Predictive Models of Typing Performance in Novel Keyboards

Publication Type:

Conference Paper

Source:

Companion Proceedings of the 18th ACM SIGCHI Symposium on Engineering Interactive Computing Systems, Association for Computing Machinery, New York, NY, USA (2026)

ISBN:

9798400726699

URL:

https://doi.org/10.1145/3807968.3816000

Keywords:

Spatial Memory, text entry, User studies

Abstract:

Novel mobile keyboard layouts require users to learn new spatial mappings, leading to performance improvements with practice. We examine how typing behaviour evolves while learning an unfamiliar virtual keyboard layout, aiming to assess the number of typing sessions required to obtain data that allow reliable predictive modelling of performance over sustained use. We find that logistic function models outperform the power function models proposed by past literature, in terms of robustness and predictive ability.

Download full text: