alan Templeton, and we decided on a schedule that had 10 talks, including a longer after-lunch one for him. We had 25 minute talks and lots of breaks. After all, what people really want is to talk to each other. Our schedule had six talks in the morning, with a 45 minute break and four talks in the afternoon with an hour break. There were two hours off before the dinner and many of us went to the bar in the Knight center then. You might have lightening talks, or the like. We had a roast at the evening dinner, of course, but it was not open to all. We provided the break food and the lunch to everyone.

10. Have posters. The solution to not having many talks is having posters. Everyone wants to share their science and posters are a great way to do it. Local people can also contribute posters. Posters also give substance to the breaks. We had about 20 posters.

11. Take a group photo. This is something I did not think of, but someone else did, fortunately. We took a photo of Alan and all his students and postdocs, at least the ones that could come.

12. Accept help. In some ways the hardest part of the organization was deciding what the event should look like. Once we decided on a Monday symposium with a party the night before and a party the evening of, things started to fall into place. But I couldn’t have done it without a ton of help from Karla Garcia in our office.

13. Take photos. It is a wonderful thing to see people greet each other. The research team of a brilliant professor like Alan Templeton is like a family. The students helped and supported each other through some of the hardest moments of life, and generally succeeded in ways that carved their careers.

14. Have fun. It was so exciting to realize we had all the food ready for the Sunday dinner, thanks to our helpful students. We were overwhelmed with all the arrivals and the fond greetings. The poster boards filled. The talks were given. There were no glitches. Food arrived, was set out by our wonderful Biology Department Staff, and was eaten. It was an event I’ll never forget. It was an event that taught me a lot. After all, people do science, but it is the science that brought us together.

15. Make a photo book. I have to say I just love these. I use Shutterfly. There is something nice about real paper. Alan, your book will be in your hands soon.  Here is the link, if you are curious, way down there at the bottom:

 

 

 

 

 

 

 

 

 

http://share.shutterfly.com/share/received/projectdetail.sfly?fid=2c24bf42207f6cc15b2f2841c4d4a52e&sid=0EcNHDFuxZM3rA&key=0EcNHDFuxZM3rA&shareProjectTitle=Photo+Book&pid=SFLY&cid=SHARE3PSJXX

There, that is it. It was some work to put together, but what could be a better use of my time than to gather together a great team of scientists, a great symposium, and honor to one who has given us so much.  Thank you, Alan! I hope to see much of you in retirement.

How to throw a retirement party for 65 people at home for well under $500

   Why do so many people cater their parties? They end up with soggy food that should be hot but is tepid and is certainly overcooked and overpriced. It just isn’t that hard to have a great party at home for a reasonable sum and not too much work, either before, during, or after the party, especially if you have a few post-docs or grad students to help out, like Debbie Brock, jeff smith, and Katie Geist. Maybe if I share a few tips on what we did, others will take the plunge! For smaller parties of under 20 we usually make pizza, making it ourselves, mostly during the event. But a large gathering needs a different approach.

First of all, it helps to understand that parties are generally not about the food or drink. They are about the people. So think about the geography of your home. Where do you want people to gather? Plan for a few places and distribute drinks coolers and food at them, though the main food can be on a central table. Most people won’t be sitting down for long, so you can get away with a quarter as many seats as people, or even fewer. Lots of our seats were unused. We had chairs, food, and drinks on the front porch, in the living room, the dining room, and the back deck.

Music is good. We put Pandora on the Cat Stevens station and it played throughout the house.

It is June, so we decided not to have any hot food. This also makes serving easier. You just put everything out and forget about it. We had way too much of some things, so I’ll be writing about correct amounts, not what we had.

The longer people are at a party, the more they eat and drink. It seems that three hours is a typical length for a gathering, but a larger one will last longer. I was delighted that ours lasted from 5:30 to 11:00 pm, a five and a half hour success.

OK, here is the list for 65 people:

Drinks:

Wine: one case, 8 red bottles 3 white bottles

Beer: 72 bottles

Soda water: 24 cans

Plain water: from faucet in a pitcher with ice

Orange juice, one gallon

Lemonade, homemade, one gallon

Assorted syrups for water: raspberry, hazelnut, black currant, cherry

Coffee can also be good, but I forgot to make it.

