One drawback of this work is that it has tested only college students, although they are obviously not alone in multitasking. As technology prices drop, younger children have increased access to mp3 players, smart phones, gaming platforms, etc., and with increased access comes increased use. Is the relationship between multitasking habits and cognition different for younger children?
In a recent study, Matthew Cain and colleagues (in press) tested 73 subjects (average age = 14.4 years) on a battery of cognitive and personality tests. They largely replicated previous work, and added some interesting extensions.
First, they found that frequency of media multitasking was negatively associated with working memory span measures. They did not find a relationship with the ability to filter information in working memory, which contrasts with some previous work (Ophir et al, 2009). Cain et al speculated that Ophir et al may have used a filtering task that was also demanding of span.
Second, Cain et al wanted to use a measure that had more real-world validity, so they correlated media multitasking scores with student scores on the state tests for math in English (for Massachusetts, where the work was conducted). A significant negative correlation was observed. Scores were not related to standard lab measures of math and English ability, however, leading the researchers to suggest that scores on the state test might have been affected because test administration is longer, so scores would more likely be affected if students were prone to distraction. In other words, students might have been equally proficient in math and English, but less able to maintain attention for a long duration.
Third, researchers wanted to show that media multitasking was not related to cognition across the board, so they administered some tests of cognitive ability they thought were unlikely to be related, and indeed found that manual dexterity, and the unconscious learning of a complex category, were not related to media multitasking.
Fourth, researchers tested whether media multitasking was related to personality characteristics and beliefs that are known to be related to academic achievement. They observed no relationship with grit or conscientiousness, but did see a robust negative relationship with growth mindset. The authors expected to see relationships of all three with media multitasking. Why is not clear to me. If you think that most teens know that multitasking while reading or studying is not conducive to good performance, you might predict a negative correlation with grit and conscientiousness, but it’s not clear that most teens are aware of the relationship. The observed relationship with growth mindset—a belief about the nature of intelligence—puzzles me, and the authors say little about it.
Perhaps the most important aspect of these results is the age of the children tested. The results show that the relationship between media multitasking and cognition that we see in college students is observable by middle adolescence. That marginally increases the probability that the direction of causality is cognition prompting media multitasking behavior, not multitasking prompting changes in cognition. The authors are appropriately cautious in discussing causality but the results are a bit easier to account for with that explanation, the idea being that younger children have had less multitasking experience, and so there has been less opportunity for multitasking to influence cognition—yet the relationship is present.
One point about this study does make me uneasy. The median self-reported estimate of media use was 18 hours/day. A good number of subjects offered weekly estimates of media use that amounted to more than 24 hours per day. The authors note that these seem like overestimates, and so they suggest that these raw figures are not really interpretable.
But they go on to take the subject reports of frequency of media multitasking at face value. For each medium, subjects were asked how often they engaged with that medium while doing something else, choosing among four categories (Never, a little of the time, some of the time, most of the time). Why trust these estimates, if their estimates of the number of hours spent with media are not trustworthy? Both are retrospective memory reports, and the judgment of media multitasking has not been validated, as far as I can tell.
That validation ought to be high on the priority list for researchers in this area. It would be a terrible shame if all of these effects that we are tying to media multitasking actually ought to be tied to accuracy and bias in memory reports.