Unskilled and Unaware of It: The Competence-Confidence Kill Zone

Jul 25, 2024

Article voiceover

1×

0:00

-5:57

First of all, welcome and thank you to all the new subscribers. I feel a lot of pressure to make this a great blog! Please do share the blog and comment on your experiences as a learner, expert, coach or mentor. This should be a joint & ongoing learning experience through the exchange of ideas. I am not the expert, just a curious People Geek who will act as a tour guide of the evidence and theories on why people behave in certain ways.

I have a particular interest in expertise. How we measure it, how we identify it, how we develop it, but also its unintended consequences.

The Dunning-Kruger effect is one of the most highly replicable findings in social Psychology (Mazor & Fleming, 2021), albeit with some counter arguments (Gignac & Zajenkowski, 2020). In their original paper, titled “Unskilled and Unaware of It”, Justin Kruger and David Dunning (1999) through a series of experiments showed how novices greatly over-estimate their ability. Or another way of saying it, their confidence curve grows at a greater rate beyond their competence curve. This curve became so de-coupled that in their performances across tests of humour, grammar and logic participants placed in the 12^th% of performance estimated themselves to be in the 62^nd%.

What causes this systematic over confidence in novices and why as leaders and HR practitioners should we care? The cause is down to metacognition, the reason we should care is because of the kill zone (dramatic but true). Let’s expand.

Deficient metacognitive skills mean that novices are unable to decipher good performance from poor performance as the performance environment is too noisy. They cannot separate luck, randomness and competent execution. From an ecological psychology perspective we might say they are less attuned to “specifying information” (Michaels & Palatinus, 2014), the information in the environment in amongst all the noise that distinguishes what higher performers attune to and use to make decisions and take actions, compared to lower performers. This inability to focus on specifying information and the focusing on all the noise is what might be experienced as cognitive overload.

Interestingly, cognitive invariants like varying personality traits (such as openess) have no effect. (Chen & Wall, 2022)

The most effective way of reducing this confidence-competence gap, was by a re-education of attention away from un-specifying noise, to the specifying information that allowed them to detect how successful their actions were.

The Kill Zone

The reason this matters was captured so well in a recent podcast with Adam Grant I snipd it below. The most dangerous time is not at the super early novice level. These individuals are so new to things, they have not experienced barely any success so are far more open to feedback from outside of the environment. The danger is that sweet spot zone of experiencing some continued success with competence, but not yet where they can distinguish between luck and the edge of their competence- the beginners bubble. Think teenage drivers or more concerning- pilots at the stage of 600-800 (Craig, 2001; Knecht, 2013) hours or surgeons 15-20 surgeries in. This is where the subjective learning curve has moved far beyond the actual learning curve. Have a listen to David Dunning explaining this:

Conversely, research in the Financial Services Sector show how more experienced fund managers were less responsive to feedback on their funds poor performance compared to less experienced managers (Gaba et al., 2023). Why might we see an increase in over confidence as experience grows in this study? Well the differences here may be explained by the limitation of this study correlating experience as measured by years of service for increased competence. Sackett et al. (2023) showed how tenure only had a .07 correlation with performance. This can also be showing the weakness of external feedback over the strength of specifying information. Feedback is knowledge about performance vs specifying information is knowledge of performance. This shows the results may be less of an overconfidence, but more a weakness in the effectiveness of feedback that is not directly from the environment. What this study did not measure is, that with good coaching, the re-education of attention to more specifying information would this improve the adaptation of their fund strategy?

Limitations

Much of this research is done in laboratory environments and some of the original work was quite subjective e.g. the funniness of a joke. Knecht (2013) also caveat due to the sample size of pilot data the results are “tentative” in support of the kill zone. Thank you to Adam Grant for providing the reference David Dunning refers to in the clip above.

So what are the takeaways?

Scaffolded learning, whether that is through coaching, mentoring or use of feedback should be about educating attention to the specifying information. Specifying Information is the information in the environment that the learner should be attuning their attention to in amongst all the noise. This goes beyond just telling them in verbal constructs what to look out for, but actually creating practice environments that replicate the specifying information via role plays, theatrical learning or safe practice environments. Knowledge about the environment (arm chair expertise) is not the same as knowledge of the specifying information in the environment.

In roles where the cost of failure is high, coaching and mentoring should go on beyond the initial phase of demonstrating competence. There is good evidence they will get worse before they get better. Learning is not a nice linear curve, it is just like the financial markets, it can crash over a short term. Look out for those early signs of over confidence.

Simply put, expertise is not linear and very weakly correlated with tenure.

People Geek's Substack

Discussion about this post