If individuals are asked to estimate some quantity--such as the number of errors that a colleague has committed in one year--their final judgment is biased by any numbers they had recently observed. This tendency, which reflects a heuristic called anchoring and adjustment, diminishes if individuals press downwards on a table while they contemplate the issue.
Tversky and Kahneman (1974) introduced the concept of anchoring and adjustment. Specifically, when individuals need to reach some judgment--perhaps the price they are willing to pay to purchase a particular car or the number of jelly beans in a jar--they form an initial judgment from some simple feature and then adjust this estimate, called an anchor, to form a final judgment. The adjustment, however, is usually conservative, and hence the final judgment is usually biased towards the anchor.
To illustrate, in a study conduced by Tversky and Kahneman (1974), a random number was generated by spinning a wheel. Participants were then asked to specify whether this random number was higher or lower than was the percentage of nations that are located in Africa--referred to as a comparative question. Finally, participants were instructed to estimate the percentage of nations that are located in Africa--an absolute question. Participants who had received a high random number were more inclined to overestimate the percentage of nations that are located in Africa. The anchor, as represented by the random number, biased their final estimate.
Anchoring and adjustment is assumed to bias estimates of risk and uncertainty (Yamagishi, 1994), judgments of self-efficacy (Cervone & Peake, 1986), and predictions of future performance (Czaczkes & Ganzach, 1996). As a consequence, the anchoring and adjustment heuristic is often touted as robust and persistent (Chapman & Johnson, 2002).
Anchoring and adjustment has been demonstrated outside laboratories in natural settings. In a study by Northcraft and Neale (1987), for example, participants received comprehensive information regarding a property, including either a high or low list price, before touring the property. Finally, participants were asked to estimate the actual value of the property. When the list price was high, final estimates also tended to be elevated--an observation that applied to both amateur and experts in real estate.
Anchoring and adjustment also explains why sales employees often introduce the premium brand or option first. For example, some sales employees might describe the most expensive model first, such as the premium TV, before presenting the alternatives. Other sales employees might describe a typical model first, such as an average TV, before presenting the alternatives. If customers are introduced to the premium rather than average model first, they are especially likely to purchase a more expensive option, as shown by Donoho (2003).
Presumably, when sales employees describe the most expensive model first, this premium alternative becomes the anchor. The customers will adjust their preferences only marginally from this anchor.
Although the reliability of this anchoring effect has been established (e.g., Mussweiler & Strack, 2004), the mechanisms that underpin these findings have been debated vigorously (e.g., Epley & Gilovich, 2006). Initially, the finding that final estimates are biased towards initial anchors was ascribed to insufficient adjustment: Individuals simply do not adjust their initial estimates adequately (Epley & Gilovich, 2004).
Alternatively, the anchor--the initial estimate--might activate arguments or memories that support the validity of this value (Strack & Mussweiler, 1997). In the study conducted by Tversky and Kahneman (1974), for example, when participants were whether or not the answer exceeds the initial anchor, such as 30%, they consider evidence and arguments that might corroborate this value. They might, for example, recognize that Africa constitutes less than 30% of the landmass on Earth but, nevertheless, includes many smaller nations. They might also recall an instance in which a teacher maintained the percentage of nations that are located in Africa exceeds 25%. This evidence then biases their final estimate.
The selective accessibility model, proposed by Mussweiler (1997), formalizes this explanation. According to this model, the attention and memory of individuals is, in general, biased towards evidence that confirms their hypotheses or estimates. As a consequence, attention and memory will tend to be directed towards the anchor.
Wegener, Petty, Detweiler-Bedell, Jarvis, and Blair (2001) extended this model, highlighting that only some anchors will generate evidence and memories that confirm the initial value. Extreme anchors--that is, estimates that are patently too high or low-might generate evidence or memories that contradict the initial value. For instance, if participants are asked to specify whether the percentage of nations that are located in Africa is greater or less than 98%, evidence that contradicts this value is likely to be salient. As a consequence, the final estimate of participants will not be biased towards this anchor. Consistent with this premise, Wegener et al. (2001) did discover that extreme anchors are less inclined to bias estimates, contrary to predictions derived from the selective accessibility model.
Pahl and Eiser (2007) demonstrated that anchoring and adjustment can be applied to mitigate the tendency of individuals to overestimate their character, competence, and contributions relatives to peers--an inclination they designate as comparative self-positivity. Specifically, after participants were asked whether or not they typically engage in more than eight acts each week that could harm an ecosystem, their tendency to perceive themselves as more concerned with the environment than a typical person subsides. If instructed to specify whether or not they typically engage in more than two acts each week that could harm an ecosystem, this tendency is amplified. They begin with a low anchor-two acts-and, therefore, tend to assume they seldom engage in acts that could damage the environment.
