Mind-Wandering
Mind-wandering is thinking about something other than what you are currently doing. Matthew Killingsworth and Daniel Gilbert measured it in 2010 by pinging 2,250 adults on their iPhones at random moments: minds wandered in 46.9 per cent of samples, and in at least 30 per cent of samples during every activity except sex. People were less happy when their minds wandered. The finding that carries the weight is that people who wandered somewhere pleasant were no happier than people attending to what they were doing.
What it is
The study was published in Science on 12 November 2010, runs to a single page, and is the closest thing anyone has to a measurement of what people's minds are doing and how it feels. Killingsworth, then a doctoral student, built a web application that ran on iPhones and contacted people at random moments during their waking hours through a site called Track Your Happiness. It asked three things. How do you feel, what are you doing, and are you thinking about something other than what you are doing. Killingsworth is first author; Daniel Gilbert, the better-known name and Killingsworth's supervisor at Harvard, is second. The instrument is the contribution, and it belongs to the first author.
The headline number is that minds wandered in 46.9 per cent of samples, and that the rate stayed above 30 per cent during every activity except sex. People were less happy when their minds wandered, and that held during every activity, including the ones they liked least. Comparing the two predictors directly, what people were thinking about accounted for 10.8 per cent of the within-person variance in happiness, where the activity they were engaged in accounted for 4.6 per cent. On this evidence the contents of attention matter more to how a moment feels than the moment does.
The result the popular retelling loses is the one about pleasant topics. Minds wandered to pleasant subjects more often than unpleasant ones, and people thinking about something pleasant were no happier than people attending to what they were doing. The coefficient was an uplift of -0.52 and it did not reach significance. Neutral topics came in at -7.2 and unpleasant topics at -23.9, against an overall coefficient of -8.79 for wandering as such. Read together, that says the cost was in leaving, not in what was left for. Wandering to something nice was not a gain, it was merely not a loss. A wellness version of the doctrine that reduces to "stop worrying" is not what this data shows.
The causal reading is thinner than the headline. The authors ran a time-lag analysis, and mind-wandering in one sample predicted unhappiness in the next while the reverse did not hold, which is the standard argument for direction in experience-sampling data. The coefficient there was -1.07, against a concurrent association of -9.04. The direction is indicated. It is not nailed down, and an eightfold gap between the concurrent and lagged coefficients is a large amount of the effect that the time-lag analysis does not account for.
There is also a measurement argument. Paul Seli and colleagues argued in 2018, in Consciousness and Cognition, that the 46.9 per cent figure is an artefact of forcing a yes-or-no judgment about an inner state that does not divide cleanly. Attention is graded, drifting rather than switching, and a binary probe pushes ambiguous states into whichever box the wording favours. Their critique is methodological rather than a failed replication, and this page carries no figure from it because only the abstract and published summary were consulted.