email - July 2017

Scientific Bias

Is the scientific method valid if the scientist is biased?

Dear Do-While,

My question is about the scientific method. I know from your writing that you believe firmly in its importance. I can't deny that it serves to keep science honest, as long as scientists have the integrity to remain objective, but that's actually the problem with it. They don't.

It's the "develop a theory" part that I see as a flaw. In a field where brainpower is everything, who wants to admit that they are wrong? Nobody wants to submit research with a conclusion that says, "The evidence disproved my idea, so that's that, I guess."

However, I am unable to think of a better way. Simply observing without forming a theory would produce information without any significance attached. If you have the time, would you care to share your thoughts?

Also, if you know of any good, OBJECTIVE sources of science, for the pure sake of knowledge, not atheism and politics couched in scientific terms, let me know! I'm not smart enough to make much sense of many of the articles in peer-reviewed journals.


Both of Johnís questions have to do with bias in science. Does the scientistís expectation of a desired outcome negate the validity of the experiment? Are there any objective scientific sources? They are two good questions which deserve to be answered. First, letís examine if a desired outcome necessarily invalidates an experiment.

Experimental Expectations

Yes, we firmly believe in the scientific method. The experimental part of the method is what gives it objective integrity. The experiment succeeds or fails regardless of what the scientist believes. The scientific method is a reliable way to discover the truth.

We also recognize, as John does, that some scientists are dishonest. Dishonest scientists can ignore results they donít like. They might fool themselves into believing that unexpected data points are ďoutliers,Ē and need not be taken into consideration. They might even make up data to conform to what they think would have been the experimental outcome if their equipment had been more accurate. The blame rests on the dishonest scientist, not the method.

But John isnít really concerned with dishonest scientists. He wants to know if having a preconceived theory in mind invalidates the scientific method. That is, if the scientist is trying to prove a theory, can the experiment still be unbiased? Does a prejudicial expectation somehow invalidate the objectiveness of an experiment? No, it doesnít. The goal of an experiment is usually to prove (or disprove) an expected outcome, and thatís good.

Sometimes a scientist does conduct an experiment without any clue as to what the outcome will be. This happens often in genetic research. The scientist might damage a gene in a fruit fly to see what kind of birth defect it will cause. Thatís how the functions of many genes have been discovered, and it is certainly an unbiased methodóbut it has two drawbacks.

First, doing an experiment without having any expectations is inefficient. A chemist might randomly mix some chemicals together to see what happens, but it is unlikely that anything useful will result. If he is lucky, he might stumble on a new kind of glue, or a more powerful explosive. But if he is looking for a new kind of glue he would be better off mixing together chemicals that are known (or suspected) to have adhesive properties, rather than things that are likely to explode when mixed together.

Second, if you donít know what you are looking for, there is a good chance you wonít see it. If a chemist just mixes some random chemicals together, it might create a really good cleanser. But if he isnít looking for a good cleanser he probably wonít think to test his concoction on different kinds of stains.

Our point is that if one has no preconceived expectations of the experimental outcome when mixing chemicals, the probability of success is low. Furthermore, in the unlikely event that something useful is produced, the value of the product might not be recognized because the scientist wasnít looking for it.

On the other hand, there is the danger that if a scientist mixes some chemicals together in an attempt to make a better glue, he might produce a better cleanser, but he might not notice how well it removes stains because he is so intent on seeing how well it makes things stick together.

A properly designed experiment will unambiguously succeed or fail depending upon whether the theory is right or wrong. It could be that the experiment is flawed, and the outcome is misinterpreted. Thatís what critical peer review is for. Thatís why the methods have to be reported as well as the results. It allows other scientists to repeat the experiment to see if they get the same results, and to see if there are other effects that werenít noted by the originator of the experiment.

Yes, scientists are biased. They do experiments expecting their bias to be confirmed. There is nothing wrong with that. The experiment will reveal the truth, regardless of the scientistís bias. Of course, an unethical scientist might misreport the result of an experiment that doesnít turn out as desired; but that is a failure of the scientist, not a failure of the scientific method.

Objective Sources

In regards to Johnís second question, we suggest to John, and all our readers, that one should operate under the presumption that no source, including Science Against Evolution, is objective. Question everything and everyone. Be skeptical, and honestly evaluate every idea for yourself. Listen to both sides.

Johnís Response

We didnít want John to have to wait a month for his answer, so we mailed him a preliminary draft of this column. Here is part of his response:

I was thinking of the penchant of a large number of scientists to put the fate of their "pet theory" ahead of scientific integrity. I believe this is why I see so many articles riddled with the second causative, or words like "might", "could", "possibly", etc. I find it sad that the science available to the general public (who, like me, have a difficult time with the technical language in journals like Science or Nature) is left up to what individuals like Dawkins, Coyne or the editors of National Geographic choose to tell us.

The reason why so many articles are riddled with weasel words is because there is no experimental proofóit is all speculation, not science.

When people tell you things, you have to check them out for yourself. There is no shortcut.

Quick links to
Science Against Evolution
Home Page
Back issues of
(our newsletter)
Web Site
of the Month
Topical Index