The marshmallow test measured something other than willpower
A famous experiment about patience got a large, careful do-over in 2018 — and patience wasn't the variable doing the work.
In the 1970s and '80s, Walter Mischel sat preschoolers alone with a marshmallow: eat it now, or wait and get two. Years later, the waiters were doing better in school and life, and the lesson everyone drew was tidy — teach kids self-control, and outcomes follow.
In 2018, the psychologist Tyler Watts and colleagues redid the study with nearly a thousand children, this time a group that looked like the country rather than like the faculty children in Mischel's original Stanford sample. The wait-and-achievement link was about half the size reported originally, and it mostly disappeared once family income and a mother's education were factored in. Most of the predictive power, oddly, came from crossing a bare twenty-second threshold — waiting five minutes conferred almost nothing extra over waiting twenty seconds.
What the marshmallow test was tracking, it turns out, was less a child's grit than her circumstances: a stable home teaches you that waiting pays off, because it usually has. Poverty teaches the opposite lesson, correctly. The moral wasn't “build more willpower.” It was “fix the conditions that make patience a reasonable bet.”