1/ Does student smartphone use in class harm learning? Perhaps not. New paper w/ @andbjn, @RandomSquawker & @daviddlassen. We continuously monitored 470 university students’ device use & academic performance over 2 yrs... http://doi.org/10.1177/0956797620956613 #AcademicTwitter #psychtwitter

2/ Brief experimental studies have largely found negative effects of in-class mobile device use on learning. However, studies outside of the lab are less convincing: largely rely on self-reported device use & cross-sectional analyses w/ only 1 row of measurements per student!

3/ Self-reports of screen use lack accuracy & validity (see research by @jesstpiotrowski @dscheykopp @claesdevreese @theoaraujo @dougaparry @BritDavidson @cjsewall9 @jake_fisher @hnmiecz @dsquintana @davidaellis, @DrSallyAndrews, @mscharkow). As @ShuhBillSkee wrote:

4/ “studying self-reported behaviour ≠ studying behaviour”

https://twitter.com/ShuhBillSkee/status/1310923044626599937?s=20

By contrast, our study employed an app-based tracker to automatically log how much time students use their phones in multiple courses over several terms...

5/ We also leveraged geolocation data to measure screen use that transpired when students were physically present in class with their peers...

6/ Consistent with other research and to absolutely nobody’s surprise, we found that students used their mobile devices frequently in class. But was this use negatively associated with how well they did?

7/ If we ignore the panel structure of our data and model the relationship between smartphone use cross-sectionally with an extensive set of common controls — similar to the majority of previous research — we find that device use is negatively associated with performance, BUT…

8/ if we make use of the data’s panel structure to control for all stable unobserved student & course-specific characteristics, the estimated negative effect of in-class smartphone use on academic performance is reduced by almost two-thirds & the confidence interval contains zero

9/ Our study strongly suggests that prior cross-sectional research likely overestimates the negative effect of smartphone use on learning considerably.

10/ Our findings directly challenge recent “good news” that smartphone communication findings may be underestimated & that logged measures will reveal larger effects than self-reports. Correlations probably aren't showing what some think @icahdq https://twitter.com/icahdq/status/1306986527919136769?s=20

11/ Even with a rich set of “known” controls and logged smartphone use measures, it is very likely that researchers using correlational designs are omitting other confounding individual and contextual variables that influence both device use and performance.

12/ Our study is consistent with recent work by @OrbenAmy, @ShuhBillSkee, @CaveroRedondo, @davidaellis & @MasurPhil indicating that existing evidence about the adverse effects of screen use may be highly dependent on the context of use and the cross-sectional approaches used

13/ Prior work suggests that in-class smartphone use is largely composed of non-academic/social media use. A recent meta-analysis on social media use and its association with academic performance, found only minor to negligible associations @AppelMarkus, in-line with our findings

14/ We strongly encourage researchers investigating the social and psychological impacts of mobile devices to employ high-res and unobtrusive measures of device use, over extended periods of time and across multiple courses & terms.

15/ Researchers should investigate other psychological traits (ex. self-control), and contextual factors (ex. course or classroom characteristics) that may jointly influence both device use and learning outcomes.

16/ Our uncertainty about not-well-understood psychological, social and environmental mechanisms that may jointly influence smartphone use
and academic performance should be very high.

17/ Critically, our study also contributes to growing evidence that the impact of #screentime on #Wellbeing ( @Orben @NiklasJohannes), #Mentalhealth ( @H_Shawberry) and #learning may actually be quite minor, at least for the younger age group predominantly investigated.

At this point in time, educators, researchers, policy makers and popular communicators should exercise caution and avoid making causal interpretations on the basis of existing cross-sectional observational studies.

#highereducation #Screentime #education #SocialDilemma