Abstract Details

It is more important than ever that researchers have access to well-designed cognitive assessments that can be administered remotely.  Currently, there are few tools that can be easily and widely deployed across the lifespan, in different research settings, and with diverse participants. The Mobile Toolbox (MTB) is a library of app-based cognitive tests within a research platform that enables both study management and data collection for at-home smartphone administration. It is available for both iOS and Android devices. MTB will fill the gap for early detection of cognitive deficits related to age-associated changes.  

To provide initial validity evidence supporting use of the Mobile Toolbox (MTB) for remote cognitive assessment. 

We evaluated the validity of MTB’s cognitive assessments among participants recruited by a market research firm. These tests measure executive function, language, memory, and processing speed. The NIH Toolbox Cognition Battery and a subset of external validity measures were also administered.   

A total of 1140 individuals took the MTB core 8 measures and the NIH Toolbox v3. Of these, 99 took the MTB in person at the same time as external validity measures.  These 99 had a mean age of 50 (SD=18), and 64% were female.  This is consistent with the sample demographics for the larger number of participants who took the MTB measures at home. The majority of participants completed the MTB measures in one session, though they were allowed to return to the tests over the course of 2 weeks. Measure-specific reliability estimates (IRT-based reliability, split-half reliability, or Cronbach’s alpha) suggests acceptable reliability for remote large-scale assessments (rxx>0.65). Scores were highly correlated with NIHTB v3 and external validity measures of comparable domains.