Part 4: How We Think Differently
The Crucial Lesson of the Beady Eye
In the early 90s, I’d become enraptured about identifying hidden emotions from “unconscious” hand movements (or subconscious or preconscious …) We were using this complicated hand-movement exercise with “cutting-edge technology” (at that time, the mouse). But, the rich and fascinating results can’t be replicated when the tests are repeated. What now?
I was working with another guy on a parallel quest, who like me was drawn moth-like to the flame of this hand-movement technology. We were commiserating about it not working. I proposed some complicated fixes to the technology. He listened to me patiently. “You might be over-thinking this, Steve. Not that I blame you – an over-educated guy from Harvard working with another over-educated guy from Stanford. But I’m from the Midwest, and we usually try to keep it simple. We don’t over-think as much as you guys from the coasts.” Then he told me about the Beady Eye.
A little earlier the federal government noticed how many Americans are depressed, and the toll that depression takes on their job productivity. The National Institutes of Health were tasked to come up with the gold standard assessment for depression. They convened a panel of eminent psychometricians and collected all the best assessments out there. From these they distilled an instrument with 120-odd questions which identified depressed people with 95% accuracy.
But 120 questions is way too many for everyday use by psychologists and primary caregivers. The panel chose about 50 questions, which still were 90% accurate at identifying depressed people.
Still too long for everyday use! After long debate and analysis, the panel shortened the battery to 21 questions. Much more practical and still 80% accurate. It’s become the updated Beck Depression Index, or “BDI” (say it out loud) to practitioners.
Was their work done? Not quite. One of the eminent panelists (from the Midwest like my colleague?) pointed out that 1 of the 21 questions by itself had almost 75% accuracy in identifying who’s depressed. That penetrating question was “Are you depressed?”
Hmmmmm. Too simple. Way way way too easy. Raises questions for another day or another blog, like “Why have so many psychologists with so many assessment instruments for so many syndromes and disorders?”
This story was a revelation for me! Stop “tricking” people with distracting exercises, hoping they’d revealed their inner feelings. Ask them directly and find the ways to let them answer more accurately and completely. Remarkably, the absolute best question is “How do you feel?” For us, the hard part was creating instrument designs that let almost everyone answer that simple question easily and fully.