Epley, Keysar, Van Boven, and Gilovich (2004) showed that individuals tend to overestimate the extent to which the attitudes and perspective of everyone else is similar to their own--a consequence of anchoring and adjustment. In their study, participants were granted the opportunity to consume half a can of Coke and then half a can of Pepsi--with the labels attached. As a consequence, these participants could readily distinguish between Coke and Pepsi. These participants tended to overestimate the percentage of other individuals who could determine which of two unlabelled cups of cola contained Coke and which contained Pepsi. That is, they imposed their own awareness of the difference between Code and Pepsi onto other individuals who had not received the same level of exposure to these drinks. This bias dissipated when participants were encouraged to contemplate the attitudes and perspectives of other individuals more carefully and cautiously.
Because of the mechanisms that underpin anchoring and adjustment, people often project their own motivations onto other individuals. That is, they assume that both they and other people tend to share similar motivations and preferences.
Douglas and Sutton (2011) invoked this concept of projection to explain a fascinating discovery. Specifically, as their research showed, people who believe in conspiracy theories are the very people who would be more willing to engage in such conspiracies themselves. To illustrate, the people who believe the September 11 attacks were orchestrated by the US government are also more willing to concede that, if they were government officials, they may also have condoned or ordered this attack.
According to Douglas and Sutton (2011), individuals cannot be certain of the motivations and preferences of other people. To predict these motivations or preferences, they imagine their own needs, goals, or values in these settings. They may then project these needs, goals, or values onto other people. For example, if they would be willing to conspire, they assume other people would be willing to conspire as well: They thus endorse conspiracy theories.
To demonstrate, in one study, some participants completed an exercise that highlighted their tendencies to engage in moral behavior. That is, they were asked to remember a time in which they helped someone. Other participants did not complete this exercise.
Next, they read 17 statements about conspiracies, such as the notion that terrorists were not responsible for the September 11 attacks or that Princess Diana did not actually die in an accident. They were asked to imagine, for a moment, they were one of the conspirators. Next, they were asked to indicate the likelihood they would be willing to perpetrate these conspiracies. Finally, they were asked to indicate the probability these conspiracies are genuine.
Interestingly, if people claimed to be somewhat willing to perpetrate these conspiracies, they were more likely to believe these conspiracies were genuine. Furthermore, when moral identity was salient, people were both less willing to perpetrate these conspiracies and less likely to believe these conspiracies were genuine.
Wansink, Kent, and Hoch (1998) showed that customers tend to purchase more goods from a retailer when each price refers to several items, such as "$2.00 for 4 cans of Heinz soup", than one item, such as "$0.50 for each can of Heinz soup". The number of items described in the price--such as 4 cans or 1 can--acts as an anchor and thus biases the final judgment on how many cans to purchase.
In the paradigm applied by Tversky and Kahneman (1974), participants first considered whether or not the anchor exceeds the correct answer before communicating their final judgment. Many variants of this paradigm have demonstrated the anchoring and adjustment effect. First, even when the anchor is entirely arbitrary, the effect persists (Wilson, Houston, Etling, & Brekke, 1996). That is, the effect is maintained even when participants compare the anchor to an estimate of one quantity-such as the percentage of nations that are located in Africa--but then need to estimate another quantity-such as physicians and surgeons in the local telephone book. Second, the effect persists even when participants are encouraged to minimize the bias, either through incentives to be accurate or through warnings about the ubiquity of this distortion (Wilson et al., 1996). The effect does diminish, however, if participants have developed extensive knowledge about the topic in which they need to form some estimate (Wilson et al., 1996).
Interestingly, the anchoring and adjustment bias diminishes if individuals are asked to place their palms on a table and press downwards (Epley & Gilovich, 2004). When individuals engage in this movement, the initial estimate is less likely to bias the final judgment. This movement tends to activate a specific set of muscles, including the triceps. Throughout evolution, these muscles tend to be activated when humans rejected some object--that is, shoved objects away from the body--and hence activate a state called avoidance. As a consequence, when individuals activate these muscles, they unwittingly reject the first number they retrieved, and this bias thus dissipates.
To increase sales, prices should refer to multiple items. For example, suppose that most customers tend to purchase 6 boxes. The price for 10 boxes should be specified, such as "$20 per 10 boxes". Second, salespersons should casually specify the number of items that anothe customer purchased. They might mention that "Frank yesterday purchased 12 boxes".
