Thursday, June 04, 2009

Preventing Bias

I had to report for jury duty today.

Any long-term reader of this blog will recognize that I put a lot of moral stock in the institution of trial by jury.

We are all plagued by confirmation bias, by tribal prejudices, by an inclination to see as 'proved' that which we want to believe is true. These are not just afflictions that affect ‘the other guy’ and for which each of us is blissfully immune. They are standard features of human psychology – dispositions to error (and even to unjustified violence) that we must protect ourselves (and each other) against.

Which is why the institution of trial buy jury makes sense. We present the evidence (ideally) to a jury that lacks bias, that has no pre-conceived notions relevant to the case to be confirmed, and has no personal stake in the outcome. By presenting the evidence to them, we can hopefully avoid some of the dispositions to error.

Scientists are fighting against the same dispositions when they invented the procedures that define the scientific method. In recognition of these human flaws, they insist that the data be taken in a way that another researcher can replicate the results, or that the person recording the data does not know what thesis he is trying to confirm or deny (and has no personal stake in the thesis being tested).

Requiring these methods is not an insult to scientists – or to people generally. It is recognition of a common human trait and the need to take these psychological facts into account when testing our beliefs (or hypothesis’).

This is part of what this blog is about. I do not pretend to have all of the right answers to moral questions. In fact, I have repeatedly asserted that I know that I am wrong about at least one thing. Part of the reason for this blog is to present a set of ideas for ‘peer review’ – that review coming from the members of the studio audience.

I encourage readers to read the comments to each post, because they will help to identify when my own dispositions to confirmation bias and tribal prejudices are getting in the way of clear reason.

This was one of the problems with the Gitmo hearings. The Bush Administration had a vested interest in finding weapons of mass destruction, or in finding something that would link Al Quida to Saddam Hussein. It was plagued by confirmation bias – thinking it even had the authority to rewrite scientific publications so that they confirmed the administration's views. It should have adopted some method of presenting the evidence to third-parties, but its members were too arrogant to consider even the possibility of error.

This appears, so far, to be one of the merits of the Obama administration. Obama seems more modest, more inclined to seek advice from others – not just for political show, but for the sake of actually listening to what they have to say. He seems to better recognize the psychological dispositions that make us prone to error.

Of course, this may just be my own confirmation bias and prejudices coming through.

Another way in which we can deal with confirmation bias and tribal prejudices is to simply adopt the jury technique of a presumption of innocence. If you read something in which somebody ways, "The people of the tribe that you and I belong to are superior to the people of that other tribe over there," I would recommend starting with a presumption of innocence.

See if you can find holes in the reasoning. You will have to take a close, hard look because confirmation bias and tribal prejudices will blind you and make those holes hard to see. , and ask yourself if you are not seeing them because they are not there, or because you have been blinded by your prejudices.

Think of yourself as being on a jury. If you do not qualify to be on a jury for the issue you are investigating, then perhaps you should be offering your ideas as suggestions for others to review but one that certainly cannot be tagged as free from bias.

I often think that a little more humility is in order.

1 comment:

Stephen L. Gibson said...

Well said, oh wise one :-). When a friend recently said "I never learned anything with my mouth open," I suggested the opposite. If we genuinely care to listen to and learn from the responses, asking questions or positing concepts is an immensely valuable way to learn. Thanks for all your efforts.