Abstract
We reflect back on our 2004 monograph reviewing the implications of faulty self-judgment for health, education, and the workplace. The review proved popular, no doubt because the importance of accurate self-assessment is best reflected in just how broad the literature is that touches on this topic. We discuss opportunities and challenges to be found in the future study of self-judgment accuracy and error, and suggest that designing interventions aimed at improving self-judgments may prove to be a worthwhile but complex and nuanced task.
“Know Thyself” admonished the inscription on the Temple of Apollo, where the Delphic Oracle of the ancient Greek world was housed. Philosophers and social commentators have been repeating this exhortation ever since. “The peak of knowledge is perfect self-knowledge,” asserted Richard of Saint-Victor, the 12th-century Scottish theologian. “Self-knowledge is the beginning of self-improvement,” advised Spanish philosopher Baltasar Gracián in the 17th century. “Knowing others is intelligence,” observed the Chinese philosopher Laozi all the way back in 6th century BCE, “but knowing yourself is true wisdom.”
The Original Message
Thus, it is not a surprise that a monograph examining the accuracy of self-views—and its relevance to education, health, and the workplace—would be something of a natural topic in psychology as well as something of a hit to readers, especially when the article suggested that knowing thyself was a challenging task that people only imperfectly executed, at best (Dunning, Heath, & Suls, 2004).
The article contained two main headlines. First, people’s self-perceptions of skill and the reality of that skill correlate, at best, only moderately. At worst, they correlate not at all (see also Mabe & West, 1982; Zell & Krizan, 2014). Indeed, at times, the judgments of other people anticipate a person’s outcomes better than that person’s own self-judgments (e.g., Risucci, Torolani, & Ward, 1989).
Second, in a myriad of ways, people overrate themselves. They tend to hold false perceptions of superiority when comparing their competence and character to that of their peers (Alicke & Govorun, 2005). When contemplating the future, they overpredict favorable outcomes while underpredicting unpleasant ones (Weinstein, 1980). They provide overly optimistic estimates of how quickly they will complete tasks and projects (Buehler, Griffin, & Ross, 1994). They imbue their predictions with too much confidence, overestimating the odds their estimates and forecasts will prove right (Lichtenstein, Fischhoff, & Phillips, 1982).
The monograph described how these basic difficulties in self-judgment applied to health, education, and the workplace. In health, for example, we described how people underestimate their health risks and then fail to moderate mistaken views of invulnerability even after being asked to list attributes that put them at risk, for example, for obesity or alcoholism. However, if asked to list protective attributes instead, they were more than happy to provide revised estimates suggesting even more invulnerability (Weinstein & Klein, 1995). In education, we described how badly students judge their comprehension of educational materials (Maki & Berry, 1984). In the business world, we described how CEOs on average tend to make overly hubristic strategic decisions—only to have the value of their firm’s stock punished by outsiders as a consequence (Malmendier & Tate, 2008).
Legacy
The Delphic Oracle was correct; accurate self-knowledge is crucial, and so it is not a surprise that the topic of self-assessment has arisen separately in social, cognitive, personality, abnormal, organizational, education, health, and developmental psychology (Dunning, 2005; Vazire & Wilson, 2012; Wilson, 2009). Every corner of psychology contains some version of the flawed self-assessment issue. For example, an active area in neuroscience has arisen to explain why people hold such overly rosy views of self (Beer, 2014), as well as overly optimistic forecasts about the future (Sharot & Garrett, 2016). Indeed, if one ventures outside of psychology, one sees active research in self-assessment and its flaws. Flawed self-views are studied in physics (Kohl & Finkelstein, 2005), finance (Gervais & O’Dean, 2001), medicine (Eva & Regehr, 2005), and computer science (Murphy & Tenenberg, 2005), to name just a few disciplines.
Thus, the major contribution of the monograph was to take an extremely broad literature on self-assessment and distill it all together, as much as possible, into a single source document. Basic findings on the topic were reviewed, the fundamental psychological mechanisms underlying those findings were delineated, and specific problems arising in health, education, and business because of faulty self-views were described.
However, in performing that service, we were bound to fail. There is just too much of a broad literature to corral into one document. Furthermore, the extreme breadth of self-assessment scholarship means that it will always be an active literature, and so any review of it is likely out of date just as soon as its last sentence is written.
Moving Forward
Looking back at the 12 years since the publication of the monograph, would we have done anything differently? Probably not, given that space in the original monograph was limited. The intervening years, however, have sharpened some issues, some which have been addressed, and some of which have yet to be addressed adequately by scientific research as it moves forward.
Nudging self-accuracy
First and foremost, the intervening years in psychological research have taken a traditional emphasis on documenting problems, and understanding the psychological dynamics that produce them, to a newer focus on engineering psychological techniques that reduce or eliminate those problems. In short, we live in the era of “nudge,” concentrating on interventions designed to improve the human condition and produce social benefit (Thaler & Sunstein, 2008). What can people do, for example, to help their peers fulfill that ancient admonition from the Delphic Oracle about knowing thyself?
To be sure, we talked about possible nudges in the original monograph. In education, for example, we pointed out that students could profit from peer assessments to achieve better self-understanding of their academic performance, in that peers tend to provide evaluations that better match what instructors think (Lennon, 1995; Sullivan, Hitchcock, & Dunnington, 1999). We also discussed educational practices to avoid. Namely, massed study (what students know as “cramming”) leads to quick learning and student satisfaction. Unfortunately, it also leads to rapid forgetting, leaving students with illusory self-impressions of competence about material they soon no longer remember. Instead, best practices suggest that educators should spread out training over time, making it more random and difficult. This “distributed” approach produces less student satisfaction and confidence, but actually promotes better long-term learning and retention of material—in short, a less overconfident but more competent student (Bjork, 1999).
