Blind FLM web-based tools for keystroke-level predictive assessment of visually impaired smartphone interaction

Abstract

The keystroke-level model (KLM) is a predictive model used to evaluate motor behaviour in skilled error-free user interaction involving conventional techniques, i.e. mouse and keyboard. A blind fingerstroke-level model (blind FLM) was recently introduced as an extension of KLM to assess visually impaired interaction on smartphones. The model comprises six operators that are used to calculate the time required for a visually impaired expert user to accomplish a task on a smartphone. In this paper, we present two blind FLM tools: calculator and editor. These tools enable designers to create behavioural models of user tasks from which reliable estimates of skilled user task times can be computed. Each tool was used to model a sample task on YouTube to assess its performance against previously recorded values. Both tools accurately predicted user performance with an average error of 1.27%.