Progress

I went to the Mammal Society spring conference last weekend, a two-day mammal-fest at which I presented some of my work on mapping the distribution of UK mammals. It was my first time at a Mammal Society conference and I really enjoyed it.

I am used to much bigger conferences so with no parallel sessions this one felt more like a symposium. This has both advantages and disadvantages of course. On one hand you get to see all the talks so no worries about missing something interesting and after a while you know people’s first name and they know yours. On the other hand, if you have insufferable people that love asking questions or simply making comments based on their infinite wisdom, then you are stuck with them.

Overall I have to say I was quite impressed by the level of the talks and by the pointed questions asked after each one of them. Although, to be honest, I am inclined to forget not so good talks if they have loads of pictures in them. Obviously a mammal conference is bound to not disappoint in that department. So now I really want to work on hedgehogs and dormice and rabbits…

One thing I noticed about myself is how far I have come in several key skills. First in terms of giving a talk; I think I have always been able to control my nerves to an extent by practicing my talk over and over, but I can distinctly remember feeling much more nervous that what I felt this weekend. While it could be a fluke, I’d like to think that it demonstrates that I have levelled up in this skill. I don’t think I will ever stop being a bit nervous but it’s nice to see that with a bit more experience comes more control.

Moreover the thing that used to terrify me even more than giving a talk was answering questions about it. Last weekend, even though my presentation contained some controversial elements, my feeling was ‘bring it on’, kind of like playing a game of ‘how nasty can questions be’ (actually they weren’t).

Then I have progressed in terms of asking questions to other speakers. I know some people, even junior people, are very comfortable with raising their hand after a talk. It’s always been another excuse to be nervous for me, and I usually avoid it even if I have a good question in mind. Not this weekend though! I asked several questions without trembling too much…

Finally I am getting better at networking or at least engaging people I don’t know in conversation and forcing myself out of my shell. Not only did I talk to people at lunch and tea breaks but I marched myself down to the dinner and sat at a random table when my first instinct was to not go at all. Being in a room full of people I don’t know in a social setting is usually the definition of a nightmare but this weekend I did it and chatted away with some very interesting people.

So overall, this experience was a win on a professional and personal level. My talk was successful and I met interesting people, and I pushed the boundaries of what I usually can tolerate.

Advertisement

Intuition

Do you think the instinct for golden research ideas (i.e. not just good but really high impact good) is something that can be learned? Or is it something that either you have or you don’t? I suppose that it might be a bit of both- some people are naturally good at seeing the big picture instantly and sniffing out great prospects, but most people get better at it over time and with experience.

Besides the fact that as your career advances and your reputation grows people are more likely to pitch their best ideas for collaborations to you, I think that it takes many successes and failures (in pursuing ideas and trying to publish them) to “calibrate” your intuition to the point where you can say, just from a simple conversation, whether it will be worthwhile* to be involved in a project or not.

I have a colleague who seems to have collected a random assortment of cool data over time, and alongside it, plenty of ideas of what could be done with it. He is the kind of person for whom you think “I really want to work together, the specific subject doesn’t really matter”. We all know people like this, don’t we? And he wants to work with me as we have complimentary skillsets.

However, I haven’t quite honed in my intuition yet, so how do I determine what is a worthwhile pursuit for me? The first step, I suppose, is to define what output I consider being worth my time. Is any peer-reviewed publication better than no publication, regardless of where it gets published? At this stage of my career I am inclined to say yes, as long as the subject is broadly related to the main subject I am interested in. But then, there is no doubt that for side collaborations, there is a direct correlation between how much time I want to invest into a project and how high the impact of the final** journal is.  That’s when a bit more experience would be very useful I think.

For now, my immediate response is “why not but…” For example: why not, but can I see the data?; why not, but should we also involve this person who is an expert in that particular subject?; why not, but have you thought about these possible caveats/limitations/confounding effects that may prevent us from drawing conclusions from the data?

__________________

* what is worthwhile will be different for different people
** where it gets published as opposed to where one would like to get it published

High and Low

If there is one quality that all scientists, no matter what their discipline, must possess, it is resilience. Your dog just died? Oh well, have some nasty comments by well-intended reviewers on your beloved paper. You’ve just received a massive bill and have no idea how you’re going to pay it? Don’t care, that grant you spent the best part of a month writing, rejected without any constructive criticism.

