Home Interviewers Interview Suite Premium Suite Blog Contact

Why job applicants are like bank robbers

August 20, 2019

McArthur Wheeler

It was 1995, Pittsburgh, USA. The police officer knocked on the door. His colleagues were behind him, out of sight, with their guns unholstered. There was silence. After a while the door swung open, and McArthur Wheeler stood before them. There was no question Wheeler was the man the police wanted. They saw his face on security cameras. The officer asked him who he was. Wheeler confirmed.

‘We are investigating two bank robberies in Brighton Heights and Swissvale,’ the police officer said. The officer remembered how he got here. The eleven o’clock news yesterday ran a segment on the bank robberies, showing video footage, asking for witnesses. Someone then called the precinct and tipped them off.

Wheeler’s eyes widened; his face blanched. ‘I would like to take you to the precinct and question you,’ said the officer. Will he resist? the police officer thought, and each of his colleagues tightened his grip on the handle of his gun.

Wheeler sighed. But he didn’t resist. He was silent. They cuffed him, and placed him in the back of the car. During the drive back, Wheeler kept murmuring. ‘But I wore the juice, but I wore the juice.’

At the precinct, Wheeler was ushered into a room. The police inspector entered, puzzled about this case. He was about to become more puzzled. The inspector lit a cigarette and sat down opposite Wheeler. Smoke filled the room.

‘I do not know how you found me,’ said Wheeler. ‘I did everything right. I wore the juice.’ The inspector raised an eyebrow. The video footage of the banks clearly showed Wheeler’s face. Here was a man who made no attempt to hide his face, robbing two banks in broad daylight. Why was he surprised we found him? the inspector thought.

‘Someone told me how to fool the cameras,’ Wheeler said. ‘I didn’t believe it at first. Lemon juice. Lemon juice. It was to make my face invisible to the cameras.’ Wheeler shifted in his seat. ‘Of course, I am no fool. I didn’t believe it. I tested it. I put lemon juice on my face. It stung my eyes terribly.’ Wheeler pointed at his eyes. ‘And then I took a camera and shot a photo of myself. Do you know what the photo showed? Do you?’

Wheeler paused, waiting for a reaction. ‘What did it show?’ said the inspector. ‘It showed nothing,’ exclaimed Wheeler. ‘I wasn’t on it. I don’t know how you found me. When I was at the bank, I wore the juice.’

David Dunning

The above story sounds like fiction. But it isn’t. The Pittsburgh Post-Gazette broke it in 1995 (Fuoco, 1995). Wheeler robbed two banks, believing lemon juice would blur his face on security cameras.

Later, David Dunning read Wheeler’s story in the 1996 World Almanac (Morris, 2010). Dunning was a professor of psychology at Cornell.

Wheeler was three things:

  1. overconfident in his method,
  2. overconfident in his ability to distinguish good from bad methods, and
  3. ignorant of his overconfidence.

Dunning wondered whether Wheeler wasn’t the only one who was overconfident and ignorant. Dunning knew that the average person can be overconfident. For example, in a survey of engineers at one company, 42% thought their work ranked in the top 5% (Zenger, 1992). In a survey of college professors, 94% thought they do above-average work (Cross, 1977).

But what Dunning didn’t know was how competence influenced overconfidence. Are better performers less overconfident? Or more? So Dunning designed a series of studies and published an article in 1999. The article would become seminal.

Overconfidence and competence

Dunning and his graduate student Justin Kruger (1999) recruited university students for three studies. Study 1 asked students to rate how funny jokes were and then compared student ratings with the ratings of professional comedians. Study 2 asked students to complete 20 logical reasoning questions from the Law School Admission Test. Study 3 asked students to complete a grammar test from the National Teacher Examination guide.

In all three studies, Kruger and Dunning (1999) measured how well students did and how well they thought they did. Findings were consistent.

1. On average, students thought they were in the 61st percentile of performance.

2. The worse students performed, the more they overestimated themselves.

Distinguishing good from bad

Not only did Kruger and Dunning (1999) show that students could not correctly assess their own performance, they showed that students could not correctly asses the performance of others.

In a second part of Study 3, Kruger and Dunning (1999) asked students who were the worst and best performers on the grammar test to return. Kruger and Dunning (1999) wanted them to rate the tests of their peers. Those who did worst on the grammar test were least accurate in their assessment of their peers (r = .37). Those who did best on the grammar test were most accurate in their assessment of their peers (r = .66).

Kruger and Dunning (1999) concluded that the more grammatically students could write, the better they could recognise correct grammar in others. In abstract terms, it takes competence to recognise competence.

Further research

Kruger and Dunning’s (1999) article motivated more studies in a variety of domains with a variety of populations. We now have good evidence that overconfidence is often strongest among the worst performers. For example, Motta et al (2018) measured how well people did on a test on autism and vaccines. The worst performers were the most overconfident. Ehrlinger et al (2008) measured how much gun owners knew about gun safety. The worst performers, again, were the most overconfident.

Is there no help?

Ignorance is corrigible. One can help people to calibrate their performance estimates. Kruger and Dunning’s (1999) conducted another study. In that study, the authors asked students to complete logic tests and then to estimate how well they did.

