Neuropsychological assessment by videoconference: Mean score differences, individual-level agreement, and connectivity-related disruptions.

Objective

To examine mean score differences and individual-level agreement between videoconference and in-person administration of a multidomain neuropsychological assessment battery, and associations between connectivity parameters and task-specific disruptions. Method: In a randomized, counterbalanced within-participant design, 110 healthy Portuguese adults completed the battery by videoconference and in person, one month apart. Participants were predominantly young to middle-aged and digitally proficient; only 2.7% were aged ≥60 years. Mean differences were examined using paired t tests, Cohen's dz, and Bayes factors, and agreement using intraclass correlation coefficients and Bland-Altman analyses. Sensitivity analyses examined administration-order effects, and logistic regression assessed latency, jitter, and packet loss as predictors of disruptions.

Results

Mean score differences were small. MoCA scores were slightly higher by videoconference (M = 26.99, SD = 2.46) than in person (M = 26.42, SD = 2.39; dz = -0.26, p = .011), but this depended on administration order and was consistent with practice effects. No other measure differed between modalities (all ps > .05; |dz| ≤ 0.19). Individual-level agreement varied across measures (ICC = 0.20-0.82), with lower agreement for several derived verbal-memory indices. Technical disruptions were infrequent and task-specific; higher jitter was associated with disruption in MoCA Letter Series/Vigilance (OR = 1.19) and IFS Go/No-Go (OR = 1.10).

Conclusions

Videoconference and in-person administration produced small mean score differences but variable individual-level agreement. Findings support task-specific interpretation and indicate that small mean differences do not imply consistent individual-level agreement. Generalization to older and clinical populations remains limited.