Recognition Heuristic
From PsychWiki - A Collaborative Psychology Wiki
Gerd Gigerenzer (1991) has pioneered a cognitive view of heuristics as “fast and frugal” methods by which humans are able to make adaptive decisions based on their environments while utilizing cognitive resources such as knowledge, time, and computation in as efficient a manner as possible (Gigerenzer and Todd, 2000).
Contents |
Gigerenzer (1991) vs. Kahneman and Tversky (1974)
This theory stands against the “heuristics-and-biases” program suggested by Kahneman and Tyversky (1974) with the expressed purpose of understanding the mental shortcuts people take to make both valid and invalid judgements (Gigerenzer, 1996). Although these two approaches are similar in many aspects of their research, the main challenge to the heuristics-as-biases program that Gigerenzer has asserted is that Kahneman and Tversky’s research has emphasized the inaccurate decisions people sometimes come to via the use of heuristics and the ways in which they prevent one from arriving at correct answers to questions in the realm of probability theory (Gigerenzer and Goldstein, 2002). In contrast, Gigerenzer views heuristics as tools that humans use, for the most part, with great accuracy and that ultimately heuristics are more useful for solving even complex probability-based problems because they mirror the way humans actually think instead of expending the energy to consider each alternative in a mechanical manner (Gigerenzer and Todd, 2000).
The Recognition Heuristic and Why it Works
One of the alternative heuristics that Gigerenzer suggests exists that has been strongly supported by the data in research studies is what he calls the recognition heuristic (Gigerenzer and Goldstein, 2002). One of the major critiques of Kahneman and Tversky that Gigerenzer offers is that their representativeness, availability, and anchoring-and-adjustment heuristics are too vague and could refer to any number of cognitive processes and that their interpretation depends solely upon the orientation and personal perspective of the researcher (Gigerenzer, 1996). Alternatively, he presents the recognition heuristic, which is utilized to select one object from a subset based on some criterion (Gigerenzer and Goldstein, 2002). The recognition heuristic is actually based, to a certain extent, on the ignorance of the individual and is only applicable in situations where that ignorance is strongly correlated with the criterion being recognized (Gigerenzer and Goldstein, 2002). Gigerenzer and Goldstein (2002) suggest that the mechanism through which the recognition heuristic operates is ecological rationality, the idea that humans have an ability to exploit the structure of information in its natural environment. For example, the natural structure of human society as well as the way the human mind operates means that larger cities tend to be more salient to individuals. The recognition heuristic exploits this natural structure of human society by allowing an individual to determine the more populous of two cities with reasonable accuracy based on whether or not they recognize it (Gigerenzer and Goldstein 2002).
Research Methodology and Data
To elucidate exactly how this mechanism operates, the discussion of two research studies conducted on the recognition heuristic follows. One, conducted by Gigerenzer and Hoffrage (1995), asked twelve Americans and twelve Germans whether San Diego or San Antonio had a larger population. Although they would naturally be less familiar with the population of U.S. cities, all of the Germans correctly responded that San Diego had the larger population, because it was the more recognizable of the two choices (2002). In this case, there exists a strong, positive correlation between recognition and the criterion (population of the city) (2002). This point is vital in that the recognition heuristic only applies when there is a correlation between recognition and the criterion being studied (2002). A second study, conducted by Ayton and Onkal (1997), found that Turkish people, who had little or no knowledge or interest in soccer, were only 3% less accurate in predicting the outcome of the F.A. cup than more highly invested English soccer fans (2002). This data suggests that the recognition heuristic is as good at predicting outcomes that would be very difficult to quantitatively compute as populations that were not utilizing this heuristic (2002). He has also done some short term studies in which the stocks picked using the recognition heuristic led to equal stock market returns as those chosen by professional stock-pickers (Gigerenzer and Todd, 2000). All of the aforementioned findings were statistically significant. In general, Gigerenzer’s experiments showed that over 90% of people will act in accordance with the recognition heuristic, even if they have been given explicit information that should lead them to stop making this decision (2000). All of the aforementioned findings were statistically significant.
Critiques of This Approach
Oppenheim (2003) suggests that Gigerenzer's (1991) theory is flawed in its assumption that the recognition heuristic was the only mechanism used by participants in Gigerenzaer's studies to determine which of a subset of alternatives to pick. He provides the specific example of a study by Gigerenzer and Goldstein (1999) in which Americans chose the larger of two German cities with 93% accuracy out of a sample of Germany's thirty largest cities in instances where one city was recognized and the other was not(Oppenheim 2003). Oppenheim (2003), however, proposes that the knowledge of the size of the city also may have played a role in the participants' choices. For example, he posits that participants not only recognize the city of Berlin, but also have the knowledge that it is one of the largest cities in Germany (Oppenheim 2003). According to Oppenheim, the recognition heuristic would suggest that a participant would choose a recognizable city over an unrecognizable city even if they knew the former to be very small. He tested this theory through a series of experiments. In the first experiment (2003), Oppenheim recruited fifty participants from the Stanford community and presented them with a survey in which one of six local cities was paired with one of nineteen fictional cities (controlling for the possibility that the participant might recognize the names of both cities. The participants were asked to indicate which city was larger, with the knowledge that some of the local cities were quite small. He found that the participants indicated that the local city was larger only 37% of the time, a statistically significant finding (Oppenheim 2003). In the second experiment (2003), he used cities that are highly recognizable for reasons other than size instead of local cities. Examples of cities used are Chernobyl (nuclear accident), New Haven (site of prestigious university), and Nantucket (subject of popular limerick). Once again, he paired them with fictional cities and asked participants, who were Stanford undergraduates required to complete the survey as part of a course, to indicate which of the two was larger. He found that participants chose the recognizable city as larger only 40% of the time (Oppenheim 2003). This data suggests that recognition is not the only cue utilized by humans in making this type of determination and are inconsistent with the predictions made by the recognition heuristic (2003).
Conclusion and Practical Applications
The recognition heuristic is just one of many such tools that Gigerenzer suggests that humans use to make the best, most efficient decisions but it does provide an excellent example of his theories and is backed by considerable data (Gigerenzer and Todd, 2000). The practical implications of this theory are not limited to describing human behavior but could also be influential in the future development of artificial intelligence technology that could be created to think the way that humans do rather than in the more serial and inherently less efficient method by which computers currently solve problems. The essential theory behind this view of heuristics is that “To behave adaptively in the face of environmental challenges, organisms must be able to make inferences that are fast, frugal, and accurate” (736) and these heuristics are the mechanism through which humans are able to do just that (2000).
Works Cited
Gigerenzer. “On Narrow Norms and Vague Heuristics: A Reply to Kahneman and Tversky." 1996. Psychological Review. Volume 103. 592-596.
Gigerenzer and Goldstein. “Models of Ecological Rationality: The Recognition Heuristic." 2002. Psychological Review. Volume 109. 75-90.
Gigerenzer and Todd. “Precis of Simple heuristics that make us smart." 2000. Behavioral and Brain Sciences. Volume 23. 727-780.
Oppenheim. "Not so fast! (and not so frugal!): rethinking the recognition heuristic." 2003. Cognition. Volume 90. B1-B9.
