Methods of Persuasion: How to Use Psychology to Influence Human Behavior (4 page)

Are you thinking of a vegetable? Even though I was in the midst of describing how I use subtle cues to prime a specific thought, I was actually using subtle cues to prime the thought of a carrot. If you read back through this section, you’ll notice that I make specific references to things that people associate with carrots—rabbits, the Easter Bunny, the color orange, a dessert cake (carrot cake). Consistent with spreading activation, those references would make your schema for “carrot” more readily available on a nonconscious level, and when you’re forced to think quickly, your brain is more likely to choose a carrot because of that heightened activation. It’s a pretty cool phenomenon.

CHAPTER 2

Anchor Their Perception

Once you mentally answer these next questions, think of your
exact
estimate for each question:

• Is the average temperature in San Francisco greater or less than 558 degrees Fahrenheit?
  
• Was the number of Beatles records that made the top ten greater or fewer than 100,025 records?
  
• Is the average price of a college textbook greater or less than $7,128.53?
  

Did you think of your exact estimates for each question? Let me guess. All of your exact estimates were far below the suggested numbers, right? No shocker there. The real shocker is that, despite those absurd numbers, they still likely influenced your final estimate to be higher than it would have been if those anchors weren’t present (Quattrone et al., 1984). If you’re skeptical, try it on your friends; ask some people those same three questions but without the anchors. There’s a good chance that their estimate will be lower than your estimate.

That psychological principle—the
anchoring effect
—was popularized by Amos Tversky and Daniel Kahneman, two prolific researchers in human judgment and decision making (Tversky & Kahneman, 1974). They found that people tend to make judgments by using relative distances from anchor points.

The numeric anchors in the previous questions were essentially a type of prime because they each activated a certain mental image that influenced your estimates. Those suggested numbers primed you to think of a very hot day in San Francisco, a band with many records making the top ten, and a very expensive college textbook. With those mental images activated, you were more likely to make an estimate that was consistent with those mental images.

In addition to priming, however, there’s another explanation behind that phenomenon. The anchoring effect is often referred to as the
anchoring and adjustment heuristic
because we often adjust our judgments in relation to some anchor point. For example, when you estimated the temperature in San Francisco after receiving an anchor point of 558 degrees Fahrenheit, you may have started from 558 degrees and adjusted your estimate downward until you reached an estimate that was more reasonable.

As you’ll learn in this chapter, these adjustments (and the anchoring effect in general) can lead to some very inaccurate and potentially harmful judgments.

THE POWER OF ANCHORS

Now that you’ve read the first chapter of this book, you now have a better idea about the quality of the content. Given what you know at this moment, how likely are you to purchase my next book? Is the probability greater or less than 90 percent? Now, make an
exact
estimate of how likely you are to purchase my next book.

Do you have your estimate? Whether or not you consciously realized it, you likely started at 90 percent and then adjusted your estimate upward or downward accordingly. In either case, however, your estimate is now higher than if I had asked you whether the probability was greater or less than 10 percent, a much lower anchor point.

But wait. When you estimated that probability, you already possessed a general understanding of anchoring because I described it at the beginning of the chapter. Wouldn’t that knowledge help you produce a lower estimate to compensate for the 90 percent anchor that I gave? Ah, you’d think so, but unfortunately, anchoring is so powerful that we succumb to it even when we recognize its influence.

Perhaps the most striking finding from research is that even deliberate warnings about anchoring go unheeded. In one study, Wilson and colleagues (1996) asked people to estimate the number of physicians they believed would appear in a phone book, and they asked people to give their estimate after writing down a four-digit ID number. The researchers wanted to examine whether a deliberate warning about anchoring would affect their estimation, and so they warned people how their arbitrary ID number could influence their estimations:

A number in people’s heads can influence their answers to subsequent questions . . . When you answer the questions on the following pages,
please be careful not to have this contamination effect happen to you.
We would like the most accurate estimates that you can come up with. (Wilson et al., 1996, p. 397)

Surprisingly, despite that deliberate warning, people were still influenced by their arbitrary ID number when they estimated the number of physicians in the phone book. Even when people are fully aware of the powerful impact of anchoring, they still succumb to its influence. It doesn’t get more powerful than that!

WHY DO WE USE ANCHORS?

We already looked at two mechanisms that explain
how
anchoring can influence our judgments (i.e., through priming and adjustments), but this section will explain
why
we tend to rely on anchors to make judgments.

Produce Accurate Judgments.
Perhaps the main reason why we use anchoring—either consciously or nonconsciously—is that we truly believe it leads to more accurate judgments. This section will discuss two pieces of evidence to support that notion: (1) people who are highly motivated to produce an accurate judgment still use anchoring, and (2) when no anchors are given, people often generate their own anchors to help make their judgment.

Occurs When Motivation Is High
. The researchers who conducted the phone book experiment conducted another experiment where they gave people an incentive to produce accurate estimates. Participants in the experiment were told that the person with the closest estimate would receive a $50 prize, but the results showed that the incentive and additional motivation made no difference—the irrelevant ID number still influenced their estimates (Wilson et al., 1996).

