Gregory A. Clark: Critical thinking is the basis of science

Al Hartmann | The Salt Lake Tribune Eleventh graders make glass slides and count Paramecium through microscopes, part of a national science competition in biology class Wednesday October 26 at Salt Lake Center for Science Education. New test scores from the National Assessment of Educational Progress show Utah students leading the nation in science.

In one sense, we’re all natural-born scientists. Give a toddler an egg, and sooner or later, they’ll drop it, just to see what happens — as life and Neil deGrasse Tyson remind us.

So why do we need to teach scientific experimentation, as the new proposed Science with Engineering Education (SEEd) Standards would have us do? And what’s up with the claim that the scientific method is counterintuitive?

Good questions. Let’s play, “I have a hypothesis.”

I (nature, metaphorically speaking) give you an observation: a triplet sequence of numbers, “2-4-6.” You (the scientist) try to determine the general rule that underlies such triplet sequences. You can’t know for sure what that rule is, given the present data. But you might venture a reasonable guess — a “hypothesis.”

First, write down your hypothesis, so you don’t change it later. To test your hypothesis experimentally, you’ll generate a new triplet sequence of numbers (methods). I’ll tell you whether your sequence fits the actual general rule (results). Repeat. You get as many tries as you like.

Write down the first sequence you want tested, for future reference. OK, now it’s answer time. The real point of this exercise — a seminal experiment by English psychologist Peter Wason in 1960 — isn’t to see whether you correctly hypothesized the general rule. Instead, the point is to see how you test your hypothesis.

If you’re like most people, you test your hypothesis the wrong way: by generating sequences that fit your hypothesis, rather than ones that don’t. Check yours; did you?

If your hypothesis was, say, “a sequence of even numbers, increasing by two,” then you probably generated a test sequence like that, such as “4-6-8.” To which my response would be, “Yes. Your sequence fits the general rule.”

So, you try another test sequence that also fits your hypothesis, perhaps “6-8-10.” Again I reply, “Yes.” And on it goes. Forever.

But here’s the thing. You’ll never discover the real general rule, which is this: “any ascending sequence.” Every sequence that fits your hypothesized but incorrect rule (“increase by two”) also fits the actual rule (“increase by any amount”). So you’ll never find out you’re wrong.


In contrast, if you generate a single test sequence that violates your rule -- say, “1-2-3” -- but I nonetheless say, “Yes,” then you know immediately that your original hypothesis is wrong: It’s not “increase by two.”

Easy-peasy. But infrequently done.

Takeaway Number 1: Our natural tendency is to look for evidence that we’re right and avoid evidence that we’re wrong. This tendency is so common it has a name: “confirmation bias.” The problem is, this approach often fails to find the correct answer, despite convincing us that it did.

Takeaway Number 2: We need to train students in a better way to think. The scientific method embodies that improvement. In science, we try to overcome confirmation bias. We seek potential disconfirmation, to see if our hypothesis is wrong, rather than seeking only confirmation that it’s right. We investigate alternatives. Ideas that survive serious challenge are more likely to be correct.

So, too, the proposed SEEd standards will try to teach students “to think critically and experiment” (The Salt Lake Tribune, Mar. 28).

Unfortunately, many Utahns find science threatening — perhaps because they find critical thinking and objective evaluation threatening. Even in science classes, they seek the comfort of religion, as the continuing uproar over the proposed standards shows. Religion doesn’t belong in public school science classes for a simple reason: It’s not science.

Honest science doesn’t evaluate its findings according to whether they conform to scripture, popular consensus or the religious beliefs of legislators. One can’t legislate the similarity between human and chimpanzee DNA any more than one can legislate the value of pi, as Indiana legislators once tried to do.

Science education must do what science itself does: Present scientific findings and explanations accurately and fully, regardless of whether they conflict with religious beliefs. Religion often vilifies knowledge. In Christianity, consuming forbidden fruit from the tree of knowledge constitutes the original sin.

People remain free to believe that at home. But that’s a horrible, inherently incompatible approach to science education. Honest science education, unfettered by religion, deserves our support. Voice it before 11:45 p.m., April 11, here.

Gregory Clark

Gregory A. Clark is an associate professor of biomedical engineering at the University of Utah. The views expressed are his own, and don’t necessarily reflect the U.’s.