Matthew Killingsworth

Who they are

Killingsworth is a senior fellow at the Wharton School of the University of Pennsylvania and the founder and director of trackyourhappiness.org, which he built while completing a PhD in psychology at Harvard. His method has stayed the same across his career: rather than asking people to remember how happy they were, catch them at random and ask them now. Experience sampling is older than his application, but the smartphone made it possible at a scale the technique had never previously reached.

The 2010 result is the one this site leans on. "A Wandering Mind Is an Unhappy Mind" appeared in Science on 12 November 2010 as a Brevia, a two-page format with most of the substance in the supporting online material, which is where anyone citing it should be reading. The analysed sample was 2,250 adults, 58.8 per cent male, 73.9 per cent resident in the United States, mean age 34, drawn from a database of nearly a quarter of a million samples contributed by roughly 5,000 people in 83 countries. Minds wandered in 46.9 per cent of samples, and in at least 30 per cent of samples during every activity except sex. What people were thinking about predicted their happiness better than what they were doing, accounting for 10.8 per cent of the within-person variance in happiness against 4.6 per cent for the activity itself.

The causal reading is thinner than the headline. A time-lag analysis showed mind wandering in one sample predicting unhappiness in the next while the reverse did not hold, but the coefficient in that analysis is far smaller than the concurrent association, at -1.07 against -9.04. The direction is indicated rather than settled, and Killingsworth and Gilbert say so in the paper. His later work has used the same sampling method on a different question, income and well-being, including a 2023 adversarial collaboration with Daniel Kahneman and Barbara Mellers that reconciled two datasets which had appeared to contradict each other.

Key papers

Killingsworth has not published a book. His influence runs through journal papers and through the sampling platform itself, which has supplied data to work well beyond his own.

Killingsworth, M. A., & Gilbert, D. T. (2010), "A Wandering Mind Is an Unhappy Mind," Science, 330(6006), 932. The mind-wandering study. A Brevia, so read the supporting online material alongside it.

Killingsworth, M. A. (2021), "Experienced well-being rises with income, even above $75,000 per year," Proceedings of the National Academy of Sciences, 118(4), e2016976118. Experience sampling applied to the income question, finding no plateau where earlier work had reported one.

Killingsworth, M. A., Kahneman, D., & Mellers, B. (2023), "Income and emotional well-being: A conflict resolved," Proceedings of the National Academy of Sciences, 120(10), e2208661120. An adversarial collaboration between researchers who had reached opposing conclusions, reanalysing both datasets together.

Key ideas

Real-time sampling beats recall. Asking people how they feel right now, at random moments, avoids the reconstruction errors that make retrospective happiness reports unreliable.

Minds wander in roughly 47 per cent of sampled moments, and people are less happy when they do, across every activity measured including the least enjoyable ones.

What you are thinking about predicts your happiness better than what you are doing.

Wandering to a pleasant topic did not make people happier than staying with the task, which narrows what the finding can be used to argue.

The causal claim rests on a time-lag analysis whose coefficient is much smaller than the concurrent association, so the direction is indicated rather than established.

Concepts on this site

Mind-Wandering

Tanha

Papanca

Find them elsewhere

LinkedIn is the only social account linked from his own site. No X, Bluesky or Google Scholar profile is linked from his site, happiness-science.org, trackyourhappiness.org or his Wharton page, so none was recorded

Checked 2026-09-10. Handles and affiliations go out of date, so treat this block as a snapshot rather than a standing fact.

---