I am biased…but so are you!

All of us who do science walk around with a lot of baggage.  So when we encounter a new piece of scientific work, we are biased about it when we pick it up. The key is what we do about this.

“Now wait” GG hears you cry “I always keep an open mind when I read a new paper.”

Really?  So you pick up a piece of young earth creation science and say “yep, as likely to be right as wrong.” Either you are a saint of some kind or lying to yourself.

So let’s back up.  What is bias, anyways?  It is favoring one view over another; because the term is typically used to suggest we are prejudging something, it has pejorative overtones.

So how are we biased? If you are reading a scientific paper about a topic you work on, you have probably read a lot about that topic.  And, odds are, you have digested this and come to a working conclusion, probably event A happened because of process B. You pick up a new paper that says A didn’t happen.  Are you really going to read it without having in mind all that work you read about event A? Of course not; for the paper to convince you, it must have evidence that is greater than the sum of evidence in the other material you’ve read.  In short, it must overcome your bias, where that favoritism was earned in sense from the weight of the prior work you’ve read or done.

Now maybe you have another word for that, but bias is what we’ll use here. It is not that you are rejecting the work out of hand, merely that you are more or less skeptical of the results based on the previous material you’ve digested.

That last phrase “more or less skeptical” should have you worried, because what you really want to do is treat each piece of work with an equal amount of skepticism. And this is where what you do with your bias matters.

If you recognize your bias, you need to be a bit more charitable with a paper whose conclusions are at odds with your own: don’t put words in the mouths of the authors that aren’t there, for instance.  Don’t expect that they will refute all your arguments. See what the paper has to offer and then add it to the scales you have about how the earth works rather than tossing it to the side. On the flip side, papers that sound like what you believe you need to treat more harshly. Does this look right because it really is, or are you overlooking flaws in the argument? Is this really new, or is it rehashing your favorite arguments? In sum, tear it apart before laying it on top of your existing reasons to support some position. This is not easy to do.

However, if you don’t recognize your bias, you are fated to plunge down the same path forever.  If you are lucky or exceptionally perceptive, this may not matter as you might have divined the correct path.  Congratulations. But if you are human like the rest of us, you will reach the wrong conclusion somewhere; if you approach papers with your bias hidden from you, it will be easy to dismiss contrary papers and equally easy to elevate sympathetic papers. Don’t think this happens?  GG saw both behaviors in a review of a paper: one reviewer saw phantom arguments made by others in the paper in question and argued vehemently against arguments that were objectively not present in the paper; because the conclusions were ones he didn’t agree with, he in essence presumed the worst and made the new paper accept all the sins of the old arguments.  On the other hand, another review was glowingly favorable, but there was no real attempt to see if the paper’s arguments were sound (in fact, a major piece of logic was missing that became evident from the negative review). Yet these two should have been ideal reviewers: they were both extremely familiar with the observations at the root of the paper. But they both effectively acted on their biases seemingly without recognizing the presence of those biases.

Arguably the hardest thing to do as you get more involved with science over the years is to read papers contrary to your biases at all. Why? There is only so much time in the day; if a paper is unlikely to overturn the weight of your experiences, you may feel it unlikely to be worth the time, while a paper paralleling your thoughts might reveal new insights to further build upon earlier work in the favored direction. Many times this behavior works out well, but there are those times when you are wrong and you are then missing the chance to be convinced you are wrong.  This is not an easy balancing act. Younger scientists are more omnivorous with literature, not having acquired the responsibilities demanding more of their time but also not having spent so much time working over older literature that has now been sufficiently accepted as to require little cognitive effort (e.g., for instance, that the viscosity of the asthenosphere is much less than the lithosphere). So in many ways it is us old coots who are at greatest risk of making this mistake.

So accept that you are biased.  As long as you know why you are biased and you understand the need to correct for it, you can do OK. The first step is to read the work you think is wrong.  The second step is to allow that paper to be itself and not the answer to every criticism. The third is to mentally keep score so you are not overweighting your favored theory, but also so you don’t overweight the most recent papers. (One reason bias can be insidious is that you may well use it internally to allow you to discard all the details that went into making that bias; thus when confronted with new information that appears to overcome your bias, you might have forgotten all the reasons that bias developed in the first place–and change your mind for the wrong reasons! Joseph LeConte’s late abandonment of glacial action in Yosemite is an example).

This is a lot harder than it sounds. GG is sure he has failed at times. The point is not to be critical of work you are disposed against, but to be just as critical or more critical of the work that seems to support your world view. So, good luck, and try to read a paper you disagree with this month. You never know, it might change your mind.

Advertisements

Tags:

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

%d bloggers like this: