How We Know What Isn't So (12 page)

i. Hedonic asymmetries.
One entrance to the Psychology Department at Cornell University is a set of six doors that are open during all but the wee hours of the night. For an unknown reason, however, the person who unlocks the doors each morning often fails to unlock one of the six, and the one that remains locked varies randomly from day to day. Because I approach the building from different directions on different occasions, the door through which I enter also varies haphazardly from time to time. It stands to reason, then, that I should happen to pick the locked door only occasionally, say, once in six entries if one door really were left locked each morning. Objectively, I acknowledge that that must be the case. Subjectively, however, it seems that the custodian has an uncanny ability to overlook the very door I happen to select later that day. I “always” seem to select the locked door!

As this example suggests, one of the most powerful determinants of whether various outcomes are one or two-sided is whether the potential outcomes differ in their hedonic or instrumental consequences. Outcomes are two-sided if both produce the same intensity (if not the same kind) of emotion, or if both necessitate further action on the part of the individual. Many times, however, only one of the outcomes arouses much affect or demands further action, making the outcomes one-sided. My entrance to the Psychology Department is a good example: A locked door that one wants to pass through arouses anger, and the frustration and delay make the event stand out in one’s experience. Passing through an unlocked door, on the other hand, requires no effort, gives rise to no emotion, and goes unnoticed. Consequently, the encounters with a locked door dominate my memory. Similar processes are no doubt responsible for a host of folk beliefs such as “it always rains right after you wash the car,” “you usually seem to need something just after you’ve thrown it away,” “the elevator (or the bus) always seems to be heading in the wrong direction,” and, as mentioned previously, “the phone tends to ring when you’re in the shower.”

The belief that “the bus always seems to be heading in the wrong direction” is particularly interesting in this regard because of an important asymmetry between positive and negative events: Certain kinds of negative events can accumulate in ways that positive events cannot. I can become convinced that all the buses are headed in the wrong direction by observing quite a number headed the wrong way before I encounter one going in my direction. Note that the opposite cannot happen: Unless I have difficulty boarding, I never observe several going my way before I discover one headed in the opposite direction. If a bus is going in my direction, I take it. Because of this asymmetry, we can experience a certain kind of “bad streak,” but not a complementary streak of good fortune. To those who fail to recognize this fact, events can sometimes seem to be conspiring against them.

Asymmetries in hedonic and instrumental consequences can also lead to the formation and maintenance of beliefs that can have more serious consequences, such as those that induce marital conflict. Many people claim that their spouses “never” do the chores or tasks they have agreed to do. While the claim is surely justified in some cases, in others it may stem from the fact that the spouse’s failure to wash the dishes, clean the counter, or do the laundry arouses resentment and anger, and can have immediate instrumental consequences such as the need to do the tasks oneself. When the tasks are performed on time, however, the world runs smoothly and there is little to notice. Similar processes operate in the common belief among couples that they are “out of sync”—it seems one always wants to stay home and watch TV when the other needs to socialize, one wants to make love when the other “needs some space,” one is upbeat when the other is morose, etc. Here too, there may be an inherent asymmetry in the salience of relevant events that can make things seem more out of sync than they really are. Wanting to do something when the other does not is frustrating, and it can occupy the contents of one’s mental life for some time. Examples of asynchrony are therefore easily brought to mind. But again, when a couple’s passions, preferences, or moods coincide, things go smoothly and the events can be less noteworthy. Furthermore, even when they do stand out, they tend to do so by virtue of the quality of the events themselves, and not by virtue of the synchrony that produced them. They are categorized and remembered as instances of laughter, passion, or fun, and not as instances of synchrony.

ii. Pattern asymmetries.
A second variable that makes some events one-sided is whether there is an asymmetry in the numerical, spatial, or temporal pattern produced by the various outcomes. Many people report that when they wake up in the middle of the night, their digital clocks indicate that it is something like 2:22, 3:33, or 1:23 “too often.” This is no doubt because such outcomes stand out—in a way that 3:51 or 2:47 does not—as a result of the pattern or “unit” that is formed. Indeed, a great deal of numerology depends upon certain coincidences being imbued with special meaning because events such as these are so salient and memorable that they seem more common than they really are. Similarly, the widespread beliefs (discussed in Chapter 2) that basketball players shoot in streaks or that gamblers get “on a roll” stem in part from analogous processes: A run of several hits or misses in a row, or a burst of consecutive winning bets, are so much more noteworthy and memorable than a mixture of hits and misses, or losses and wins.

To continue with sports for a moment, how many times have we heard baseball announcers state that “the player who makes a great play in the field to end an inning tends to be the one who comes to bat first the next inning.” Of course this cannot really be the case (unless we are willing to believe that an effect can precede its cause), but every time there is such an occurrence we are sure to notice it. Occasions when someone else leads off the next inning escape our attention.

There’s a similar origin to some of the remarkable similarities alluded to earlier in the character or life histories of identical twins who are reunited after having been separated at birth. A match on some characteristic or dimension creates a pattern, a unit, a focus of attention; a mismatch, unless terribly egregious, generally does not. The similarities between twins are noticed and remembered, and the dissimilarities pass us by.

Asymmetries in pattern are not limited to those that are spatial, temporal, or numerical, or to the inherent difference between matches and mismatches. Some derive from their relation to broader theories that we hold. Popular superstition informs us that the period of the full moon is an unusually dangerous time of the month. Consequently, any homicide, suicide, or accident during that time will command our attention and be linked to the full moon—
even if we do not believe in the superstition
. Similar events during other periods of the month will be thought of exclusively as what they are, and not as tragedies that happened in the
absence
of a full moon. This asymmetry leaves us with a distorted view of the relevant evidence and appears to lend empirical support to groundless superstition. Likewise, we notice when a crime is committed by someone on drugs or by a member of an ethnic minority. A link forms in our heads. But when the criminal is drug-free and a member of the majority, we focus on the crime itself and not on the absence of drugs or the perpetrator’s mainstream ethnicity.

iii. “Definitional” asymmetries.
Certain events are one-sided almost by definition. The outcomes relevant to the belief that “I can always tell when someone has had a facelift” is one example. Those that one detects lend support to the belief, but those that go undetected are simply that—undetected. They do not disconfirm the belief except in those rare instances in which an unsuspected person reveals that he or she has secretly undergone such surgery. The belief held by many people, including many clinicians, that a person can only overcome some problem (drinking, drug abuse, procrastination, etc.) after hitting “rock bottom” follows a similar pattern. Because there is no real definition of what constitutes “rock bottom,” it is hard to know what a disconfirmation of this belief would look like. Evidence that is inconsistent with the belief cannot stand out and is not remembered. The belief that people cannot profit from advice unless they are “truly ready” to receive it follows a similar logic. If the person benefits from the advice, he was obviously ready; if the advice is unheeded, however, the person must not have been “in the right place” to receive it. The very nature of the belief makes it impossible for it to be disconfirmed.

iv. Base-rate departures.
Perhaps the most common determinant of whether an event is one-sided is the base-rate frequency of the different possible outcomes. When certain outcomes occur frequently enough, they become part of our experiential background and go unnoticed. Departures from normality, in contrast, can generate surprise and draw attention. The unexpected can sometimes be unusually memorable.

Consider cases of cancer remission. Sadly, people who are diagnosed as having certain forms of cancer rarely recover. An instance in which someone does recover, therefore, is rather noteworthy, particularly if the person did anything unconventional to try to effect a cure (such as visiting a faith healer or travelling to Mexico for Laetrile therapy). Because we do not expect people to get better, we hardly notice any time someone tries an uncoventional treatment and it fails; when such treatments are successful, in contrast, the outcome violates our expectations and stands out in our memory.

Similarly, people’s beliefs in certain “jinxes” are partly due to the vividness of outcomes that depart from the base rate. People will often say things such as, “I hope I don’t jinx him, but Fred has never picked a losing stock.” It is easy to see how a concern about jinxing someone might arise. If a person has experienced such a large number of positive outcomes that it is worthy of comment, an additional success is not, by itself, terribly noteworthy. A subsequent failure, on the other hand, violates the typical pattern of success and thus stands out in the person’s experience. Examples of earlier jinxes are therefore easy to recall.

One of the most interesting classes of events that depart from the baserate and thus stand out in everyday experience is what sociologist Erving Goffman referred to as “negatively eventful actions,” or those actions and customs that are so common and automatic that we only become aware of them when someone fails to honor them.
²⁴
All of us have a preferred distance that we like to maintain from others—a “personal space” that governs the physical closeness of our interactions. Few of us, however, are aware of the precise dimensions or even the existence of such a bubble until someone invades it. It is only when someone violates the spacing norm that we even notice that it exists. Similarly, we tend to face forward in an elevator, pass fellow pedestrians on the right, and talk to people of different status with different styles of speech. All of this occurs with minimal awareness until we encounter someone who fails to uphold the norms.

Goffman’s negatively eventful actions are perfect examples of one-sided events: The outcome is perceived as an event only when it comes out one way. Although the “expectations” in Goffman’s examples are generally vague and unarticulated, it is the
dis
confirmations that tend to stand out. Interestingly, these kinds of
dis
confirmations tend not to undermine a person’s pre-existing beliefs. A norm violation of the type Goffman describes certainly does not diminish one’s expectations about how people will behave in the future. If anything, it strengthens those expectations by making them more explicit.

These asymmetries of pattern, hedonic consequences, etc., as well as one-sided events more generally, all serve to distort the evidential record that a person consults to evaluate the validity of various beliefs. For the most part, these asymmetries tend to accentuate information that is consistent with a person’s expectations and pre-existing beliefs. As a result, people tend to see in a body of evidence what they expect to see. What people expect to see, furthermore, is often what they want to see, and so the biasing effect of their preconceptions is often exacerbated by the biasing effect of their preferences and motives. This latter effect serves as the subject of the next chapter.

*
A blind observer is a person who is unaware of either the hypothesis under investigation or the specific condition of the experiment that is being run at any given time (e.g., treatment or control group). Because the observer is blind in this way, his or her expectations about what “should” happen in the experiment cannot bias his or her behavior.

*
Fittingly, P. T. Barnum also said that a good circus, like these bogus personality descriptions, should have “something for everybody.”

*
This is not to suggest that these two beliefs are false. Indeed, both have been subsequently supported by more rigorous evidence: Identical twins do tend to have somewhat similar personalities (see R. Plomin. Special section on developmental behavior genetics.
Child Development
, 1983,
54
, 331-55), and stress can in fact lead to cancer (see L. S. Sklar, & H. Anisman, Stress and Cancer.
Psychological Bulletin
, 1981, 89, 369-406). Note, however, that our confidence in the validity of these beliefs must rest on this more carefully collected evidence, and not on common anecdotal accounts that are subject to the problem of multiple endpoints.