Is a Little Knowledge Really a Dangerous Thing?

New research explores the dangers of overconfidence and the beginner's bubble.

Posted Nov 13, 2017

A little learning is a dangerous thing / Drink deep, or taste not the Pierian spring / There shallow draughts intoxicate the brain / And drinking largely sobers us again. —Alexander Pope

How dangerous is it to be overconfident?   

While we may be impressed by people willing to make bold statements (especially at election time), the fact is that, more often than not, being overconfident can be a grave mistake. It's no wonder then that Nobel laureate Daniel Kahneman stated in his 2011 book, Thinking Fast or Slow, that, if he had a magic wand to banish one judgmental bias from the world, he'd choose overconfidence.  

Ironically enough, however, the people who are most likely to be overconfident are the beginners who have just received a taste of whatever skill or topic they are trying to learn. According to the popular "four stages of competence" model developed in the 1970s by Noel Burch and Thomas Gordon, the first and, arguably, the most dangerous stage is what they termed unconscious incompetence. In other words, when someone is in the early stages of learning a new skill, he or she may not realize how much more there is to learn.  

Looking at popular culture, one classic example of an unconscious incompetent is the Sorcerer's Apprentice who, deciding to save himself some work, uses his master's magic book to do the work for him and, of course, failing miserably. Another popular meme that highlights the inability of many people to recognize their own incompetence is the famous Dunning-Kruger effect, which evolved out of a classic 1999 study by social psychologists David Dunning and Justin Kruger.  While it tends to be applied largely to mediocre people who overestimate their ability, the effect can be applied to any beginner in a new field who may not yet realize more training is needed.

A new research article published in The Journal of Personality and Social Psychology presents the results of six studies testing overconfidence in beginners and what it means for their performance. The authors, Carmen Sanchez of Cornell University and David Dunning of the University of Michigan, provided impressive evidence for what they called the "beginner's bubble hypothesis." According to this hypothesis, people who start out learning a task or skill usually begin with a feeling of caution or uncertainty until, following a few early successes, the "beginner's bubble" sets in. At this point, what initially seemed difficult appears much easier than originally feared. This is when the overconfidence kicks in and beginners become unconscious incompetents. Of course, what usually follows is a "correction period" when the overconfidence flattens out and skill continues to improve.    

According to Sanchez and Dunning, this kind of early overconfidence is most likely to occur in dealing with what they called probabilistic learning tasks. This involves trying to make predictions about uncertain events based on whatever clues people may rely on to guess what will happen in the future. Whether the task involves guessing which direction the stock market is going to take in the next six months, who is going to win the next election, which job applicant should be hired, etc., there is always going to be some uncertainty and, not surprisingly, the damage that can result from overconfidence in making such predictions can sometimes be disastrous.  

To test the beginner's bubble hypothesis, Sanchez and Dunning designed four laboratory studies and two real-world studies looking specifically at how people developed expertise over time and what it meant in terms of the kind of judgments they made. The four laboratory studies examined how people made predictions about uncertain events (using scenarios ranging from diagnosing someone during a zombie apocalypse to how accurate a lie detector was).   

As predicted, all four studies showed strong evidence for a beginner's bubble and also allowed them to test a possible explanation for why a beginner's bubble would form in the first place: a process they termed exuberant theorizing. This suggests that people often develop their own theories on how to interpret information they receive and, based on a few early successes, become much more confident in the validity of those theories than the actual facts might warrant. Even though these theories were wrong more often than they were right, people continued to believe in them until additional experience taught them differently. 

Recognizing the limits of this kind of laboratory research, Sanchez and Dunning decided to test the existence of a beginner's bubble with a skill people might require in the real world, i.e., financial competence as measured using the 2012 and 2015 Financial Industry Regulatory Authority (FINRA) survey on financial capability. This annual survey examines financial habits of people across the United States. Along with basic questions about demographic factors (age, gender, education, etc.), the survey tested respondents in terms of financial literacy (using items such as “If interest rates rise, what will typically happen to bond prices?") as well as asking them to estimate their perceived financial literacy on a seven-point scale.

Looking at the 24,814 Americans surveyed in 2012 and 25,901 in 2015, the researchers compared how respondents viewed their perceived financial literacy (how financially astute they believed they were) to their actual financial literacy (how well they did on the financial questions asked in the survey. As expected, participants in the 18 to 24 age group showed a sharp surge in how they viewed their financial expertise only to flatten out or decline until middle age (45 to 54 age range). 

Even when actual financial knowledge, income, and level of education were factored in, the evidence for a "beginner's bubble" remained strong. When looking at financial literacy, however, age and experience appear to win out. In other words, actual financial knowledge rose as people grew older and more experienced regardless of other factors such as gender, level of education, and income.

So what can we conclude from this research? As Sanchez and Dunning point out in discussing their results, people in their third decade of life often lack the experience to make proper financial decisions; unfortunately, they are also least likely to realize this. If anything, they are often prone to overconfidence. While they may be hesitant at first, this passes quickly as the beginner's bubble sets in and the consequences can be, well, disastrous. Sadly, even experts can be prone to these kinds of errors because of their own exuberant theorizing when it comes to making predictions about financial markets, political events, election results, etc.   

While more research is definitely needed, these research results highlight the dangers of overconfidence, especially for people just starting out who may have an inflated view of their own ability. Though time and experience can help to some extent, people also need to realize that we are always going to be vulnerable to exuberant theorizing no matter how experienced we believe ourselves to be.   

So think long and hard about making that risky decision you've been contemplating, whether financial or otherwise. Chances are, you might need a little more experience first.


Sanchez, C., & Dunning, D. (2017, November 2). Overconfidence Among Beginners: Is a Little Learning a Dangerous Thing?. Journal of Personality and Social Psychology. Advance online publication.