Figure 5.

Summary of key findings

(A) We captured real-world behavioral dynamics using a joint-interval distribution of next intervals derived from smartphone touchscreen interactions. The cognitive test focused on sensorimotor processing correlated with distinct parts of the dynamics in contrast to the tests focused on executive functions.

(B) We used parametric modeling to link the chronological age with the smartphone behavioral dynamics, and according to this model young vs. old adults generate distinct dynamics. While youthful behavior was characterized by higher behavioral probability densities at the shorter intervals, the elderly showed a higher density at the longer intervals.

(C) Of all the examined factors, chronological age best explained the inter-individual differences in the smartphone behavioral dynamics.