The Marshmallow test is one of those famous experiments almost everybody knows. Check this popular reenactment.

This 1970 study on delayed gratification was led by psychologist Walter Mischel, a professor at Stanford University. In the original study, a child was offered a choice between one small but immediate reward or two small rewards if they waited for some time. They also used Oreo Cookies, by the way. 

But afterwards, years later, it was discovered that this delayed gratification was a predictor for better-performing adolescents.

However, while there have been mixed replications earlier, e.g., concerning the influence of culture, a new replication examining the reliability of the test for behaviour in adult life shows very small results, stating that the test is not a reliable measure.

Jessica Sperber and colleagues conclude:

the present analysis found little evidence for the Marshmallow Test’s predictive validity across a wide set of adult outcomes. This lack of association raises doubts regarding the possibility of a causal relation between delay of gratification ability and later functioning, as the few bivariate associations that were observed were largely explained by confounding characteristics. Similarly, Ahmed et al. (2019) found that delay of gratification did not predict adolescent EF independent of other early childhood EF’s, including working memory, sustained attention, and impulsivity. This evidence suggests that improving performance on the Marshmallow Test in early childhood will not likely produce effects on adult health, wealth, or behavior, unless boosts in delay of gratification coincide with other changes to broader personal and environmental characteristics. Furthermore, the overall lack of bivariate relations also suggests that using the Marshmallow Test as a type of “screener” for later adult difficulties may be ill-advised.

Abstract of the study:

This study extends the analytic approach conducted by Watts et al. (2018) to examine the long-term predictive validity of delay of gratification. Participants (n = 702; 83% White, 46% male) completed the Marshmallow Test at 54 months (1995-1996) and survey measures at age 26 (2017-2018). Using a preregistered analysis, Marshmallow Test performance was not strongly predictive of adult achievement, health, or behavior. Although modest bivariate associations were detected with educational attainment (r = .17) and body mass index (r = -.17), almost all regression-adjusted coefficients were nonsignificant. No clear pattern of moderation was detected between delay of gratification and either socioeconomic status or sex. Results indicate that Marshmallow Test performance does not reliably predict adult outcomes. The predictive and construct validity of the ability to delay of gratification are discussed.