The National Institute of Health (NIH) Executive Abilities: Measures and Instruments for Neurobehavioral Evaluation and Research (EXAMINER) project developed a laptop-based battery of executive functioning tests that were intended for use as an outcome measure for clinical trials and observational research. The project was completed in 2013 and the EXAMINER has been implemented in a myriad of research studies. Empirical investigations, along with feedback from participants and research teams, have identified opportunities for improving the battery. Moreover, tablets have recently gained traction over laptops as a convenient and powerful assessment tool. Thus, a new tablet version of the EXAMINER was created on the UCSF Tablet-based Cognitive Assessment Tool platform (TabCAT-EXAMINER). Here we describe the development of the updated EXAMINER battery, the creation of a global executive functioning factor score, and present validation analyses.

A diagnostically heterogeneous sample (e.g., cognitively unimpaired (CU), mild cognitive impairment (MCI) and dementia due to Alzheimer’s disease (AD), behavioral variant frontotemporal dementia (bvFTD), and Parkinson’s Disease) of 1328 individuals (agemean=65.71, ageSD=10.68) completed up to five tablet-based assessments (Flanker, Set-Shifting, Match, Dot Counting, Running Dots) and four verbal fluency measures (animals, vegetables, L-words, and F-words). A single factor CFA model was fit and multilevel item response theory was used to extract factor scores. Using linear regressions, the resultant score was validated through associations with six independent executive functioning assessments, demographics, and disease severity (Clinical Dementia Rating Scale; CDR). We also examined performance across a subset of diagnostic groups: CU, n=507; MCI, n=216; AD dementia, n=100; Parkinson’s disease, n=33, and bvFTD, n=52. Voxel based morphometry analysis was conducted with CU participants who had available MRI (n=196) to evaluate the neural substrates of TabCAT-EXAMINER performance.

Factor loadings suggest the factor score was moderately to highly associated with all components (range: .54 to .85). Younger age (r=-.26; 95%CI: -.31, -.2) and higher education (r=.28, 95%CI: .22, .33) were associated with better performance. Convergent validity was supported by positive associations with independent tests of executive functioning (r’s range: .43 to .67) and associations with disease severity as defined by CDR-box score (r=-.61, 95%CI: -.64, -.57) and diagnosis (CU>MCI>dementia; p’s<.001). There were significant differences between the CU and neurodegenerative groups (CU > PD, bvFTD, AD; ps < .006); the AD group performed significantly worse than the bvFTD group (p=.008). Within the CU subsample that completed an MRI, higher factor score was associated with larger volume in frontal and temporal regions, as well as areas within the thalamus and cerebellum.

The TabCAT-EXAMINER is a tablet-based executive functioning battery developed for observational research and clinical trials. The construct validity of the IRT-generated factor score was supported by expected associations with demographics, independent measures of executive functioning, disease severity, and brain volume. Future research will study the utility of this score for tracking longitudinal change and for use in diverse samples.