A priority-based model for smartphone screen touches

The distribution of intervals between human actions such as email posts or keyboard strokes demonstrates distinct properties at short vs long time scales. For instance, at long time scales, it has been shown that those inter-event intervals follow a scale-invariant (or power-law) distribution. In contrast, little do we know about the events that occur at the shorter time-scales and how they relate to the scale-invariant pattern. Here, we analysed the intervals between smartphone screen touches of 84 individuals which span several orders of magnitudes (from milliseconds to days). To capture these intervals, we propose a priority-based generative model for smartphone touching events. At short-time scale, the model is governed by refractory effects, while at longer time scales, the inter-touch intervals are governed by the priority difference between smartphone tasks and other tasks. The flexibility of the model allows to capture inter-individual variations at short and long time scales while its tractability enables very efficient model fitting. We show that each individuals has a specific power-low exponent which is tightly related to the effective refractory time constant suggesting that the fine motor control properties (which influence short intervals) are related to the higher cognitive processes (which affect longer intervals).