So yeah, resilience is something that we need to learn and learn fast. I am not complaining, really, because that’s one useful skill for surviving what life throws at you in general.

What is more difficult to deal with, in my opinion, is the whole uncertainty that goes hand in hand with academia. One week, you might have two grants under review and two papers that, fingers crossed, might get published in high impact high reputation journals. Even if you get only one of each through, you can see how it will positively impact your career prospects, maybe give you a fighting chance at that job you desperately want?

Because that’s another thing we have to do, thinking of the long-term career impact of what we choose today. We know that between the submission of a paper or a grant, and any outcome, positive or negative, months will pass, sometime a year, maybe more?! So when things are under review, and that “yes” is still an option, it’s easy to see how your career will benefit and dream plan accordingly. But ignoring the uncertainty doesn’t make it go away.

Sometimes, the very next week, you’ve received the dreaded no(s). The hard part, I think, isn’t so much the short-term effect; you have to rewrite and resubmit whatever came back and that’s ok (you are resilient!!). I think it’s the long-term impact, because every rejection pushes back that career boost you are ultimately looking for. It’s going to be another x months before you are in an improved position, and you don’t know when, and you don’t even know if you will eventually get there. Yet, you have to keep at it.

Despite this, I love being a scientist because there are many highs alongside the lows.  In fact, if I look back at the past few months, there has been some really good highs; I received an award for my PhD from the RSPB, and one of my paper was highly commended for Journal of Applied Ecology’s young investigator award. So, I apologise if this post feels a bit gloomy because in truth, life is fairly good right now and no, my dog has not died. Also I don’t have a dog, just adorable cats*. I guess the months of winter have also been a bit dark at times, and I didn’t feel like writing about it then. I don’t know about you, but I am ready for spring.

________

*Sorry I coudn’t help it

Let’s get technical

I am not going to lie, I consider myself an R nerd. I love it. Anything I can do in R, I will,and would not be caught dead using another stats package (with the exception of OpenBUGS….maybe).

Lately my work has involved a lot of geospatial work and, unfortunately that entails using ArcGIS. Of course, being an R aficionado, I have been doing as much as possible in R and complaining a lot about ArcGIS- I mean who wrote those help files?? Unfortunately, there are still times when ArcGIS is the best choice and me, not being content to click away, I’ve decided that I would learn Python.

I didn’t know anything about the language but it turns out that Python is not just great for geospatial work. In fact it is a versatile, powerful and object-oriented programming language that can handle pretty much everything.

Of course, I am not sure that anything can beat R in terms of statistical modelling* but now that I’ve spent a good amount of time learning the basis of Python, I am wondering whether I won’t eventually switch to it for quantitative modelling. At my current level, it is still not efficient enough; for example, it currently takes me about an hour to write a fully functioning individual-based model, perhaps half a day if it is spatially-explicit** and there is no way I can beat that in Python. But eventually??

I’ve seen some posts about people switching from Matlab to Python but not really anything about switching from R. The often-cited advantage of Python? Its versatility and flexibility, and the fact that it’s free, which is not an advantage over R obviously. The whole object-oriented design is what’s appealing to me.

If you are already comfortable with a programming language, picking Python’s basics is going to be fairly easy. The syntax is slightly different but the reasoning is absolutely the same so you can get if statements and for loops running pretty much instantly. It’s the particular features that are harder to pick up, like event-driven programming and the elusive class system- and the bloody indenting! Also, I’ve been trying to install libraries on a computer where I don’t have admin rights, and let me tell you that was not easy. Eventually, I prevailed, using source code and a command shell. This is not my favourite part of the Python experience even though I love a challenge.

How am I learning this new language then? I’ve done a lot of exercises online, and I am now doing the coursera “Introduction to Interactive Programming in Python” course, which I highly recommend as it is good fun***.

So, any ecologist using Python as their main language who wants to share their experience and advice? Sound off in the comments or on Twitter @AChauvenet.

____________________________________

*although Python does have a library or two that lets you use R’s functionalities

**when I started in R, for my MSc, this simple IBM took me a few weeks to write! I love seeing progress like that.

***Disclaimer: I once translated some code between matlab and R just for fun

A post full of questions (and it’s about co-first/co-last authorships)

Have you heard of this new* practice of having co-first and co-last authors on scientific papers? Well I had, and yet when I found an example of it in recent ecological literature I was baffled. I had thought that this was becoming populaer in other distant fields like astrophysics or biomedicine, but ecology? I didn’t think it was done.

