Why is our confidence disproportionate to the difficulty of a task?

The Hard-easy effect

, explained.
Bias

What is the hard-easy effect?

The hard-easy effect, also known as the discriminability effect or the difficulty effect, occurs when we incorrectly predict our ability to complete tasks depending on their level of difficulty. It suggests that we are overly confident in how successful we will be at hard tasks and under-confident about how successful we will be at easy ones. 

Where this bias occurs

The hard-easy effect occurs in any situation where we are asked to predict or comment on our success regarding a task. Our confidence usually doesn’t match our actual ability, which can cause us to be ill-prepared when taking on hard tasks, and stress over easy tasks. While the hard-easy effect is usually demonstrated through our inability to correctly identify whether we got an answer right or wrong, it also can have negative real-life applications.  

For example, imagine that you are learning how to drive. To get your license, you have to complete both a written exam and a driving test. Arguably, the driving test is the more difficult component of obtaining a license, because it is based on skill and knowledge, whereas the written exam is mostly based on knowledge. However, according to the hard-easy effect, you are likely to be over-confident in your ability to pass the driving test and under-confident on how likely you are to pass the written exam. As a result, you barely practice driving and focus all your time on studying for the written task. When it comes to the day of the exam, the hard-easy effect may have made it easy for you to pass the written exam, but since you were too confident about your driving skills, you fail the driving test.

Debias Your Organization

Most of us work & live in environments that aren’t optimized for solid decision-making. We work with organizations of all kinds to identify sources of cognitive bias & develop tailored solutions.

Learn about our work

Individual effects

The aforementioned hypothetical example of incorrectly predicting one’s success for different components of obtaining a license demonstrates that the hard-easy effect can have serious impacts on our behavior, not just our beliefs. The hard-easy effect describes an incorrect mindset, which is based on our inability to accurately make predictions about the future. Our present behavior is usually modeled based on our future predictions, which means that we make decisions that are not in our best interest. 

Systemic effects

The hard-easy effect may seem like a bias that is reserved for small issues, like trivia and exams. However, overconfidence, which is half the problem of the hard-easy effect, can have some pretty dire consequences on our society.

Researchers have suggested that overconfidence may be the explanation behind wars, strikes, litigation, and entrepreneurial failures.1 Military battle, taking legal action, and running a successful business are all incredibly difficult and complex tasks. When we overestimate our ability to be successful in these hard tasks, we may embark on endeavors that we are not fully prepared for. We don’t do the necessary work required for success, and failed endeavors, in the aforementioned areas, can have pretty negative consequences on society. Engaging in wars that we are sure to lose can lead to unnecessary deaths, litigation is often very expensive for any party involved, and failed businesses can lead to disappointed investors. 

Additionally, if we are bad at predicting our own abilities, then it follows that we are even worse at predicting the ability of others, whom we have less knowledge about.1 If we are incorrect about how successful our peers will be on a task, it is difficult for companies to efficiently delegate and run operations. Easy tasks may be assigned to people with positions higher up in the company, due to under-confidence in easy tasks, whilst hard tasks may be assigned to less experienced employees.1

Why it happens

There are multiple different theories as to what is behind the hard-easy effect.

Most of the people who investigate the hard/easy effect believe that it occurs as a result of other biases. The hard-easy effect is very similar to the Dunning-Kruger effect, which also describes the gap that exists between perceived and actual competence. The gap may occur because when we lack the knowledge to be good at a particular topic, we also lack the knowledge of our deficiency of that topic. A lack of understanding of the topic makes it difficult for us to assess how much we know about it. 

When it comes to memory recollection, our accuracy is hindered by cognitive biases. This leaves a lot of room for errors when we try to retrieve information. To answer a difficult question, we have to try and recollect that information, making us susceptible to memory biases. However, while these biases affect the accuracy of our recollection, they do not affect our confidence. This may account for our overconfidence in answering difficult questions.2

Our overconfidence may also be caused by confirmation bias. Confirmation bias stipulates that we tend to give greater weight to things that confirm our pre-existing beliefs and ignore evidence that goes against them. In rating our confidence on whether we have the correct answer, our brain remembers the process through which it arrived at that answer instead of looking to other pieces of knowledge that would lead to a different answer. We are unlikely to try and think of reasons as to why our answer might be wrong, we focus on why it might be right.3

But what explains our lack of confidence for easy questions? Little research has tried to pinpoint the cause of the hard-easy effect, but perhaps it may also be due to another cognitive bias, bikeshedding. Bikeshedding describes our tendency to spend a disproportionate amount of time on trivial tasks. The longer we spend thinking of an answer to an easy question, the more likely we are to conjure up evidence that our answer is wrong. This may then cause us to be less confident in our answers, compared to hard questions where because we don’t know enough about it, we only consider the evidence that points towards us being right. 

