Thursday, October 06, 2005

Journalism and Science

Lindsay of Majikthise has an interesting discussion on journalism and objectivity. In particular, she argues that journalism has taken objectivity too much towards disinterestedness.

I want to take that analysis in a slightly different direction, but make a similar point, using the analogy she draws between science and journalism.

A point that I like to make is that science isn't an encyclopedia, it's a process, a scientific method, if you will. First, you make some observations, then you develop a hypothesis, generate predictions from that hypothesis, and you test the hypothesis by conducting an experiment. Finally, you generate a new hypothesis based on your new results, and go back to the first step.

At the end of the experiment, something has changed. There are no longer two sides to the original story, there's just one, but the new story has two (or more) sides. If a scientist uses valid processes, it doesn't matter that she has a personal opinion about the hypothesis; the data tells her if the hypothesis is valid or not. How you choose your experiments expresses personal bias, but, at least in principle, there's relatively little room for bias to creep into the experiment itself or the analysis.

Journalists ought to approach their work in the same way. How you choose what story you write about is a reflection of bias. You make some observations and develop a hypothesis about what happened or is happening. Then you ask yourself what must be true if your hypothesis is wrong and what must be true if the hypothesis is false. Then you talk to the people who can tell you about those things. As Lindsay says, it doesn't matter so much when you're reporting things like the shipping news, or crop yields. But on politics and policy, it does. What stories a reporter chooses to follow will reflect personal biases. I care about the schools, so I cover the nonsense in the school board. I care less about the minutiae of tax law, so I didn't cover the TABOR tour through Kansas.

Once a journalist decides to cover something, the choices are more constrained. A piece of journalism is a hypothesis test. When covering the trial in Dover, the question shouldn't be, "What was said in court today?" but "Did what was said today support or falsify a hypothesis about the teaching of evolution or creationism?"

In Dover, the journalists need to pick their hypothesis. "Clueless rural rubes advocate creationism" is a possibility, but a bad one. "Intelligent design is creationism in a rented lab coat" is better. "How can Intelligent design be taught?" is a fair question to raise, leading to a series of hypotheses: IDC can/cannot be taught in science class, IDC can/cannot be taught in a history class, IDC can/cannot be taught in a philosophy class, etc. Each is a very different story, leading to different questions. For some, legal precedent is most relevant, for others, the current curriculum in the Dover schools is relevant. None requires a "he said, she said" format, because the reporter writes a story after learning something useful to the reader.

Journalism schools foolishly teach their students to "balance" quotations. Balancing sources is wise, but you shouldn't be obliged to quote every source. The only part of j-schools which don't teach this "balance" are the science journalism classes, which advocate talking to the researchers in the field and even letting sources vet an article for accuracy before it goes to press. One presumes that they do this because scientists and their research have already gone through a formalized vetting process.