In an experience-sampling study that bridged laboratory, ecological, and individual-differences approaches to mind-wandering research, 72 subjects completed an executive-control task with periodic thought probes (reported by McVay & Kane, 2009) and then carried PDAs for a week that signaled them eight times daily to report immediately whether their thoughts were off task. Subjects who reported more mind wandering during the laboratory task endorsed more mind-wandering experiences during everyday life (and were more likely to report worries as off-task thought content). We also conceptually replicated laboratory findings that mind wandering predicts task performance: Subjects rated their daily-life performance to be impaired when they reported off-task thoughts, with greatest impairment when subjects' mind wandering lacked metaconsciousness. The propensity to mind wander appears to be a stable cognitive characteristic and seems to predict performance difficulties in daily life, just as it does in the laboratory.
The study of mind wandering provides a novel means to explore fundamental issues of consciousness. For example, the commonplace experience of moving one's eyes across a page without comprehending a thing suggests the startling conclusion that we are sometimes unaware of our own conscious experience; if we "knew" our thoughts were elsewhere, we would return to reading or drop the charade (Schooler, Reichle, & Halpern, 2004). Moreover, despite controversy about the causal functions of consciousness (e.g., Morsella, 2005; Rosenthal, 2008; Wegner, 2002), field and laboratory studies of human performance (e.g., Reason, 1990; Smallwood et al., 2004) indicate that errors increase when people report experiencing task-unrelated thoughts (TUTs). Mind wandering thus co-occurs with events of scientific and practical interest. It is also beginning to figure into general theories of executive control, metacognition, and the "default mode" brain network (e.g., Bar, 2007; Buckner & Carroll, 2007; Burgess, Dumontheil, & Gilbert, 2007; Mason et al., 2007; Schooler, 2002; Smallwood & Schooler, 2006); we have argued, for example, that unwanted mind-wandering experiences represent momentary failures of goal maintenance that reflect, in part, enduring individual differences in executive control (Kane et al., 2007; McVay & Kane, 2009).
Like most areas of cognitive investigation, mind- wandering research is dominated by laboratory and neuroimaging methods; here, subjects engage in an ongoing task that is periodically interrupted for them to report or categorize their current thoughts (e.g., as on or off task; Giambra, 1995; Mason et al., 2007; Smallwood, McSpadden, & Schooler, 2007). Such thought-probe responses appear to be valid: TUT reports vary systematically with experimental manipulations, such as memory load, stimulus pacing, and task practice (e.g., Antrobus, Singer, & Greenberg, 1966; Teasdale, Proctor, Lloyd, & Baddeley, 1993); TUTs show a reliable neural signature (e.g., Mason et al., 2007); and task errors can increase by 25% during TUTs as opposed to on-task thoughts (McVay & Kane, 2009; Schooler et al., 2004). As well, individual differences in TUT rates are reliable across different primary tasks and across substantial test-retest lags (Giambra, 1995; Grodsky & Giambra, 1990/1991) and they are predicted by objective cognitive ability assessments (McVay & Kane, 2009).
Unlike some heavily investigated cognitive phenomena, however, mind wandering seems ubiquitous in everyday life. Perhaps for this reason, researchers have also investigated TUTs in ecologically valid contexts, by inserting thought probes into normal classroom activities (e.g., Cameron & Giuntoli, 1972; Geerligs, 1995) or by electronically paging ("beeping") subjects to answer questions about their thoughts, emotions, and environmental context during unconstrained daily activities (e.g., Hurlburt, 1979). …