Why it is important

We face uncertainty on a daily basis, and in order to effectively maneuver throughout the world, we need to be able to deal with that uncertainty in a rational manner. The hard-easy effect suggests that we approach uncertainty irrationally, as we are not good predictors of our abilities or outcomes. We constantly have to try and make predictions about our own future behavior, and this informs many of our present-day decisions. If our decisions are based on an inaccurate perception of our abilities, these decisions may lead to suboptimal outcomes.

The hard-easy effect essentially suggests that we are not being realistic with our abilities. When it is extrapolated outside of our misplaced confidence in whether we got an answer right, our overconfidence in being right can lead us to be narrow-minded. In thinking we are right, we will be unlikely to absorb information that contradicts us. In attempting to complete a difficult task, the hard-easy effect indicates that we are likely to believe we can do it by ourselves, and therefore we may not be seeking out help when we need it. 

 On the other side of the coin, if we don’t believe we can achieve success in easy tasks, we may decide not to take on those tasks. That means that we are selling ourselves short and not taking advantage of opportunities that come our way. 

How to avoid it

There are a number of biases that affect our ability to make accurate predictions, which makes it difficult to try and ensure that our confidence is well calibrated to our performance. However, a follow-up study conducted by the researchers who first demonstrated the hard-easy effect suggests that the more knowledgeable people are on a topic, the more accurate they are with their confidence in their answers.4

When we know a lot about a topic, we are able to feel more confident about whether the question being asked is something we know the answer to. If we know very little about a topic, all of our answers are essentially guesses, making it more difficult to predict whether they are right. To become more accurate in predicting our success, we may need to be more knowledgeable about the topics on which we are making predictions. However, Lichtenstein found that knowledgeability only reduced the margin of error when it comes to being overconfident on hard tasks. The results showed that knowledgeable people were still under-confident in answering easy questions.4

How it started

Psychologists Sarah Lichtenstein and Baruch Fischhoff, researchers with a special interest in behavioral psychology, first examined the hard-easy effect in 1977.4 Realizing the ubiquity of prediction-making in our daily lives, the researchers wanted to examine how confidence came into play when it comes to probability self-assessments. 

Lichtenstein and Fischhoff conducted a number of experiments to examine whether people made accurate predictions of their own success and failure.4 Their experiments ranged in difficulty to examine how self-assessment varied depending on the difficulty of a task. 

In an experiment that incorporated a more difficult task, Lichtenstein and Fischhoff asked participants to determine the nationality of the artist for 12 different drawings. The only two options were that the artist was a European child, or that the artist was an Asian child. After providing their answer to each drawing, the participants were also asked to indicate how probable they thought it was that they’d gotten the answer right. The results of the experiment showed that participants had gotten the answer right 53% of the time.4 However, the mean response for correct answers was almost 68%.4 This shows a discrepancy between confidence and actual performance, with participants believing that they performed better than in reality. 

In a subsequent experiment, Lichtenstein and Fischhoff asked participants to answer 150 general trivia questions and to once again note how the probability that their answer was right. In examining the results, the researchers split the questions into “easy” and “hard” ones. Lichtenstein and Fischhoff found that for the easy questions, there was a discrepancy between the number of people who got the answer right and the probability they had indicated of it being right. Participants showed under confidence, as they indicated a 60% probability of being right when 75% of participants answered it correctly.4

By comparing the results from various experiments, Lichtenstein and Fischhoff concluded that people show over-confidence when it comes to predicting their success for hard tasks and show under confidence when it comes to predicting their success for easy tasks.4 This phenomenon later became known as the hard-easy effect.

Example 1 - The hard-easy effect in holistic tests

Most of the studies that have shown evidence of the hard-easy effect have looked at the levels of confidence and success rates for individual questions. They measure confidence and success based on individual items. What happens if a whole exam is easy, or hard? Is the hard-easy effect still seen?

This was a question that Lichtenstein and Fischhoff, the psychologists who first researched the hard-easy effect, wanted to answer. In a subsequent study, the researchers used the results from earlier tests to determine a set of easy questions and a test of hard questions.4 Participants then either received a test with only easy questions, or a test with only hard questions. 

Lichtenstein and Fischhoff still found that participants who had been administered the easy test displayed under confidence when predicting how successful they’d been on the exam, while participants who had been administered the hard test were overly confident in how well they’d performed.4 These results indicate that the hard-easy effect is seen for holistic tasks as well as subset tasks. 

