The overconfidence effect is a natural bias toward believing that we’re better at something than we actually are. The overconfidence effect can distort belief in the accuracy of a strong memory, estimations of how long it will take to get things done, judgment about our intelligence compared to others, and even the reliability of eyewitness accounts. It can sabotage communication during conflict, too.
I’m leading a mediation training and the participants are in a mediation roleplay. In one trio, I notice a man escalating the conflict. I walk over to watch what’s unfolding.
The man is repeatedly interrupting the woman with whom he’s in conflict. Each time, the mediator scowls at the man. With each scowl from the mediator, the man seems to become bolder, increasing both the volume of his voice and the length of his interruption.
I call a time out. I ask the mediator what he’s noticing.
“His behavior,” he replies, gesturing toward the interrupting man, “is getting to be a real problem.”
I would agree with you, I say to the mediator. What have you tried so far to address it?
He replies that he’s been trying to follow my lowest level of intervention first rule of thumb. I ask what his low-level intervention has been and if he thinks it’s working.
“Each time he’s interrupted, I’ve scowled at him to discourage the behavior. I don’t think it’s working, though. I may need to go up a level.”
I ask if he thinks the interrupting man understood what he was conveying. “No question,” replies the mediator, “We made eye contact each time.”
I ask the interrupting man if he’s noticed the scowls.
“I sure have,” he replies. Then he gestures toward the woman. “But I thought the mediator was scowling because of the obnoxious things she was saying.”
A quick exercise
In the following very brief audio, I’m tapping out the beginning of a famous song on my husband’s djembe. What song is it?
Actor, director, and author Alan Alda does a similar exercise with the audience when he’s speaking about science and communication. A conflict resolution colleague wrote to tell me about seeing Alda conduct the casual experiment:
He took a volunteer from the audience and asked her to select one of the three songs he had listed on his phone. He directed her, “Once you have the song in your head, please tap it out.” (Picture someone thumping on a podium to the melody in their head.)
She went on for about a minute with what sounded to me a little like Morse code. Alan asked her, how confident are you that the audience will recognize this song? She offered 70%. So in her mind, 70% of the audience would correctly guess the song she tapped out. He asked the audience, “How many of you think you know this song?” About 50% raised their hands. He asked a few of them and each one was wrong.
— Lora Barrett
Lora went on to say that Alda typically sees signal senders with similar or stronger confidence that the audience will recognize the song. And many in the audience typically believe they know the song. But their confidence turns out to be disproportionate to reality.
Alda’s exercise does a really nice job of illuminating the overconfidence effect, the phenomenon in which our confidence in our own ability is quite a bit greater than our actual performance.
Check out the signal
Overconfidence in our ability to send a clear signal can create a communication gap we never intend. Overconfidence in our ability to understand their signal can widen that gap further.
Sometimes we think we’re so abundantly clear that no one could possibly miss our point. My husband, a very articulate college professor, once received a course evaluation in which a student took him to task for never explaining “what the hell an S.A. exam is.” He and I stared at the evaluation for a long while before it hit us: Essay exam. My husband had told the class that most tests during the term would be essay exams.
Sometimes we think that someone knows us so well they should know exactly what we mean. This is a particularly insidious assumption in relationships with spouses and partners, parents and children, and a source of pain we could avoid. Our loved ones are not mind readers and we create pain for ourselves and for them when we expect them to be.
Sometimes we think we know them so well that we know exactly what they’re saying, even before they’ve finished their sentence. There are all sorts of cognitive traps that trick us into certainty about the signal we’re receiving, from confirmation bias to reflexive loops and more.
What if, instead of running with what we heard, we checked it out first? What if, instead of forging ahead assuming they “get us,” we made sure they really did?
Oh, and that song I was tapping out on the djembe? It was Somewhere Over the Rainbow. Here’s Iz Kamakawiwoʻole with my favorite version:
