← Back to the shelf
ShareXBlueskyLinkedInEmail

Psychology paper

Revisiting the Marshmallow Test: A Conceptual Replication

Tyler W. Watts, Greg J. Duncan, Haonan Quan · 2018

Read this as

New to this. Just the gist and why it matters.

How deep

The gist, about 2 minutes.

In one line

The short of it

The famous finding that a child's ability to delay eating a marshmallow predicts lifelong success mostly melts away once you account for the child's background. Willpower mattered far less than circumstances.

Why it exists

The problem it tackles

The original 1970s marshmallow studies suggested that a preschooler's self-control strongly predicted later achievement. That story shaped parenting and policy. This paper asked whether the effect survives a larger sample and proper statistical controls.

In plain words

The key idea

Re-run the delay-of-gratification test with many more children and a more representative sample, then check whether waiting still predicts later outcomes once you control for family income, home environment, and early cognitive skills.

Think of it like: Like re-testing the claim that kids who own more books read better, and finding that once you account for how wealthy the household is, the books themselves explain much less than you thought.

What they did

The method

The authors used about 900 children from a diverse national sample and controlled for family and background factors the original studies did not.

What they found

The results

The predictive power of the marshmallow test shrank sharply, roughly by half or more, once background was controlled, and largely vanished for children whose mothers had degrees.

Does it hold up?

The honest part. What is solid, and where to be careful.

  • A conceptual replication, not an exact one; measures and ages differ from the original studies.
  • Covariate adjustment is powerful but imperfect and can be sensitive to which controls are chosen.
  • It does not prove self-control is irrelevant, only that its independent effect is much smaller than claimed.
  • The popular 'willpower predicts success' narrative is not supported once background is controlled.
  • Much of the original effect appears to reflect a child's environment and resources, not an inner trait.

So what, for you

If you are a curious newcomer

You have probably heard that kids who resist a marshmallow do better in life. The honest update: most of that was really about the kid's home and resources, not a magic willpower gene.

ShareXBlueskyLinkedInEmail