Chapter 4
Lesson 7: The Falsifiability Criterion of Science
In the previous lesson, you learned that the concepts used in a claim must be linked to observable events if the claim is to be tested. For example, the claim that "all swans are white" is a testable claim. Operational definitions of "swan" and "white" could be developed. We then could travel around the world observing whether or not swans are white. Each new observation of a white swan would support the claim that "all swans are white." Thus, the claim that "all swans are white" is "verifiable." A claim is verifiable when it is possible to imagine a set of observations that would support the claim. However, there is an important problem with verifiability in this example: even if we observe a million white swans in a row, the next swan we observe still might not be white. Although we become more confident that the claim is true as we observe each new instance of a white swan, we can never be certain that it is true. In other words, the claim can never be "proved".
Scientific claims are similar to the one about swans: in science, we are attempting to discover universal truths--claims that are true for all instances of a particular phenomenon. For example, if we claim that a gene on Chromosome 7 is a cause of a particular mental disorder, we are saying that the gene will influence (in combination with other factors) the development of the disorder regardless of the time or place we are in. Nevertheless, no matter how many times we observe that people who inherit the gene (in combination with the other essential factors) develop the mental disorder, we can never say with absolute certainty that this gene always influences the development of the disorder. Although a large number of verifications of the association will tend to make us certain beyond a reasonable doubt that the gene is a cause of the mental disorder, we can never prove the claim.
Karl Popper famous philosopher of science, noted this problem with the verifiability of universal claims and stated that there is a solution. In 1934, he argued that, although we cannot prove a universal claim to be true, we can prove that it is false. For example, if we observe just one black swan, the claim that "all swans are white" has been disproved. When it is possible to imagine a set of observations that would show a claim to be false, we say that the claim is falsifiable.
Popper argued that, in order for a claim to be a "scientific claim," it must be falsifiable--a requirement that is referred to as the falsifiability criterion of science. According to this criterion, any claim that looks like a scientific claim, but that is not falsifiable, would be a pseudoscientific claim: the claim looks scientific but actually is not. One example of a pseudoscientific claim involves the notion that there exist "human energy fields" emanating from our bodies that result in illness when disturbed in various ways. Some proponents of this idea have stated that these energy fields cannot be measured objectively. These people claim that the only way to detect human energy fields in others is by "sensing" them with one's body, which is too subjective a method to be of any scientific use. There is no way to falsify the claim that human energy fields exist and, thus, it seems to be a pseudoscientific claim.
Requirements for Falsifiability
There are several requirements that must be satisfied in order for a claim to falsifiable.
(1) Specify relevant observations. One must specify in advance which observations will support the claim and which observations will not. This is necessary because, if we already believe a claim to be true, we tend to accept observations consistent with the claim and explain away observations that are inconsistent with it. For example, some believers in "psi" (which may defined as the factor underlying psychic phenomena such as extrasensory perception) may readily accept positive experimental results but attempt to explain away negative results by proposing new (and sometimes bizarre) phenomena.
As one instance of this, we can examine the so-called "shyness effect," which refers to the fact, long noted in parapsychological research, that psi seems to disappear in the presence of skeptics. Those who believe that psi exists have explained this result by proposing that skeptics produce what might be called "negative psi," which is claimed to cancel out the "positive psi" of experimental subjects. If this is true, then any attempt to test psi scientifically is made much more difficult, to say the least: anytime a researcher fails to find psi, it could be claimed that the researcher was too skeptical and, thus, was producing negative psi. In fact, some have even argued that the shyness effect could occur backwards in time and in places far away from the location of the study. These critics claim that a skeptic who merely reads about the results of a psi study may change the results of the study through his or her skepticism. It should be obvious that this concept of "retroactive psi" would make any claims about psi unfalsifiable: by explaining away any negative findings as being due to the effects of skeptics in the future who read about the study, there is no way to test the existence of psi.
The examples of negative psi and retroactive psi illustrates the use of an ad hoc hypothesis, which is defined as a speculation developed in order to explain away findings that falsify a claim. The research of Benjamin Rush, a famous physician of the late-eighteenth and early-nineteenth centuries, provides another example of ad hoc hypothesizing. During a yellow-fever epidemic in 1793, Rush used bloodletting (in which blood is drained from an open vein) to treat the ill. He did this because he thought that yellow fever involved inflammation in the brain, and bloodletting had been a standard treatment for inflammation for many centuries (Arnebeck, undated). Nevertheless, many criticized Rush's method of treatment. According to Keith Stanovich (2001):
Critics charged that his treatments were more dangerous than the disease. However, following the epidemic, Rush became even more confident of the effectiveness of his treatment, even though several of his patients had died. Why? ... [I]f the patient got better, this improvement was taken [by Rush] as proof that bloodletting worked. If the patient died, it merely meant [to Rush] that the patient had been too ill for any treatment to work. (p. 23)
Instead of considering the frequent death of patients to be a disconfirmation of his claim, Rush created an ad hoc hypothesis that explained away their deaths. Thus, he was able to hold on to his claim that bloodletting cured yellow fever (as long as the patient was not too "far gone").
(2) Develop operational definitions. Predictions made to test a claim must contain concepts that can be operationally defined. In Lesson 6, you learned why this is necessary.
