Monday, August 3, 2015

Fear of missing out

There are so many tools, techniques, communication platforms and methods out there for ecologists that it can be hard to keep track of them all. Whether it be LaTeX, Linux, or Twitter, someone has probably suggested that your research life will improve if you use their favourite tool.

Online science has allowed fantastic networking and information exchange. However, one side effect of the increased usage of social media can be the feeling that you can’t keep up. The phrase ‘Fear of Missing Out’ in social psychology was coined to describe the anxiety arising from looking at social media and fearing that you have chosen how to spend our time poorly. But this can extend to academic working life too – that nagging fear that your colleagues are doing cooler things (going to more conferences, using more cutting edge techniques, publishing more papers) than you. It can be disconcerting to see that there’s a cool R package you know nothing about, or to receive yet another invite to a work-based social media platform you’ve never heard of.

There are people who are the early adopters and the technological trendsetters; people who derive their success from their methodological/technological skill sets. These people are fantastic, since they provide the examples for others in the field to follow. But for most researchers, it is practical to recognize that knowing every new technique and tool doesn’t promise success.

Graduate students may suffer this fear of falling behind most intensely, given how closely it relates to imposter syndrome and the general skills gap that grad students have to overcome. But I’d argue for most students, that balance is key. Otherwise it can be a slippery slope: maybe you learn R, then find that people recommend knitR and R Markdown, then you learn Shiny to show your results beautifully, then someone suggests things would be faster in a lower level language – and maybe months have gone by. That’s not necessarily bad, intellectual freedom and self-teaching are some of the best parts of graduate school, and if you are going to learn a bunch of new techniques, that is the perfect time. But aficionados will make arguments for everything from Python, Linux, Emacs (apparently the one true operating system), Github, to LaTeX/BibTeX, etc, etc. And every one of these is a useful tool, but maybe not always a necessary tool.

Researchers can never learn everything, and some things will fall out of fashion as quickly as they arose. Sure, it is less than ideal to be a researcher who hasn’t learned a new approach since the advent of SAS. But we need to balance necessity (is it fundamental to my research?), the time investment, and the utility of these skills for future endeavours (e.g. if you are applying for a data analyst job after grad school, learning a few languages might be smart - but if you will rarely use it again, perhaps it is not so important). The danger for many students is that either the time investment in learning a less-than-fundamental tool is very high, or else they learn a number of tools but never master any of them. This isn't to say you can shirk on quantitative skills - on the contrary, having those skills is far more important than mastering any specific program or language. I can think of fantastic ecologists who don’t know what Twitter is, who still use SAS, who have never learned R, and who produce brilliant work. The questions, the ideas, and the knowledge matter too, after all.

(*I should note that I'm a quantitative/theoretical ecologist, and I think every ecologist should have a strong quantitative education! Just that you can do that using lots of different methods and doing it well should always be the first priority.)

3 comments:

Anonymous said...

Hi,
You are absolutely right. I've obtained my PhD in 2013 and I am just completing a one-year postdoc position. During this year I've been confronted to all the new computer methods that ecologists are using, figuring I should master as many as possible to keep in the move. I quickly got into panic and despaired about not beeing able to cary out my scientific project and conjointly find time to get all those skills. My sad feeling is that an ecologist should now be as well a statistician and a data scientist. Just look at the post-doc positions that are offered. What is needed to get a position are first of all "outsanding quantitative skills".There are always more positions in ecology that are opened for data scientists with ecological knowledge as a desirable option... Is that just my feeling?

Andrew MacDonald said...

WOW. thank you for writing this! Fear Of Missing Out is definitely a thing -- in life in general, but perhaps particularly in science. I mean humans probably have always felt it to some extent -- since inevitably one dies before one has experienced most things -- but we've never had a means for **everyone we know** to show us pictures of **everything they are doing**.

I think its worse in science because it is such a complicated pursuit, and so many skills go into composing an excellent scientist. Programming, knowledge of the literature, taxonomy, public speaking, writing, lab/field technical skills. There is no limit to how good you can be in any of these pursuits. Many (most?) good scientists can do them all, but not all are essential. How good should you be in any one of them? Graduate students get extremely conflicted advice on the subject.

That said, I think you might unfairly be criticising the computer side of things. There are so many slippery slopes waiting for graduate students, and every tool is "less-than-fundamental". How can we choose what to learn? (then again, perhaps I am just being defensive because i love R/Rmd/Github/Twitter)

Caroline Tucker said...

Hey Andrew - great comment, I agree totally about the pressure on scientists, and grad students in particular, to learn every possible skill.

I'm definitely being a bit imprecise in mostly mentioning the computer side of things (and wanted to avoid everyone feeling the need to defend their favourite statistical approach or writing technique...)! FOMO absolutely extends to every aspect of our lives of scientists (and seriously, kudos for mastering and enjoying all of R/Rmd/Github/Twitter - I swear i'm not a Luddite and heavily use computer tools and platforms (R, Mathematica, Github, Python, Twitter, blogging)...).

This post partially came from something I was at where students were presenting some research looking for feedback using Rmd. They had been to a session on Rmd a while ago, and it's the cool thing to do I guess (in part b/c they've seen it used really well). But the time that was spent troubleshooting their code that wouldn't work, and that displayed their data imperfectly just seemed to me to be better spent actually talking about their research (even if they had to use an uncool powerpoint). Rmd is pretty useful, but it seemed like in this specific case, not necessary. So that's what I had in mind.