Food:

Cheeses: cheddar, 2 lbs., French comte 1 lb., brie 1 lb., goat, 1 lb.

Homemade bread: 6 lbs.

Mixed nuts: 2 lbs

Fruit: apricots, grapes, strawberries, tangerines (2 lbs. each)

Vegetable tray: celery, red pepper, carrots, cherry tomatoes (one tray with dip is enough)

Pasta salad 6 lbs. uncooked pasta (olives, red & yellow peppers, cherry tomatoes, mild onions, pine nuts, pesto or vinaigrette

White bean salad 3 lbs. uncooked beans (white beans, vinaigrette, rosemary, mild onions)

Brownies 2 big pans with nuts

Biscotti, homemade 2 lbs.

I made gluten free cupcakes but only one person ate them, so the GF people might just enjoy the strawberries.

There, that’s it. Plenty of food and drink for all.  The bread and cheese, fruit and brownies were big hits, as was the pasta salad. Biggest overestimate were the veggie trays. I just won’t do cokes and things, and the non-alcoholic stuff we had seemed to work. Next posts will be about organizing retirement festivities in general, including a checklist.

I have to add that I would rather have had vegetarian black bean chili for tacos with all the trimmings, pico de gallo, guacamole, jalapeños, sour cream, you know, but my nostalgia for Texas is running high enough as it is! (Can’t wait to hit Navasota and see family on Friday!) Then we could have had tortilla chips and salsa too. OK, just stop!

What to do about a low GRE verbal score

These two have already learned two languages.

How can you fix a low GRE verbal score when it is dependent on a lifetime of reading, listening to complex language, and writing? Little children have no control over the richness of language they hear. They don’t get to pick whether they were in a family where complex ideas were discussed at dinner. They don’t get to pick whether they had a lot of siblings close to their age, naturally bringing the conversation level down, or were alone and had to listen to the adults.

Today the New York Times had an article about how much harder it is for a school to raise reading test scores than it is to raise math scores. Part of it was based on a Mathematica Policy Research study of KIPP schools that showed 11 months advancement in math and 8 in reading at middle school. Part of the argument in the NYT is that children differ more in reading than math exposure at home.

But what does this have to do with improving a low GRE score? Only that it can be challenging. I think there are two basic approaches. One is to really cram hard for the test. If you study an hour a day for a few months, memorizing words, working examples, reading, and answering questions you will undoubtedly bring your score up. How much you bring it up will depend in part on how long and hard you study. It should be in the hundreds of hours, not the tens of hours. But you already knew this.

I think there is another perhaps more thoughtful approach that involves the reasons universities even care about your verbal GRE in science programs. To be an effective scientist, you need to be able to communicate what you do. You need to be able to think through verbally the steps you need to take in your research to make a strong argument that your experiment actually addresses the questions you want it to address. You need to understand the nuances of language in other people’s work, from letters of recommendation to review articles to original research. If you can somehow show directly that you have mastered these skills in spite of a low GRE, then your evaluators are less likely to hold the low GRE against you. Furthermore, this will actually help you with your career.

What does improving in this way look like?  First of all, you should set aside time to read for pleasure every day, without fail. I would choose fiction before bed time. There are a lot of older authors you can get tons of work by for very cheap on your smart phone or pad. I’m working my way through Willa Cather right now, having obtained her entire works for under a dollar. I’ll probably move to someone else after the third book. Actually, I see a lot of the individual books are completely free Kindle edition on Amazon. I read so far O Pioneers! and The Song of the Lark, and just began My Antonia. After reading Larry McMurtry about the Texas West, I wanted another perspective, much as I loved Lonesome Dove. So I chose a female writer and sodbusters. But you should choose something you like and read it not only for the story but look at how it is constructed.

Spend some time just watching comma usage. Then focus on the first sentences of paragraphs. Then see how seldom the author uses that weak linker, “and.” Don’t only let the story flow in. Examine it, maybe for the last few minutes of your reading period.

Reading is important not only because it brings joy to uncertain lives, but also because it gives you information. But it is not enough if you want to become an effective scientist. You also need to write. Write often. Write clearly. Share your writing. Read about writing. You can write every day or once a week. You should have something to say and find an audience. A blog is a natural way to write these days, but it should be focused on a topic you know something about. Your writing should be evidence based, not just opinion. This means you need links and sources.