(3) Make specific predictions. General predictions do not allow a definitive test of a claim: the chance of their being falsified is smaller than is the case with a highly specific prediction. For example, Benjamin Rush predicted that his patients would recover after bloodletting (as long as he treated them "in time"). Such a general prediction, however, is very difficult to falsify, especially given the fact that about half (or more) of yellow-fever victims eventually recover on their own without any treatment (Arnebeck, undated). Thus, Rush's observations of recovery were consistent with his claim that bloodletting is an effective treatment for yellow fever. But these observations are also consistent with other possible claims, such as that the body's own immune response causes recovery. A more specific prediction, on the other hand, would have allowed a better test of the claim because other possible claims would have been less likely to make the same prediction. For example, if Rush's claim about bloodletting had led him to predict that bloodletting would eliminate vomiting and bodily aches within three hours and fever within 24 hours, it would have been much easier to falsify the claim; and any confirming evidence would have been much better support for it. In general, then, it is easier to falsify a specific prediction than a general prediction; and any observations consistent with a specific prediction provide better support for the claim from which it was derived.
Question 7-1
Sigmund Freud proposed that we defend against the anxiety caused by conflicts between motives (such as sex) and other factors by using defense mechanisms. Freud stated that, in using a defense mechanism, we distort reality unconsciously. In other words, we pretend that something isn't true that is, or that something is true that isn't; and we do this without any awareness that we are doing so. For example, in using the defense mechanism of "reaction formation," we push an unacceptable feeling into the unconscious and then express the opposite feeling in our conscious behaviors. In this way, we distance ourselves from our true desire by expressing behavior that is diametrically opposed to it.
For instance, let's look at the case of a woman whose parents neglected her to an extreme degree when she was a child. Let's say that Karl believes the following claim to be true: people with desires left unsatisfied during childhood will become preoccupied with satisfying them during adulthood. Based on this claim, Karl predicts that the woman will be very needy and dependent as an adult because she is trying to satisfy dependency needs that were left unsatisfied by her neglectful parents.
When Karl interviews the woman, however, he is struck by how independent and even aloof she appears. She seems not to need anyone for anything. Karl explains this surprising behavior by supposing that the woman has repressed her dependency needs (pushed them into the unconscious, thereby becoming unaware of them) and, on the surface, now expresses an extreme independence from others in order to distance herself further from her unacceptable dependency needs. In other words, she is using the defense mechanism of reaction formation.
Discuss this example in terms of the falsifiability criterion of science.
Suggested Answer
Question 7-2
In Chapter 4 of your textbook, the frustration-aggression hypothesis was discussed. The frustration-aggression hypothesis makes the following claim: aggressive behavior is always caused by frustration and frustration always causes aggressive behavior. In other words, frustration is claimed to be a necessary and sufficient condition (see Lesson 1) for aggressive behavior.
Is the frustration-aggression hypothesis a scientific claim? Explain your answer. Given your experiences with frustration and aggression, does the frustration-aggression hypothesis constitute a good explanation of aggressive behavior?
Suggested Answer
Question 7-3
In Chapter 4, the learned-helplessness theory of depression was discussed. The learned-helplessness theory makes the following claim: unavoidable negative life events that are experienced repeatedly over time cause people to give up trying to deal with future negative life events. This "giving up" causes people to become depressed. In other words, it is claimed that people become passive and depressed after repeatedly experiencing uncontrollable and unpleasant events. These people no longer attempt to overcome any adversity in their lives, even when it can be avoided.
With respect to the falsifiability criterion of science, first discuss the claim that people become depressed after repeatedly experiencing unavoidable and unpleasant life events; and then discuss the research described in the textbook relevant to testing the claim.
Suggested Answer
Question 7-4
It is sometimes claimed that psychics have provided clues about crimes to police--clues that later were found to be amazingly accurate. For example, let's say that a woman has been missing for several weeks and a psychic claims that he "sees" her dead body "in a wooded area near a body of water." Later, the police find the woman's body in a clearing about one-half mile from a wooded area and about three miles from a large lake. The psychic concludes that his prediction was confirmed. Evaluate this conclusion with respect to the three requirements of falsifiability.
Suggested Answer
Question 7-5
Let's say that you want to test the claim that biological heredity is an important influence on personality. In order to test the accuracy of this claim, you make the following prediction: the behavior of relatives will be more similar than the behavior of nonrelatives. Discuss this prediction with respect to the three requirements of falsifiability.
Suggested Answer
Bibliography and References
Arnebeck, B. (n.d.). Destroying angel: Benjamin Rush, yellow fever and the birth of modern medicine. Retrieved April 14, 2002, from
Blackmore, S. (1996). In search of the light: The adventures of a parapsychologist.Amherst, NY: Prometheus Books.
Davison, G. C., & Neale, J. M. (2001). Abnormal psychology (8th ed.). New York: John Wiley & Sons.
Ford, G. (2000). Phlebotomy: The ancient art of bloodletting. Retrieved April 14, 2002, from
Hergenhahn, B. R., & Olson, M. H. (1999). An introduction to theories of personality (5th ed.). Upper Saddle River, NJ: Prentice Hall.
Magee, B. (1985). Philosophy and the real world: An introduction to Karl Popper.LaSalle, IL: Open Court.
Nickell, J. (1994). Psychic sleuths: ESP and sensational cases.Buffalo, NY: Prometheus Books.
Popper, K. R. (1959). The logic of scientific discovery.London: Routledge. (English translation)
Posner, G. (1997). A not-so-psychic detective? A case study of Noreen Renier's latest and greatest 'success' story. Skeptic, 5(4), 52-57. Retrieved April 4, 2002, from
Stanovich, K. E. (2001) How to think straight about psychology (6th ed.). New York: Longman.
RETURN TO TOP