For over 25 years — from the late 1920s until 1955 — it had been repeatedly observed that each cell in the human body contained 48 chromosomes[∂] (excepting certain cell types, such as mature reproductive cells — sperm and ova — which were found to have 1/2 that number). As required by the scientific approach (see Section 1-2), this claim was tested repeatedly by many biologists working in different laboratories in different countries. Thus, the claim that humans have 48 chromosomes in each cell of the body was accepted as very likely to be true. In fact, this number was observed so often under so many different conditions and by so many different observers that biologists accepted it as an established fact[∂]. By the 1950s, any biologist who reached a different count had to conclude that he had made a mistake(Kevles, 1985), just as people counting something other than 10 fingers on a pair of normal hands would conclude that they must have made a mistake.

In 1955, however, Joe-Hin Tjio used an improved technique for preparing cell nuclei[∂] on slides — a technique that spread apart the chromosomes so well that each could be clearly identified for the first time (Tjio & Levan, 1956). Using these preparations, Tjio repeatedly counted only 46 chromosomes in each cell. He repeated the counts many times with chromosomes from many cells and continued to count only 46 chromosomes. He asked others in his laboratory to count them. They also reached the same count. In other words, Tjio found that the chromosome number that had been repeatedly observed and reported in every scientific article published on the topic for over a quarter of a century had been wrong.

Why did so many researchers during this time period repeatedly count the wrong number of chromosomes? Although each was being properly empirical in his or her research, their primitive chromosome-preparation techniques resulted in a clump of intertwined chromosomes that could not be distinguished easily: counting the number of chromosomes in this clump involved a great deal of subjectivity[∂]. And not only were the chromosomes clumped together, but researchers also had to slice through the nucleus[∂] of the cell, mount each slice on a different slide, and then count the chromosomes by examining the slides and adding up the chromosomes found on each one. The problem: a single chromosome might end up on two slides if it had been sliced in two. Whether a researcher counted a chromosome that had been cut in two as one or two chromosomes was a "judgment call." There was no way for the researcher to know for certain whether chromosome fragments on different slides actually had been one chromosome cut in two or two chromosomes, each appearing on a different slide.

A second important subjective factor was the preconception[∂] — the prior belief — that there were 48 chromosomes in human cells. The difficulty of counting clumped and sliced chromsomes coupled with this preconception caused researchers to repeat a count when they came up with the "wrong" number of chromosomes (that is, something other than 48). Then, they would count two fragments as either one or two chromosomes, depending on whether their initial count had been high or low. In these cases, they weren't "lying" about what they had observed: they were correcting what they thought had been an inaccurate count.

The struggle to find 48 chromosomes (even when a correct count of 46 had been made) can be seen in the case of Tao-Chiuh Hsu, a biologist who, a few years before Tjio, also used an improved technique that allowed him to clearly distinguish and identify each chromosome. Hsu stated that he "had difficulty getting the count to equal forty-eight" chromosomes (Kevles, 1985, p. 241). He kept counting 46 but, because of his preconception that there were 48 chromosomes, he kept repeating his counts until he was able to count the "correct" number. Hsu's experience is an excellent example of what has come to be called the confirmation bias — the tendency to accept without question the accuracy of evidence that agrees with (confirms) one's preconceptions, and to question the accuracy of evidence that contradicts (disconfirms) one's preconceptions. The disconfirming evidence — such as Hsu's count of 46 chromosomes — is examined very carefully until problems with the evidence are found., and then distorted and exaggerated until they appear significant. At this point, the disconfirming evidence can be rejected. Since the confirming evidence is not subjected to a similar scrutiny, it is readily accepted and used to support the preconception.

To take a more everyday example of the confirmation bias, let's say that a police detective is investigating a murder — a crime observed by a bystander. If the detective already believes that a particular person committed the crime, she probably will include that suspect's picture in a "photo line-up." If the eyewitness hesitates over the picture of the suspect but then chooses a different photo with some uncertainty, the detective may ask the eyewitness to look a second time more carefully. On the other hand, if the eyewitness hesitates over the suspect's picture, looks at the others, but finally chooses the suspect, again with some uncertainty, the detective may quickly state, "that's the guy we thought did it!" You can see in this fictional example that the detective questioned the eyewitness's uncertain identification of another person but quickly accepted his uncertain identification of the person that the detective already suspected. In this manner, the detective set up a situation in which it became more likely that her prior belief would be confirmed and that an alternative possibility would be rejected.

The psychologist, Thomas Gilovich (1991), described the confirmation bias in this way:

When examining evidence relevant to a given belief, people are inclined to see what they expect to see, and conclude what they expect to conclude. Information that is consistent with our pre-existing beliefs [preconceptions] is often accepted at face value, whereas evidence that contradicts them is critically scrutinized and discounted. Our beliefs may thus be less responsive than they should to the implications of new information. (p. 50)

Because of this bias, we are unlikely to examine potential problems with evidence that supports a belief we already hold, even when these problems are relatively obvious to anyone taking the time to look. When researchers counted 48 chromosomes in their preparations, they accepted without question that this observation was accurate. And because of this bias, we are likely to closely examine potential problems with evidence that contradicts a belief we already hold, even when these problems are minor (or perhaps even nonexistent). Thus, if significant problems exist, we will find them and reject the evidence; and even if significant problems don't exist, we still may find minor problems that cause us to question the evidence more than we should. When researchers counted something other than 48 chromosomes (such as the correct number, 46 chromosomes), they assumed that this observation was inaccurate. Thus, they recounted until they felt satisfied that the cell contained 48 chromosomes (that is, they distorted their observations); or, if they still were unable to count 48 chromosomes, they probably would have discarded the data because, in their minds, the slides must have been improperly prepared. This is what Hsu meant when he said that he "had difficulty getting the count to equal forty-eight" chromosomes. He had to work hard either to misperceive the evidence until it seemed to agree with his preconception or to find a reason to reject the evidence as inaccurate.

The fact that we don't subject confirming evidence to the same degree of examination and evaluation as we do disconfirming evidence means that we are unlikely to question our preconceptions, let alone reject them at some point. Instead, because of the confirmation bias, our preconceptions are likely to become stronger and stronger over time. This is why it is important to become aware of the influence of the confirmation bias in your everyday life. This is a very difficult thing to do because typically the confirmation bias works unconsciously (without awareness). We'll look at the unconscious influence of the confirmation bias (and other "cognitive biases"[∂]) elsewhere in this text.

Study Questions for Section 1-3

1. How would you define in your own words the "confirmation bias"?
2. What is an example of the confirmation bias from your own life?
3. How does the confirmation bias limit our ability to properly test whether a claim is true or not?
4. In what way does the history of research on the number of human chromosomes illustrate the negative influence of the confirmation bias?

Go to Quiz 1-3 questions

Go to Readings Section 1-4