Admittedly, I have not been looking for it when reading papers so I could have missed other recent instances. This thought got me curious. Have I been blind to these “equal contribution” statements? If not, does it have to do with the journal where I spotted this anomaly (let’s call it publication A) ? Having never submitted a manuscript to this particular publication, I couldn’t be sure whether they actively offer the option to name co-first-last-nth authors and thus perhaps encourage the practice. So, I had a look at a few recent issues and found 5 out of 85 published articles in publication A with co-first-last authors- that’s almost 6%.

Then, because I can’t really justify checking every major ecological publications for instances of co-something authors, I decided to take a short-cut and check out all 2012 and 2013 papers in my EndNote library. That’s 108 papers in 40 different publications by the way. I can’t say for sure that it is a representative sample of the wider literature but it was the least painful way to assuage my curiosity. I found two, including the one that is the inspiration to this post (that’s c.1.8%). While I can’t say whether it is significalty different than the rate specific to publication A, at least it explains why my instincts telling me this is not common practice in the field.

Why did this paper catch my eyes in particular? Not only does it have two co-first authors but it also has two co-last authors. This is out of 7 co-authors listed. While I can see situations where you could genuinely have two first and two last authors, or at least cases when nobody can agree on who gets those coveted places and thus a polite consensus is reached, this situation remains pretty strange and makes me wonder where this trend is taking us.

Since I am assuming co-first or co-last authorships are a fairly recent development*, I don’t think proper conventions having been ironed out yet (although Nature seems to have a policy in place: two co-first-or-last authors max). Interesting fact, in the search detailed above, I found instances of up to 4 co-first authors. Where does this end? Are there rules? Can we all be co-first and co-last at the same time?Perhaps we all get the “equal contribution” treatment and remove last authors with a simple footnote? Maybe we all want to be co-last instead.

Some might think that it is a great opportunity to avoid conflict with your co-authors. Why not eliminate the ranking altogether? Let’s not quibble and agree that we’ve all contributed in different ways but in equally important manners. Except, has this ever happened on a list of more than two or three authors? Are we perhaps to aim for equal contributions then? In my view, and this has nothing to do with being a control-freak of course, this will be a case of too many cooks in the kitchen. Meshing styles when co-writing a paper is hard enough but how do you suppose we’ll be able to co-analyse data? Are all authors to redo the same analyses for the sake of it? I can see it: an R code being passed to everyone on that list, and little bits and pieces being added, some being removed along the way… That’s sound like a nightmare I could have.

Anyway, why does this matter at all? Why not get rid of the more traditional ways and adapt authors list to every situation. Well, even if we find the right conventions and guidelines on how to attribute co-whatever authorships, how are CVs going to be able to cope with these? I guess, every listed publications will have to have its own footnote….

__________________

* I think it is a fair assumption but I could be wrong

On reviewing part 2: ethics and etiquette

Useless fact of the day: I remember reading a great post on how to be a good reviewer a few months back but I can’t seem to find it again.  Oh well…

I don’t remember it taking about dealing with the non-anonymity of the authors whose short-term success you have so much power over (I know what it’s like to get a scathing review or even a fair rejection). While I understand why peer-reviewing is rarely double-blind (you need to be able to see whether you have a conflict of interest with the authors), it seems to me that knowing who the authors are may cause several problems.

First, while I am keen to not get involved with papers where there is a conflict of interest, I can see how it could be tempting to the exact opposite and do your best to sabotage not facilitate the publication of a paper that is (1) from a “rival” lab, (2) co-authored by someone who has slighted you professionally, (3) co-authored by someone that has spent years trying to get your own work down… These may be moot points as most people will tell you that it’s easy to guess who the senior author/lab is on a particular problem or system. Yet, I feel that not knowing the authors and their place in the list could limit the risk of malevolent behaviour. I think we’ve all felt that perhaps we’ve been on the receiving end of something like that at some point?

