Modelling typing disfluencies as finite mixture process

Abstract To writing anything on a keyboard at all requires us to know first what type, then activate motor programmes for finger movements, and execute these. An interruption in the information flow any of these stages leads disfluencies. capture this combination fluent typing hesitations, researchers calculate different measures from keystroke-latency data—such as mean inter-keystroke interval pause frequencies. There are two fundamental problems with this: first, summary statistics ignore important data frequently result biased estimates; second, pauses pause-related defined using threshold values which are, principle, arbitrary. We implemented series Bayesian models that aimed address both issues while providing reliable estimates individual speed statistically inferred process tested random sample 250 copy-task recordings. Our results illustrate we can model copy mixture disfluent key transitions. conclude (1) map onto cascade generate keystrokes, (2) provide principled approach detect disfluencies typing.