Example 2 - Intelligence

While Lichtenstein and Fischhoff suggest that greater knowledgeability about a specific topic makes people better at predicting their success for difficult questions, in another study they conducted, they found the same was not true for intelligence.4

We may believe that the more intelligent we are, the better we are at predicting our abilities. However, the trope about book smarts not indicating street smarts may be supported by the hard-easy effect, as academic intelligence did not seem to make people better at calibrating their confidence in their own abilities.

Lichtenstein and Fischhoff compared the results from two experiments in order to see whether intelligence had an impact on the hard-easy effect. One experiment had been conducted on undergraduate students and one was conducted on graduate students. They assumed that graduate students would, on average, have higher intelligence scores. While Lichtenstein and Fischhoff found that graduate students had better test scores, they were not significantly better at predicting their performance. As a result, the psychologists concluded that intelligence, at least in terms of academic intelligence, does not help negate the hard-easy effect.4

Summary

What it is

The hard-easy effect is a prediction bias. It suggests that we are overconfident in our ability to complete difficult tasks, and under-confident in our ability to be successful in easy tasks.

Why it happens

There is a lot of debate as to why the hard-easy effect exists. Most theories rest on the premise that other biases lead to the effect. For example, memory recollection biases mean that there is room for error when we answer a question, but that we don’t adjust our confidence accordingly. 

The disproportionate amount of information we know about easy versus difficult tasks may also be a cause of the hard-easy effect. For difficult questions, we have less knowledge about the topic, meaning that we are less likely to think of evidence that goes against our answer. When the question is easy, we may know more about the topic, and be able to come up with evidence that goes against our answer, making us less confident in it. 

Example 1 - It doesn’t just occur for subsets

While most of the studies conducted to examine the hard-easy effect measure whether confidence is accurate in predicting success for individual questions or items, it has been found that the hard-easy effect is also seen for overall tasks. If an entire exam is easy, people are under-confident in how well they did, whilst they are overconfident in how well they did in a hard test.

Example 2 - Intelligence doesn’t negate the hard-easy effect

We may believe that the more intelligent we are, the more realistic we would be, making us better at predicting our success in different tasks. However, it seems as though intelligence, at least in terms of academic intelligence, has no impact on the hard-easy effect. Being intelligent is not enough to avoid the hard-easy effect.

How to avoid it

Whilst intelligence may not negate the hard-easy effect, some research has suggested that knowledgeability about a specific topic may help counter being overconfident in how well we performed on difficult tasks. 

Sources

  1. Moore, D. A., & Healy, P. J. (2008). The trouble with overconfidence. Psychological Review115(2), 502-517. https://doi.org/10.1037/0033-295x.115.2.502
  2. Skala, D. (2008). Overconfidence in Psychology and Finance – an Interdisciplinary Literature Review. Munich Personal RePEc Archive, 33-50. https://mpra.ub.uni-muenchen.de/26386/
  3. Suantak, L., Bolger, F., & Ferrell, W. R. (1996). The hard–easy effect in subjective probability calibration. Organizational Behavior and Human Decision Processes67(2), 201-221. https://doi.org/10.1006/obhd.1996.0074
  4. Lichtenstein, S., & Fischhoff, B. (1977). Do those who know more also know more about how much they know? Organizational Behavior and Human Performance20(2), 159-183. https://doi.org/10.1016/0030-5073(77)90001-0

About the Authors

Dan Pilat's portrait

Dan Pilat

Dan is a Co-Founder and Managing Director at The Decision Lab. He is a bestselling author of Intention - a book he wrote with Wiley on the mindful application of behavioral science in organizations. Dan has a background in organizational decision making, with a BComm in Decision & Information Systems from McGill University. He has worked on enterprise-level behavioral architecture at TD Securities and BMO Capital Markets, where he advised management on the implementation of systems processing billions of dollars per week. Driven by an appetite for the latest in technology, Dan created a course on business intelligence and lectured at McGill University, and has applied behavioral science to topics such as augmented and virtual reality.

Sekoul Krastev's portrait

Dr. Sekoul Krastev

Sekoul is a Co-Founder and Managing Director at The Decision Lab. He is a bestselling author of Intention - a book he wrote with Wiley on the mindful application of behavioral science in organizations. A decision scientist with a PhD in Decision Neuroscience from McGill University, Sekoul's work has been featured in peer-reviewed journals and has been presented at conferences around the world. Sekoul previously advised management on innovation and engagement strategy at The Boston Consulting Group as well as on online media strategy at Google. He has a deep interest in the applications of behavioral science to new technology and has published on these topics in places such as the Huffington Post and Strategy & Business.

Notes illustration

Eager to learn about how behavioral science can help your organization?