Second, and this could be related back to my desire of collecting reviews for more and more prestigious journals (see reviewing part 1), it can be ego-boosting to be asked to review the work of a big name. That’s great if this causes a really thorough and fair review. However, it could also lead to a mini power-trip where you feel your duty is to find lots of flaws or perhaps no flaws at all if you admire that person. Additonally, I feel that it’s probably easier to find willing reviewers for papers co-authored by big names than for other. Of course, not having any editorial duties, I can’t say that this is true, only that I suspect it might be.

Third, how do you deal with the urge of googling the authors? That’s something I suffer from, especially if the paper isn’t as good as it could be. It’s natural for a mind used to looking for patterns and explanations, to want to explain why some stuff were done that way and not another I think. If I really want to see the publication record of the authors, or perhaps whether they are actually known for what they are writing about, I make a deal with myself: I am only allowed to look after completing the review and it must not influence what I write unless it’s to soften the blow.. Is that unethical?

Ah… if only scientists weren’t humans too.

On reviewing part 1: coincidence?

An important part of our role as scientists is to peer review other people’s papers. While this sometimes exhausting task may seem like a burden to more senior member of the scientific community, I am still at a stage where receiving an invite to act as a reviewer is like getting a present. I particulary like adding new publications to my collection of journals I have reviewed for. Although I haven’t gone as far as making a real game out of it, I am pleased when (1) I get a new one and (2) the invitation is from a journal I perceive as being more ‘important’ than the ones I already have; this importance being measured both in terms of Impact Factor (I know it’s bad) and perceived reputation. Don’t judge me.

A more sensible reason for liking receiving those invitations is that it means my name came up as a potential specialist on a subject. Of course, I don’t discount the fact that some of those invitations may be the result of a colleague suggesting my name instead of theirs, and not an editor (or perhaps associate editors?) finding me on their own. Still, I take invitations as a good sign of my profile rising.

There is a weird part though. At this stage of my career, invitations are quite sparse (sometimes months go by without one) but it seems that when I receive one, the likelihood of receiving another one within a few weeks increases: they come in clusters. For example, number this month: 3, number in the last six months: 3. Does anyone else experience this?

Now being a good little quantitatively-minded scientist, I decided to search for this pattern in actual data… I thus plotted my invitations (at least those I could trace back, which is most of them) as a function of time.

invitations to review versus time since first record

Ok, so at first glance, my gut feeling about clustered invitations may not really hold when confronted with data…. But I have encircled invitations received within 60 days of each other and perhaps with more data there could be something there? In any case, it is still irking me to get nothing for months and then suddenly three within a few weeks. Coincidence??? Probably. Conspiracy theory? Unlikely but nothing is impossible. Result of an editorial process I am not aware of? Perhaps.

What sort of non-random process could cause this though? I suppose that if all clustered invitations are on a similar subject, a same-ish search for someone to review these papers could yield my name every time. However, that would only make sense if papers on similar subjects were solely submitted at the same time and not throughout the year. That is doubtful, and, anyway, those last three invitations were absolutely not on the same subject.

Any thoughts?

On publishing during your PhD

A few weeks ago I mentioned that I had a paper coming out soon. Well it is finally out in Journal of Applied Ecology. I am not going to go on and on about it, but go have a look if you’re interested in conservation under climate change and assisted colonisation as an adaptation strategy. Also, it is free open-access!

This paper is the last one from my PhD, which means that now every one of my chapters have been published in peer-reviewed journals. Considering that I submitted a year ago, one could say that I managed to get rid of it publish all of my thesis fast. This was helped by two things: (1) I started publishing during my PhD and (2) this last paper is actually a combination of not one, not two, but three chapters.

When I finished, last August, I had about a month before starting a new job and I was faced with two choices: keep all three chapters separate, which meant 3 papers on my CV (yeah!) but most likely each with a lower impact (not so yeah), and a long time before reaping the benefits OR combine all three chapters into one higher impact paper, reduce the workload a little, and the hassle and “time to benefits” at lot. Following the wise counsel of my main supervisor, I went with the second option. In hindsight, and with a Journal of Applied Ecology paper under my belt, I think this was definitely the right choice.

The key thing, though, is that starting to publish during one’s PhD is a must. At a time when excellent young researchers are struggling to find jobs, having publications on your CV at the end of your PhD –and not just the “in preparation” smokescreen- is very important. In fact, even published (or accepted) papers are probably not going to be enough to get you a job but at least you’ll get an interview.

