People usually disburden memory to external tools to compensate for their limited internal cognitive abilities when they are not confident in their own abilities. However, blended with one’s own abilities, outside assistance may prevent people from accurately assessing the extent to which their own abilities contribute to their performance. In fact, since people tend to overestimate their knowledge and abilities across various types of self-assessments, the use of external assistance may foster people’s overconfidence in their abilities. Fisher and Oppenheimer (2021) investigated whether people’s reliance on external sources disrupts their metacognitive assessments of their own abilities and the underlying mechanism of such a relationship.
Fisher and Oppenheimer (2021) did 8 experiments to examine the underlying reasons for overconfidence across different cognitive domains. The current article focuses on introducing 4 of the experiments about how time delays and active choices of hints influence people’s metacognitive awareness of their own abilities in problem-solving. The researchers recruited American adult participants from Mturk and had them solve 8 anagrams within 90 seconds. Participants in the “Help” condition could see the first 3 letters for the solution below each anagram while participants in the “No help” condition did not receive any help. After the anagram task, participants were asked to predict their future performance on a similar anagram task without hints. The results showed that participants who received help not only solved more anagrams correctly but also predicted higher future performance than those who did not get any hints. The researchers then did a new experiment in which they randomly assigned participants to the “No help”, “Help”, or “Delay” condition. Participants in the “Delay” condition could see the hints 7 seconds after they began. The procedure remained the same. The researchers found that participants who received a time delay in help answered fewer anagrams correctly than those who received immediate help, but performed better than those who did not receive any help. Moreover, participants who received delayed help had lower estimates of their future performance than those who received help, but their estimates were still higher than those who did not receive help at all. These results suggested that receiving immediate and external help leads to overconfidence in estimating future performance on a similar task with no hints, and delayed help can mitigate but not eliminate such tendency.
The researchers also explored whether forcing people to actively seek help instead of providing help as a default would influence their estimates of future performance. They added a new “Button” condition in which participants could hit a “HINT” button below each anagram to see the first three letters of the solution. The results indicated that participants in the “Button” condition performed worse than those who received immediate help, but they did better than those who did not have any hints. These participants also showed lower estimates of their future performance than those who got hints, and their estimates were not different from those who did not receive help. These findings implied that when people had to actively seek help after they had experienced the difficulty of the task, they tended to have a better metacognitive awareness of their own abilities.
In order to further understand the relationship between outside assistance, perceived difficulty, and estimated future performance, the researchers replicated the previous experiment by measuring their perceived difficulty of the anagram task and assessing their accuracy in solving a new set of anagrams that they had made predictions. After subtracting participants’ actual performance on the additional anagram task from their estimated performance, what the researchers found was that participants in the “Button” condition were as good as participants who did not receive help in estimating future performance while participants who received help were the most overconfident. Participants in the “Button” condition also perceived the initial anagram tasks as harder than participants who received help. Mediation analysis also showed that perceived difficulty played a significant role in the overconfidence levels that varied across the “Button” and “Help” conditions.
Taken together, the findings of these experiments pointed out that people tend to assess their own abilities better when external assistance is delayed or actively chosen. Also, the feeling of difficulty is needed for people to have a more accurate understanding of their abilities. These main takeaways also have practical implications for the design and usage of learning-support technology in educational settings. For instance, for students to accurately assess their understanding of certain concepts while playing a learning game, the game designer can provide the students with a “HINT” button. In this way, students can experience the difficulty of the task first, and then decide whether they need external help to win the game.
Fisher and Oppenheimer’s (2021) study also provides insights into EPIC’s research studies about what failure means to people and what people can learn from failure. Since the perceived difficulty of tasks can help people assess their own abilities better, perhaps failure experiences can achieve the same result if people perceive the task that they fail at as difficult. Besides people’s metacognitive awareness, future research could also investigate if the level of perceived difficulty influences the extent to which people seek external help.
For more information about Fisher and Oppenheimer’s (2021) study, check out the link below for the journal article:
https://psycnet.apa.org/record/2021-38384-010
Reference:
Fisher, M., & Oppenheimer, D. M. (2021). Harder than you think: How outside assistance leads to overconfidence. Psychological Science, 32(4), 598-610.
