Personally, I'm not convinced it is a good idea to make any conclusions about reality from any fMRI study (I'm not convinced what is being measured is all that important plus even intrepreting what is being measured relies on complex statistics). Also, for this kind of thing the outliers are very important in tring to figure out what is going on. They show individual data points and eyeballing them I don't really see a lot of difference in the accuracy, with the tablet in particular being slightly worse than nearly identical phone or notes (but they had to pop out "easy questions" to make that difference look interesting). Saying there is a difference between device and paper just because the tablet condition did worse is misleading.
They mention the difference in terms of location cues in the conclusion section. I would guess UI choices affect the ability to use location sense to recall information and I could easily imagine the small menus of phones enhancing this effect (and/or larger screens in general reducing the physical location effect of the device itself). Taking a quick look on Google Scholar there is some research on this kind of thing but I'm not seeing any thing that pops out as a particularly interesting presentation. Any UI designers out there who intentionally try to trigger location sense in their design?
https://www.frontiersin.org/files/Articles/634158/fnbeh-15-6...
They mention the difference in terms of location cues in the conclusion section. I would guess UI choices affect the ability to use location sense to recall information and I could easily imagine the small menus of phones enhancing this effect (and/or larger screens in general reducing the physical location effect of the device itself). Taking a quick look on Google Scholar there is some research on this kind of thing but I'm not seeing any thing that pops out as a particularly interesting presentation. Any UI designers out there who intentionally try to trigger location sense in their design?