Slashdot Mirror
Beer-Drinking Scientist Debunks Productivity Correlation
austinpoet writes in with a blog post debunking the theory we discussed a few days back that scientists' beer consumption is linearly correlated with the quality of their work. Chris Mack, Gentleman Scientist and beer drinker, has analyzed the paper and found it is severely flawed. From his analysis: "The discovered linear relationship between beer consumption and scientific output had a correlation coefficient (R-squared) of only about 0.5 — not very high by my standards, though I suspect many biologists would be happy to get one that high in their work... Thus, the entire study came down to only one conclusion: the five worst ornithologists in the Czech Republic drank a lot of beer."
beer > coffee/caffeine
Gone!
C'mon, I thought the (ancedotal) evidence proving(?) that beer is and isn't good for productivity is adequete! It should say that beer, in certain levels, is good for productivity, and in excess ... it is bad. Really, people write papers to prove this?
I think it's safe to say that the paper they are "debunking" was meant as a joke.
Scientists Claim there is a direct correlation b/w pot smokers and an amazing talent to link string theory with life on mars
too scared to forget random user names
It's more about the quality of their beer. Not that I have anything against Pilsen. I think they make a perfectly fine beer over there.
What?
So beer may or may not hinder a scientist's creative abilities. On the flip side, will scientists ever start taking drugs in order to improve their skills? Would this ever lead to drug testing researchers that announce amazing new scientific breakthroughs? (sort of far fetched but an interesting idea nonetheless).
This has to be a lost Monty Python sketch, right?
"Flag on the moon. How did it get there?"
More research is needed.
So it's safe to drink beer again. And to think I was actually going to cut down!
They've looked for a linear correlation, so if what you've said is true then the analysis they used wouldn't find it.
In order to find a correlation where the input IV (beer consumption) has an optimal value, you would have to do the regression on a transformation of the variable. Perhaps a quadratic would suffice, or else abs(X - k) for some unknown value of k.
The comic xkcd was there first and called this effect the Ballmer Peak. Most likely, this effect was also tried in Vista and Vista SP1 design meetings, but the balance was all wrong and didn't come out as (they) expected.
I know scientists who devote their entire lives to their work, never go out, never have a good time, have no children (or never see them), etc. etc.
Are they *better* scientists? I don't think so.
Are they *more productive* scientists? Not in every case, but on average, yeah, I'd say they are. There are situations where spending all your time on work and neglecting other aspects of your life is a self-defeating proposition, especially in creative work (which generally includes science, although what scientists actually *do* varies a lot from one scientist to another.)
But burn-out aside, if you're willing to sacrifice other aspects of your life, you can get more science done. Pretending that this is not, generally speaking, true, because you want to pretend that it doesn't cost you anything to have a life, is not productive.
That said, the article-author is right about the statistics. Bad Czechs!
The good and new comes from no quarter where it is looked for, and is always something different from what is expected.
I had a friend who always cracked open a cold one when he sat down to work (while at home, of course). I could never understand it - but he worked like a maniac. And he did it for years... until he failed a drug test and was fired. He was a manager for a large pharmaceutical manufacturer. Go figure...
If you were one of the five worst scientists in a field in the Czech Republic, you'd probably turn to drink, right?
it depends on the kind of drinker you are, do you drink moderately and only open that first alcoholic beverage later in the evening (after supper)? you know anybody that pops the top off any alcoholic beverage too early in the day and drinks excessively until they are slobbering & stumbling recklessly wont be a good anything (especially a scientist)...
i drink a mixed drink every evening after supper daily and only one, using a shotglass to measure the amount, i do enjoy a mild buzz but i hate being drunk and i dislike drunks since they can cause lots of problems (loss of careers/jobs, wrecked marriages, even cause fatal traffic accidents on the road)...
moderation is the key...
Politics is Treachery, Religion is Brainwashing
They gave him the bird!
I swear to God...I swear to God! That is NOT how you treat your human!
If, on the other hand, he means the correlation coefficient r=.5, that means that R^2=.25. Still, a quarter of the variance in "work quality" is explained by beer drinking. That is still very high.
His point about outlying ornithologists and the points not being independent may still be valid; determining if they are is an empirical matter. Do these outlying scientists, in fact, socialize together? What other sources of nonindependence might there be, and do they affect THIS data set? Also should we really claim that 5 out of 34 (15% of the sample!) constitute OUTLIERS? Those aren't outliers, those are a subpopulation.
He didn't debunk the study; he rather raised some interesting questions.
R-squared is the amount of variance accounted for by the variable in question. That means half their productivity is explained by beer drinking, and half on all other variables combined.
As a comparison, 0.3 is pretty much the top end R-squared in personality psychology. that field is built on correlations that account for no more than 10% of the observed variance.
To combine the two, it's far more likely that TFA didn't actually measure beer drinking, but rather how much beer those scientists who drank beer would admit to drinking. Those who'll drink it are probably more likely to relax, which will make them more productive, and those who will admit it are less likely to fall prey to negative opinions of others, a major source of which is reviewers' comments on papers submitted for publication. Such comments are often undeservedly harsh, and in many cases coming from someone who doesn't know as much as the author about the topic. That can turn away those who place great store in the opinions of others, especially perceived authorities.
Next, on to Russia and WOTKA!
"I may be synthetic, but I'm not stupid." -- Bishop 341-B
*burp*
"So long and thanks for all the fish."
Quote: "In social science .5 is huuge!"
Reply: I respectfully disagree. r^2 = .5 is a good correlation, but it is not huge, even in the social sciences. Not to mention that in a study such as this one, there are some serious lurking variables which are most likely not accounted for. If you were to control beer drinking with another variable (say, time spent in the office or some other better variable), I would dare say the t-stat would be entirely insignificant.
Quote: "so if these findings actually show causation (which, admittedly,I they might not)"...
Reply: They definitely do not show causation. This is an observational study, not an experiment. No observational study can show causation, only correlation. To determine causation, experiments using factorial design and variable controlling techniques are a must.
It beats the daily Microsoft Windows Vista article(s). Those don't get interesting no matter how many beers you drink before reading them.