A multi-faceted approach to understanding individual differences in mind-wandering

How consistent is mind wandering across situations and tasks?—A latent state–trait analysis

“…Furthermore, in line with previous studies using a single-latent-factor approach, the mind-wandering trait factor extracted in our study was negatively correlated with WMC and reasoning (McVay & Kane, 2009;Mrazek et al, 2012;Robison et al, 2020) as well as with selfdiscipline (Ralph et al, 2017). Thus, there was some evidence for construct validity of the trait variable that correlated with other constructs within the nomological network in a meaningful manner.…”

“…TUT rates observed during ongoing two-back and reading tasks were negatively correlated with the performance in the respective tasks. This is a typical finding in the literature (e.g., Robison et al, 2020;Rummel & Boywitt, 2014), which renders support to the validity of TUT reports as it shows that TUT rates correspond with a behavioral marker in a meaningful manner. The LST analyses showed that the latent TUT-rate variance components shared between the two indicator tasks (i.e., the remaining variance after CONSISTENCY OF MIND WANDERING 21 error variance and task-specific variance components are controlled for) were perfectly consistent across measurement occasions since the occasion-specific variance components were zero.…”

“…Nakamura & Csikszentmihalyi, 2002). Additionally, we assessed WMC at both measurement occasions with two complex span tasks and reasoning abilities with a matrices and a number-series task at one occasion to replicate previously observed correlations of these constructs with mind wandering (McVay & Kane, 2009;Mrazek et al, 2012;Robison et al, 2020). In doing so, we intended to test the latent trait-mind-wandering factor's construct validity but also to achieve trait-consistency benchmarks from the WMC measures as they are known to reflect trait variance to large degrees (Ilkowska & Engle, 2010).…”

“…In doing so, researchers increase reliability but may also aggregate across systematic (time-on-task) effects (but see Schubert et al, 2019;Steindorf & Rummel, 2020, for an alternative multilevel approach). Some researchers go one step further and assess mean TUT rates during different tasks and then use a latent-factor approach to extract a single common mind-wandering factor McVay & Kane, 2009;Meier, 2019;Mrazek et al, 2012;Robison et al, 2020;Robison & Unsworth, 2015;Unsworth et al, 2012). The idea behind this approach is to control for measurement-error-induced variance components within the manifest TUT rates, thereby achieving an indicator of mind-wandering that is more reliable than the manifest scores and also less task-specific.…”

“…Because mind wandering has been shown to be task-specific to some extent-for example mind wandering assessed during cognitively demanding tasks has been shown to be negatively correlated with WMC, whereas mind wandering during less demanding tasks has been shown to be uncorrelated with WMC (Robison et al, 2020;Rummel & Boywitt, 2014)-we found it necessary to differentiate between task-general and task-specific mind-wandering in addition to investigating how consistent they are across different measurement occasions.…”