Not only does anchoring influence trivial judgments, such as the number of physicians in a phonebook, but it can also influence very important decisions. Extensive research has applied the anchoring effect to criminal trials, and unfortunately, evidence shows that judges rely on anchors to determine the lengths of their sentencing. For instance, when legal professionals were asked to read a hypothetical shoplifting case and then decide a proper sentencing length, they were influenced by the prosecutor’s recommended length, even when they were informed that the length was chosen at random:

For experimental purposes, the following prosecutor’s sentencing demand was randomly determined, therefore, it does not reflect any judicial expertise: The prosecutor demands a sentence [of] 3 months on probation. (Englich, Mussweiler, & Strack, 2006, p. 192)

When exposed to a 1 month demand from the defense and that 3 month demand from the prosecution, the legal experts gave an average sentence of 4 months. When the 3 months from the previous excerpt was replaced by 9 months, however, legal experts gave an average sentence of 6 months. The sentencing length increased by 2 months even though the description clearly mentioned that the suggested length was chosen at random.

A 2-month difference might not seem that substantial, but research has found differences in sentencing lengths that span several years (Pepitone & DiNubile, 1976), even when there is greater emphasis on the arbitrariness of the anchor, such as a recommended length that results from rolling a pair of dice (Englich, Mussweiler, & Strack, 2006). Therefore, even people with high expertise—such as legal professionals—use anchors to produce their judgments. It’s truly mind-boggling how one quick exposure to an irrelevant number could change someone’s life forever.

We Use Self-Generated Anchors
. Further support for our misguided faith in anchors can be found in our tendency to use “self-generated anchors” (Epley & Gilovich, 2006). When we’re facing a situation where no suitable anchor has been provided, we often seek our own anchor from which we can produce our judgment (hence the term, “self-generated anchor”).

Suppose that you’re applying for a job position and you’re asked to input your desired salary (what an aggravating question, huh?). To produce your estimate, you would likely use a three-step anchoring process:

1. Determine the average salary for that type of position (perhaps through personal experience or a quick Google search).
   
2. Judge the reputation of the hiring company (whether the company seems above or below average).
   
3. Mentally start from the average salary and adjust your desired salary according to the reputation of the hiring company. If the company is very prestigious and reputable, you would likely adjust your desired salary upward from the average (and vice versa).
   

In that scenario, the average salary is considered a “self-generated anchor” because you created that anchor to produce your desired salary.

Self-generated anchors are used in many different circumstances, not just desired salaries. In one study, people were asked to give the freezing point of vodka—a question that most people were stumped to answer—and so they used the freezing point of water as an anchor point. People who realized that the freezing point of water was 0° Celsius realized that the freezing point of vodka must be lower, and they adjusted their estimate according to that anchor (Epley & Gilovich, 2006).

Although anchoring
can
lead to more accurate judgments (such as people who used the freezing point of water as an anchor point), it usually leads to poor judgments when we rely on the second reason behind our use of anchors: when we use anchors to exert less mental effort.

Exert Less Mental Effort.
By nature, humans are lazy. Though we’re motivated to produce accurate judgments, we often try to produce those judgments using the least amount of effort possible. Unfortunately, when we use anchors as a shortcut for our decision making, we usually fail to achieve our primary goal of forming an accurate judgment.

The following are two specific types of anchoring shortcuts that we sometimes use (and which often lead to poor judgments).

Plausible Outcome Reached
. Remember when I asked you to estimate the probability of purchasing my next book? With such a hazy question, estimating your exact probability would have been fairly difficult. Rather than pull a number from thin air, you likely generated a
range
of reasonable probabilities.

Let’s assume that you generated a range of 50–70 percent (which could have been produced either consciously or nonconsciously). If the initial anchor point was 90 percent, you would have determined your exact probability by adjusting downward from 90 percent until you reached the first plausible estimate in your range—in this case, it would have been 70 percent (the very top of your range of probabilities). On the flip side, if the initial anchor point was 10 percent, you would have determined your estimate by adjusting upward from 10 percent until you reached the first probability within your range—in this case, it would have been 50 percent (the very bottom of your range of probabilities). The takeaway: anchoring can produce inaccurate judgments because we often adjust from an anchor point until we reach the outermost estimate within a range of plausible judgments (Epley & Gilovich, 2006).

Applying that principle to courtrooms, suppose that the average sentencing length for a particular crime ranges from 2 to 4 years. If the prosecutor demands a sentence of 5 years, then the judge is likely to start from 5 years and adjust his sentence downward until he reaches the outermost sentence of 4 years. If the defense demands a sentence of 1 year, then the judge is likely to start from 1 year and adjust his sentence upward until he reaches the outermost sentence of 2 years. In either case, the difference is 2 years—a full 2 years of someone’s life would be completely dependent on an arbitrary number presented to the judge.

Availability Heuristic
. Another poor use of anchoring can be found in the
availability heuristic
, the tendency to evaluate the probability of an event by how easily an instance comes to our mind. When we receive an anchor, we might reflect on instances where that anchor is true, and if an instance comes to our mind very easily, then we might falsely assume that the anchor is accurate, and so we produce a judgment near that anchor point (Mussweiler & Strack, 2000).

In courtrooms, judges can falsely associate the ease with which a particular sentence length comes to mind with the frequency of that sentencing. If a prosecutor demands 5 years for a crime, the judge may reflect on past cases where that same sentence was given for that crime. If he can easily think of a particular instance, then he might assume that the length of his sentence should also fall near that anchor of 5 years.