Quantiphobia

Standard

illust 4 copy

A little while ago, the webcomic Piled Higher and Deeper–a source of consolation for all of us in the low/middling ranks of academia–posted a comic titled “The Spiral“, in which one of the recurring grad students cannot separate her self-worth from her data. When the latter is proving difficult, she transfers confusion about analysis to inner turmoil and self-loathing.

That simple story arc, succinct in its four frames, describes my lowest moments as a scientist.

I think it stems from my deep-seated fear of numbers. My father is a mathematician by training; my sister is a math teacher and Euclid aficionado; I’ve attended schools that emphasize quantitative skills; and I’m pursuing a profession path where progress hinges on the empirical testing of hypotheses. You’d think by now immersion/exposure therapy would have run its course, that I’d begrudgingly love numbers…but no. I’m still petrified.

Words, I love. I feel safe around words. Words allow prevarication. With words, you can dance around reality, or enhance it without strictly violating truth. Numbers? With numbers, there is a correct answer. In mathematics (and in statistics, ±95% confidence), you are either right, or you are wrong.

I hate being wrong.

What does this mean for me as a scientist? I go through a field season. The data is entered and proofed. I sit down in front of my laptop, thinking about my a priori hypotheses. These hypotheses are quantitative at their core, but are carefully cushioned with words, “…the presence of study species is correlated with x and y, with a potential for interaction of x*y, is tested against the null hypothesis that…” or the appropriate model-selection version of such verbiage. I type in the code for the test, and, heart full of hope and hit Enter to run my script.

Slide1

Or I don’t.

Instead, I scroll through csv files, update my R packages, open a stats manual, or go make a snack. Usually the snack. Something chocolate. Because I know in my heart of hearts that the moment I hit enter, something will go horribly, terribly wrong.

Slide2

Usually, it is something in my code; comma I forgot to delete, an open parenthesis. Something that I will spend hours looking for, only to have a colleague glance over my shoulder to point out what my dyslexic brain didn’t see.

But my bigger fear is that I’ll discover that I am simply wrong. When I hit enter, not only will I discover the data contains no publishable results, but that I’ve misunderstood mixed effects models, or that my experimental design was crap, or that my hypotheses were inane, tautological, and just plain bunk.

Slide3

Sure, null results are important. Certainly, I’ve performed similar tests before. And my awesome advisor made sure our experimental design was solid. None of that matters: I know that my career will be over when I hit enter.

My fears are not unique. Countless academics report signs of impostor syndrome, convinced none of their past successes are genuine, and that it’s only a matter of time before they’re discovered to be frauds (more on that here and here). Moreover, ecology departments are peopled with graduate students that blanche at the site of an integral and blame their love for REI for leading them into the sciences. Most people with similar phobias deal with their demons with more courage than I do, and are productive in spite of them.

Ultimately, I can’t procrastinate on analysis forever. Advisors and reviewers and funders require timely results, and my lovely words–so handy at crafting excuses–cannot postpone deadlines indefinitely. So I do hit Enter.

The results that arrive (after I fix my cruddy code for the umpteenth time) are unsurprising. There are no trends, there are weak trends, and there are stronger trends. Some models perform better than others–one or two significantly better, if I’m lucky. Scrolling through results summaries and unformatted data plots, I can see how some methodologies could have be improved, or how a particular variable was confounded by another.

And you know what? Being wrong isn’t that bad. Yes, inconclusive results and weakly significant correlations are pains in the a**. They might mean Science Claus won’t bring me a fat impact factor by Christmas. But science is all about proving yourself wrong a hundred times for the sake of figuring out what is correct…and it’s about living at poverty line until you do produce something publishable, in which case you might receive a postdoc at 35 and realize you won’t be able to send your kids to college unless you marry into money.

Moral of the story: Any wealthy suitors out there? I cook, I clean, and I’m great at parties.

Slide1

 

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s