Afterwards, they split students randomly into two groups. They distracted one group with a filler task. They trained the other group to solve logic tests. At the end, Kruger and Dunning (1999) allowed both groups to adjust their initial estimates. The trained group became more accurate. Overconfidence decreased.

Poor Alexander

Poor Alexander is a website that allows job applicants to record interview answers, play them back, read answer guides, and get feedback from senior hiring managers. We are a group of academics and industry professionals, who hired in industries such as banking, law, research, tech, or consumer goods.

When interviewing applicants, we found overconfidence to be common. The academic literature concurs, e.g. job applicants routinely rate their interview performance higher than interviewers do (Wagner, 1949). There is a McArthur Wheeler in all of us.

We found, moreover, that job applicants are missing out on job offers. Overconfidence and ignorance prevent them from tweaking what can easily be tweaked. We decided that we can help.

Why Poor Alexander is designed in the way it is

We put much thought into the design of our website. We spent months reviewing the psychological and pedagogical literature. Here are some of the principles behind our website.

You can record your answers on our website because …

… research shows that memory decays and distorts perception (Bahrick et al, 1996). We are biased to remember the positive, or look for the positive. Recordings provide objective data, and research shows recordings help.

In one study, for example, doctors rated how well they did in patient interviews (Hodges et al, 2001). They then saw recordings and could adjust their initial ratings. At the end, Hodges et al (2001) compared doctor ratings with that of their supervisors. Watching recordings made doctor ratings more accurate.

You can get feedback from senior hiring managers because …

… the ratings of employers are stricter than those of students, job applicants, or laypeople (Arvey and Campion, 1982). Employers give feedback that truly helps applicants because they need good job applicants.

You can practice giving answers because …

… seeing someone else give a good interview answer does not mean that we can. Reading a book on good interview answers does not mean we can replicate. Sometimes there is no relationship between experience and performance.

For example, Marteau et al (1990) asked doctors how many resuscitations they observed. They divided doctors into two groups — observed many vs observed few. Those who saw many were more confident they could perform resuscitations. An actual test of performance, however, showed that no group performed better. Confidence and experience did not translate into competence.

Psychologists found that generally the best preparation is doing. For example, students are better off practicing mock exam tasks than reading books (Pomerance et al, 2016). For job applicants, the best preparation is answering interview questions.

What is next

The psychological literature offers many insights to help job applicants. We will be discussing these insights in a series of articles. In this article, we focused on overconfidence. In the next, we will focus on underconfidence.


Arvey, R., and Campion, J, (1982), ‘The employment interview: a summary and review of recent research.’ Personnel Psychology. 35, 281–322.

Bahrick, H., Hall, L., and Berger, S., (1996), ‘Accuracy and distortion in memory for high school grades.’ Psychological Science . 7(5), 265–271.

Cross, P. (1977), ‘Not can but will college teaching be improved?’ New Directions for Higher Education . 17, 1–15

Ehrlinger, J., Johnson, K., Banner, M., Dunning, D., and Kruger, J. (2008), ‘Why the unskilled are unaware: further explorations of (absent) self-insight among the incompetent.’ Organisational Behaviour Human Decision Process. 105(1), 98–121.

Fuoco, M. (1996), ‘Trial and error: they had larceny in their hearts, but little in their heads.’ Pittsburgh Post-Gazette. March 2.

Hodges, B., Regher, G., and Martin, D., (2001), ‘Difficulties in recognizing one’s own incompetence: novice physicians who are unskilled and unaware of it.’ Academic Medicine. 76(10), S87-S89.

Kruger, J. and Dunning, D. (1999), ‘Unskilled and unaware of it: how difficulties in recognizing one’s own incompetence lead to inflated self-assessments.’ Journal of Personality and Social Psychology. 77(6), 1121–1134.

Marteu, T., Wynne, G., Kaye W., and Evans, T. (1990), ‘Resuscitation: experience without feedback increases confidence but not skill.’ British Medical Journal. 300, 849–50.

Morris, E. (2010), ‘The anosognosic’s dilemma: something’s wrong but you’ll never know what it is (part 1).’ New York Times. June 20.

Motta, M., Callaghan, T., and Sylvester, S. (2018), ‘Knowing less but presuming more: Dunning-Kruger effects and the endorsement of anti-vaccine policy attitudes.’ Social Science and Medicine . 211, 274–281.

Pomerance, L., Grenberg, J., and Walsh, K. (2016), Learning about learning. National Council on Teacher Quality: Washington, DC.

Wagner, R. (1949), ‘The employment interview: a critical summary.’ Personnel Psychology. 2(1), 17–46.

Zenger, T. (1992). ‘Why do employers only reward extreme performance? Examining the relationships among performance, pay, and turnover.’ Administrative Science Quarterly. 37, 198–219.

About this blog

Poor Alexander is a website that helps job applicants do well in interviews. Here we write about interviews, CVs, employment, money, and psychology. Our articles are based on the primary research we do, the literature we review, and our own experience.

Favorite posts

Why job applicants are like bank robbers

Job applicants and how to sell tomatoes

The money advice you need fits on an ...

How to win jobs and influence interviewers