Anchoring and adjustment can also bias the assessments of employees. That is, during some performance appraisals, employees are permitted to rate themselves, called self-ratings. These self-ratings are sometimes conceptualized as anchors by managers. The evaluations of managers, therefore, are skewed towards these self-ratings.
To illustrate, in one study, conducted by Chen and Kemp (2012), participants received considerable information about the performance of an academic, such as the publication record of this person. In addition, the participants received a self-rating of this academic on these three facets of performance: teaching, research, and service. If these self-ratings were favorable, participants were more inclined to rate the academic favorably. The self-ratings, therefore, skewed the assessment of all objectives sources of information. As subsequent studies showed, this effect persisted even if the participants were told the self-ratings tend to be biased. Similarly, this pattern was observed after individuals were informed the self-ratings were entirely contrived and arbitrary. Therefore, self-ratings, even if perceived as misguided, tend to bias evaluations.
Cervone, D., & Peake, P. K. (1986). Anchoring, efficacy, and action: The influence of judgmental heuristics on self-efficacy judgments and behavior. Journal of Personality and Social Psychology, 50, 492-501.
Chapman, G. B., & Bornstein, B. H. (1996). The more you ask for, the more you get: Anchoring in personal injury verdicts. Applied Cognitive Psychology, 10, 519-540.
Chapman, G. B., & Johnson, E. J. (1994). The limits of anchoring. Journal of Behavioral Decision Making, 7, 223-242.
Chen, Z., & Kemp, S. (2012). Lie hard: The effect of self-assessments on academic promotion decisions. Journal of Economic Psychology, 33, 578-589. doi:10.1016/j.joep.2011.11.004
Donoho, C. L. (2003). The "top-of-the-line" influence on the buyer-seller relationship. Journal of Business Research, 56, 303-309.
Douglas, K. M., & Sutton, R. M. (2011). Does it take one to know one? Endorsement of conspiracy theories is influenced by personal willingness to conspire. British Journal of Social Psychology, 50, 544-552. doi: 10.1111/j.2044-8309.2010.02018.x
Epley, N., & Gilovich, T. (2004). Are adjustments insufficient? Personality and Social Psychology Bulletin, 30, 447-460.
Epley, N., Keysar, B., Van Boven, L., & Gilovich, T. (2004). Perspective taking as egocentric anchoring and adjustment. Journal of Personality and Social Psychology, 87, 327-339.
Jacowitz, K. E., & Kahneman, D. (1995). Measures of anchoring in estimation tasks. Personality and Social Psychology Bulletin, 21, 1161-1166.
Mussweiler, T., & Strack, F. (1999). Hypothesis-consistent testing and semantic priming in the anchoring paradigm: A selective accessibility model. Journal of Experimental Social Psychology, 35, 136-164.
Mussweiler, T., & Strack, F. (2000). Numeric judgments under uncertainty: The role of knowledge in anchoring. Journal of Experimental Social Psychology, 36, 495-518.
Northcraft, G. B., & Neale, M. A. (1987). Experts, amateurs, and real estate: An anchoring-and-adjustment perspective on property pricing decisions. Organizational Behavior and Human Decision Processes, 39, 84-97.
Pahl, S., & Eiser, J. R. (2007). How malleable is comparative self-positivity? The effects of manipulating judgmental focus and accessibility. European Journal of Social Psychology, 37, 617-627.
Ritov, I. (1996). Anchoring in simulated competitive market negotiation. Organizational Behavior and Human Decision Processes, 67, 16-25.
Strack, F., & Mussweiler, T. (1997). Explaining the enigmatic anchoring effect: Mechanisms of selective accessibility. Journal of Personality and Social Psychology, 73, 437-446.
Tversky, A., & Kahneman, D. (1974). Judgment under uncertainty: Heuristics and biases. Science, 185, 1124-1131.
Wansink, B., Kent, R. J., & Hoch, S. J. (1998). An anchoring and adjustment model of purchase quantity decisions. Journal of Marketing Research, 35, 71-81.
Wegener, D. T., Petty, R. E., Detweiler-Bedell, B. T., Jarvis, W., & Blair G. (2001). Implications of attitude change theories for numerical anchoring: Anchor plausibility and the limits of anchor effectiveness. Journal of Experimental Social Psychology, 37, 62-69.
Wilson, T. D., Houston, C. E., Etling, K. M., & Brekke, N. (1996). A new look at anchoring effects: Basic anchoring and its antecedents. Journal of Experimental Psychology: General, 125, 387-402.
Wright, W. F., & Anderson, U. (1989). Effects of situation familiarity and financial incentives on use of the anchoring and adjustment heuristic for probability assessment. Organizational Behavior and Human Decision Processes, 44, 68-82.
Last Update: 6/1/2016