If Only “They” Knew
Do you ever wonder why so much software in the world seems to be overly complex, incomprehensible, or just plain messy?
And then, say, you contact “the owner” to address the situation and they see nothing wrong?
Doesn’t that infuriate you?
If only “they” knew how terrible their code was!
You might be interested in hearing about the Dunning-Kruger Effect.
In the late nineties, David Dunning and Justin Kruger, psychology professors at Cornell University decided to test several hypotheses around the behavior of incompetent people by acquiring a pool of unsuspecting undergraduates under the guise of free pizza.
Behind me you’ll see a few simple graphs of the data as I talk through the details and findings.
The test procedure was pretty simple and went something like this:
- before the test, each student would guess how well they’d do on the test (we’ll call this the “perceived test score”)
- after the test, each student would estimate their test score (we’ll call this the “estimated test score”)
The perceived and estimated test scores were very similar - no statistical significance between the perception of students’ performance before taking the test versus their estimates after taking the test (most students estimated they would get around a 60-70% grade).
The real world of test scores would, of course, look something like this, a straight linear line somewhere from the worst performance from the novice all the way to the best score from the highest ranked expert.
So what does the Dunning-Kruger study begin to show us?
The radical difference between the novices’ perceived and estimated test scores versus their actual scores is shocking, is it not? Let’s summarize this finding as “novices overestimate themselves.”
Doesn’t this start to explain a lot in the world? People who aren’t that great at something think they’re awesome and there’s really no way to convince them that they might… not… be as stupendous as they think? Or, if you’ve seen American Idol at any point during your life… you know exactly what this looks like.
So again - “novices overestimate themselves.”
[ What’s even more demoralizing is that the Dunning Kruger study hints that novices may even be likely to think that they actually performed better than their original estimate even after their’ skills are tested! ]
However, there’s another interesting nugget to pull from this data - the fact that the experts actually performed better than they expected! Let’s summarize this finding as “experts underestimate themselves.” What’s more, experts had a tendency to expect a lower score after taking the test compared to their original estimate.
Crazy, huh? Experts underestimate themselves. And after the test, the experts even thought that their original estimated performance numbers were too high!
So at this point, you might be raring’ to go and take these findings and the Dunning-Kruger Effect back to your workplace and explain to everyone that they’re a novice but they just don’t know it. Unfortunately… the Dunning-Kruger Effect, and it’s range of findings also applies to you. And unfortunately… me too. The Dunning-Kruger Effect applies to everyone - especially when you don’t realize it.
I’m somewhat of a psychology junkie. I read books on topics like “Bowen Family Systems Theory” and “Nonviolent Communication.” And - I think a good foundation of psychology is important for anyone to have - but especially in software engineering where you have people, processes, and computers all “working” (for lack of a better word) together.
So one of my goals this year was to relay some of my thoughts and findings from the psychology domain into the software engineering domain. I submitted bunches of proposals to a wide range of conferences… big… small… and everything in between. I always got a “Hey that sounds like a good talk… but… we don’t have room for it” response. That sucked. But I kept submitting. I even submitted a proposal to talk about this very topic to the DevOpsDays Chicago crew. And… was denied.
“No one realizes how much they’re missing by turning me down!” I proclaimed in my head (cause heaven forbid I actually proclaim that out loud as a stable introvert who knows all things psychology).
No one realizes how much they’re missing by turning me down!
However, two weeks before DevOpsDays Chicago, I received an email from the organizing committee that asked if I could massage my talk into an Ignite Talk. I waited to respond, to give the impression that I wasn’t just crouching at my keyboard waiting for such a request. Then… I got to work over the next two weeks putting the ideas of The Dunning Kruger Effect into a nicely packed presentation.
Keep in mind, I have never done an Ignite Talk before in my life. I’ve done lots of presentations about technologies, processes, and held training sessions… I’ve even read books and taken classes on giving better presentations! But never an Ignite Talk.
I’ve never done an Ignite Talk. But they’re easy. People do them all the time.
I began to realize that Ignite Talks - with their condensed form backed by the auto-advance slide slave driver - were really. dang. hard.
I have so many ideas and thoughts about this subject - The Dunning-Kruger Effect - and psychology in general - how do I even begin to come up with even one idea that I can relay even close to properly!
My first practice presentation went to my coworkers - bless their souls. I think they left the room more confused than when they entered. But they provided feedback. It’s “valuable” feedback, I kept telling myself.
Inside of my head though, I was screaming:
WHAT DID I GET MYSELF INTO?!?! WHY DID I EVER THINK I COULD PRESENT ON THIS SUBJECT? SHOULD I JUST CONTACT THE CONFERENCE ORGANIZERS AND TELL THEM “SOMETHING HAS COME UP AND I WON’T BE ABLE TO PRESENT… SORRY!”?
I entered a dark emotional time of my life. I questioned my presentation track record altogether - “Have I even give a single good presentation in my life?”
Nah - I thought - I played Romeo in Romeo and Juliet in Elementary School - I was pretty good…
I was probably terrible at that too.
The second, third, and however many versions of my presentation I delivered… never roused the energy and enthusiasm from the audience that I was expecting (or rather, hoping for).
After some time, I realized… shooot… maybe I’m embodying the Dunning-Kruger Effect right now and this is what it actually… truly… feels like to get unstuck as an unaware novice.
I thought I was good at presentations, thought I could succinctly deliver topics of psychology that I had read about. But those assumptions of my skills and abilities went, for the most part, untested.
I realized I was embodying the results of the Dunning-Kruger Effect.
I was free to believe that I was awesome! And it felt great. Until I actually had to produce something - to test my skills - to actually come to grips with how I’m good at some things (like battling out ideas on Twitter), but that doesn’t make me good at everything.
So what should knowing the Dunning-Kruger Effect mean to you? And me? It’s not a study to be used to tell people how they’re incompetent and fail to realize it - it shouldn’t - I say again: SHOULD NOT be used in presentations or conversations to back up your ideas on what “they” should do or what “they” should be. It should not be used in presentations to get points across like:
- “Why we should be using Docker because you’d be stupid to do anything else” or
- “Why our current practices suck and if we’d just do it Etsy’s way we’d be better off” or even
- “How and exactly why females are ruining the software industry.”
The Dunning-Kruger Effect isn’t just more ammunition to make your points
The Dunning-Kruger Effect, and my personal story of feeling what it’s like to live through the effect should drive us to the realization that a lot of times we’re more interested in the continuous improvement of systems and processes, but less so about our own personal character.
So with that - take the Dunning-Kruger Effect’s ideas and implement a small continuous improvement plan for yourself. And actually do it. If you hear yourself saying: “I could do that better” or “I can’t believe they did it that way.” Test yourself. Test your assumed awesome skills.
Don’t give up. For the betterment of your character. For the transformation of your company. For the good of this community.
Test your assumed awesome skills
If you apply the concepts of continuous improvement - the same ones you’re applying to systems around you - to yourself… think of the power that it would have to replace the current story of our “keep those geeks in the basement and away from people” industry into a story of an industry of passion, growth, and deep human relationship.
Trust me - it’ll rattle you. It’ll mess you up. But then again… “bringing the pain forward” to continuously improve, deliver, and learn always is harder than you first expected…