One wonders what other nudges might be possible to conjure to improve self-assessment. We live in a world where people have access to more and more data about one’s personal biological, physical, behavioral, or environmental information (Swan, 2013). One can slap on a wristband or wristwatch and get an online record of how many steps one has taken today, how much sleep one has gotten, and how many calories one has burned. Can this improved access to quantified data, and thus to a “quantified self,” lead to more accurate self? If so, can it be applied somehow to other areas of life, such as one’s moral character, that seem to be less amenable to quantification?
Furthermore, we live in a world where the error pattern in self-judgment is uneven across cultures. Illusory superiority, the unrealistic tendency for the average person to think she or he is anything but average, appears endemic to North American and Western Europe cultures, but is much more sparse in Far Eastern cultures (Heine, Lehman, Markus, & Kitayama, 1999). What practices or principles might be present, or absent, in East Asian cultures that lead to less bias? Can those practices be adopted more universally in places where people express a false sense of superiority? Current evidence tentatively suggests that people the world over fall prey to a few reasoning errors that lead to self-judgment error, but that self-esteem pressures to think well of oneself is specific only to a few cultures (Heine & Hamamura, 2007; Rose, Endo, Windschitl, & Suls, 2008). Can that insight be used to fashion a self-accuracy nudge in the West?
Possible complications
Whatever path the search for nudges takes, we can imagine that the path will be filled with detours and surprises. For over 100 years, people have systematically tested ways to give feedback to others to motivate self-insight and improvement, and the net effect of those interventions has been that nearly 40% make things worse, not better (Kluger & DeNisi, 1996). Simply giving people feedback about their skill does not guarantee understanding and improvement. For example, feedback often leads the people who need it the least, rather than those who need it the most, to energize themselves toward self-improvement. Consider one study in which business students were tested on their emotional intelligence, given feedback about their performance, and then offered an opportunity to buy a book on improving emotional intelligence at a discount. Of those who performed the best, nearly two thirds bought the book. Of those who performed the worst, and presumably needed the most help with their emotional intelligence, only 20% bought the book (Sheldon, Dunning, & Ames, 2014).
Indeed, one may wonder what the most beneficial type of nudge might be when it comes to self-judgment and understanding. Perhaps accuracy in self-judgment cannot be taught, in that people harbor motivations that might combat unfriendly feedback about the self—and then win (Dunning, 2015; Kunda, 1990). As a consequence, training people for self-accuracy, at least under some circumstances, may prove impossible because of self-deceptive defenses and ego-protective reasoning. In those cases, will the best nudge to produce self-assessment accuracy be to turn to some other person, some surrogate, to make the judgment? Recent evidence suggests that other people can be “trained up” to give rather accurate judgments about other people, without the undue optimism that often pervades self-judgment (Helzer & Dunning, 2012).
Then again, the best nudge may not be aimed at self-perception at all. Rather, the best way to make people more accurate about the self is not to change the judgments they make about the self but rather to change the self they are judging. A recent analysis of overconfidence suggests it may not be how optimistic or pessimistic people are about themselves that matters. Instead, overconfidence tracks their objective performance. People who perform poorly show marked levels of overconfidence; however, those who perform well show much more accuracy in their self-judgment (Dunning & Helzer, 2014). Thus, if one wants to promote self-accuracy, it may not pay to counsel people to be pessimistic. Rather, the key will be to enhance people’s objective performance, to make their actual self match the rather favorable self-views that people already characteristically hold of themselves.
Costs and benefits
Finally, in our original monograph, we examined specifically those areas where it was clear that accurate self-assessment was important. Stock traders should not be overconfident. Students should know when they have studied enough. People need to recognize where their amateur medical expertise ends and the need to consult a licensed doctor begins.
During the time of the rich literature reviewed in the monograph, there has been an active literature about whether accuracy in self-assessment is actually adaptive. The Delphic Oracle exhorted people to know themselves the truth about themselves, but there have been contrary voices arguing that the advantage runs to those who are unrealistically positive. Taylor and Brown (1988), in a classic and provocative article, proposed that overly optimistic self-views led to a happy life of creativity and productivity. Since this landmark article, dozens of articles have been published, some exhorting the superiority of judgmental realism with others championing the benefits of congenial bias.
The vast majority of work about the adaptiveness of realism versus bias is solid. However, taken as a whole, it all becomes inchoate. A comprehensive reader of the literature will see dozens of isolated demonstrations linking realism to better outcomes, juxtaposed with dozens of other isolated demonstrations showing the opposite. However, with a few exceptions, there seems to be no thematic principles that sort those situations that favor realism from those that support bias. As such, moving into the future, searching for those themes must be placed higher on the research agenda for self-psychology.
Conclusion
The ancient Greek philosopher Thales was once asked to name the most difficult task a human can face. He nominated gaining accurate self-knowledge. Thus, it is not a surprise that a monograph largely agreeing with Thales would resonate with a broad audience. However, it is our hope that the monograph, in the areas of health, education, and work, made it slightly less difficult to succeed at this essential task, making people somewhat better oracles when it comes to understanding and judging the self.
Footnotes
Acknowledgements
The views and opinions expressed in this article are those of the authors and do not necessarily represent the views of the National Institutes of Health or any other governmental agency.
Declaration of Conflicting Interests
The author(s) declared that there were no conflicts of interest with respect to the authorship or the publication of this article.