You could also begin to contribute to Wikipedia in areas that interest you. Just be careful to maintain a neutral tone, to put in citations to evidence-based research, and to see what others think of your work. After awhile you will have a portfolio of writing that you can link to in your graduate school application. You can point out that while you may not have the highest of verbal GREs, you have mastered the kinds of skills that are behind why we care about this in the first place. Furthermore you will have demonstrated initiative and drive, the grit that we look for most of all in our graduate students.  Good luck!

Why you shouldn’t say “data not shown” or “personal communication”

What makes something science is not so much the subject matter as the process. Scientific information is obtained by clear methods that others should be able to repeat. It is above all based on evidence. There are lots of different kinds of evidence and different ways of analyzing and interpreting it. Show what your evidence is. Tell how you analyzed your raw data. Explain how you interpret these data in light of theory. This is the scientific approach. It is not magic. It is not vague. It is not private.

It is susceptible to disagreement if others disagree with your methods, your analyses, or your interpretations. Part of that disagreement should not be because you had access to something secret, or your critic had access to something secret.

This is why you need to resist the temptation to refer to data you are not sharing at this time, or to communications that are not public. Science does not work by deferring to authority. It advances with data, analysis, and interpretation. These three things reveal theories that are supported. These three things are the way in which we reject theories, no matter how eminent the person advancing them is.  This is a wonderful thing about science. It is the source of all its power.

But life is complicated and you may have a piece of a puzzle that impacts a publishable study that itself is not quite ready for publication, or that belongs to a different set of co-authors. You may want to say personal communication to acknowledge an idea that was not yours. But you should not do either of these things. Restrict your story to the facts that are revealed. Save the rest for later.

I really like it that the US Proceedings of the National Academy of Sciences, PNAS, explicitly disallows these practices: “Data not shown and personal communications cannot be used to support claims in the work,” from PNAS editorial policies.

Who gets to tell the story?

Why didn’t the writer of that important new review cite your work? After all, it is on exactly that topic. This is unfair. It is easy to feel aggrieved. Why should you bother to work so hard on your research when the big boys just ignore it. What to do? Fortunately, there is an easy answer: start writing the reviews that define the field yourself.

When you write a review, you can put your work where it belongs in the scientific narrative. But be careful! The best reviews are the most fair ones. They cite as widely as is appropriate. They have a new and clear perspective, or they extend an old perspective. Many journals these days put reference limits on reviews, so work hard to cite the most relevant ones, mixing between the very newest and the historical papers that once defined the field. Don’t look for revenge by ignoring those who once ignored you. These are the scientists you want to win over to your perspective.

When you write your review, you should have something new to say. This takes work. But it is work that is rewarded, for it not only helps put your work in a worthy place, but it can also lead you in new and creative directions. Graduate students should be involved in reviews, writing or collaborating on them as they become familiar with the literature, provided they can write a fresh, new review.

Olaya Rendueles, a postdoc in the Velicer lab in Zurich told me that a number of her fellow grad students in Paris wrote review papers very early in their graduate careers because the short 3 year program required at least one publication. That would be hard to do with a data paper in such a short time.  I did a quick search and found an interesting review that she wrote entitled: Multispecies biofilms: how to avoid unfriendly neighbors. It looks interesting, but I couldn’t download it from here.

Steve Frank

It will still be true that the reviews written by newcomers will take longer to have an impact than the reviews written by leaders in the field. Just remember, those leaders got to their positions by having interesting things to say. I can hardly wait to read Steve Frank‘s latest piece on kin selection, for example. I also love the way he makes everything so accessible, retaining the rights to his books so people can download the PDFs freely. This didn’t stop me from actually buying the books for the convenience of the format.

There are other things you can do to get your work noticed. Attend various meetings. Give talks and posters. Approach scientists you admire and engage them in conversation about their work. Then work it around to yours. We scientists are social animals and remember best the work of those we know.

To become a leader in your field, you need to do the important new studies, casting pretty shells on the beach of science. But you also need to gather the shells of everyone together and put them into a pattern. Then you will get the attention you deserve.