This seems to be obvious doesn’t it? Well wrong. In ecology, traditionally a field-based discipline, you still see people finishing without a first-author or even any kind of publications to their name. The usual excuse? That person had so much field work they did not have time to write during those 3 or 4 years- but look at all the awesome data they collected (which just happens to benefit the lab as much as the student collecting it)! The truth is, I am tired of hearing the old argument that it was easy for me because I didn’t have any fieldwork. Do you think I sat there doing next to nothing during those 3 years, except write up chapters/papers? There is only one thing that was made easier by my lack of field work, I was able to work on collaborations outside of my PhD project and this added to my experience and CV even if it wasn’t first-author stuff.

At the risk of sounding righteous rude inexperienced, there is no reason for supervisors not to push their students to submit papers during their PhD. And yes, the cost of adding  these extra final unplanned months of fieldwork to a student’s PhD most likely outweighs the benefits to that student, despite all the amazingly cool data they will collect and the fact that the lab could really use it. Talk about one bearing the costs and every one else benefiting! I am not a behavioural ecologist but that does not sound right.

Disclaimer: this rant was brought to you by someone who does not supervise PhD students at the moment but had a great example of efficient supervising during her PhD.

The art of being a scientist

It is easy to see that a PhD is like an apprenticeship for researchers. This is the time when, besides specialising in your subject, you are supposed to pick up those “basic” skills that are common to the whole profession regardless of the subject*. The ones that comes to mind of course are scientific writing, or how to structure a paper, how to handle the process of peer review from the point of view of the author and the reviewer (if you are lucky), how to write effective presentations which are entertaining and well-timed, and working within deadlines.

I got all that of course, and more. I am not sure how common this is during the PhD but I also was trained in the art of collaborating. From unofficially co-supervising students to being part of a larger team working on a single project, I learnt a lot of what to do and what not to do. However, there are two things I was advised to do but wasn’t too keen on for different reasons. Of course, when advice comes from well-meaning experience you should probably listen to it….

Here are the two bits of wisdom that I thought *did not* apply to me and now I may or may not be feeling the consequences.

Set deadlines: not for yourself (that should be automatic) but to your collaborators (including supervisors). If you send a piece of work to someone for them to work on, give them a deadline by which it should be done or their input may not be taken into account.

Putting aside the fact that giving a deadline to your supervisor might feel a bit too brave, I was stupid naive enough to think that people will always get back to you in a reasonable amount of time, you know, taking into account what your personal constraints are without having to voice them. It’s not that people do not want to be reasonable, truly, but that what they consider a reasonable amount of time to sit on a paper will never be the same as you.

Establish your role from the get-go: so you’ve been asked to co-supervise a student, ask how much your input will be worth and where it “ranks” compared to the other co-supervisors. If you’ve started an exciting project with someone, set clear guidelines as to who does what, and establish your rank on the authors’ list (we all know that’s the end game here).

For co-supervising this is especially true if you share some expertise with the rest of the supervisors and are thus likely to advise on the same piece of the analysis. Turns out that two people with similar expertise are unlikely to agree on a simple piece of work, surprising isn’t it? Sometimes you can both be right, just heading into different directions. Since you well can’t stomp your foot like a child and demand that everyone bows to your point of view, establishing “ranking” beforehand can avoid a lot of pain, and back and forth.

For the “grown-up-only” collaborations, because it may feel rude (at least to me) to ask about or perhaps demand, your place on the potential authorship list, it is easy to ignore the problem until later. However, trusting that colleagues will recognise the value of your work and that the rightful place on the authors’ list will be magically ended to you when the time comes might be a tad optimistic.

*I am sure exceptions do exist.

Back from holidays

I have been back from holidays for a couple of weeks now and I know it’s been a while since I have posted anything but I am still mulling over a couple of new posts.

So …. in the meantime, I thought I’d share with you pictures I took a few days ago when I went along to a bat survey. Since I don’t get to go into the field for my own work, the opportunity to see this survey was quite super exciting for me (I mean, flying mammals*!!). Even though I had been up since 5am that morning, I could not resist the chance to see those beauties up close and personal.

These are Natterer’s bat by the way! For those not familiar (I know I wasn’t…), they are about 5cm long and weight 7-8g so quite tiny and delicate (the Bat Conservation Trust says they can go up to 12g but I didn’t see any that big).

*According to Wikipedia flying mammals are not so rare since cats (!!) are one of them by the way….