Open AccessWith practice, keyboard shortcuts become faster than menu selection: A crossover interaction.
Author(s) -
Roger W. Remington,
Ho Wang Holman Yuen,
Harold Pashler
Publication year - 2016
Publication title -
journal of experimental psychology. applied
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.004
H-Index - 84
eISSN - 1939-2192
pISSN - 1076-898X
DOI - 10.1037/xap0000069
Subject(s) - computer science , human–computer interaction , keystroke logging , task (project management) , crossover , selection (genetic algorithm) , interface (matter) , preference , graphical user interface , artificial intelligence , operating system , statistics , engineering , mathematics , systems engineering , bubble , maximum bubble pressure method
It is widely believed that a graphical user interface (GUI) is superior to a command line interface (CLI) for novice users, but less efficient than the CLI after practice. However, there appears to be no detailed study of the crossover interaction that this implies. The rate of learning may shed light on the reluctance of experienced users to adopt keyboard shortcuts, even though, when mastered, shortcut use would reduce task completion times. We report 2 experiments examining changes in the efficiency of and preference for keyboard input versus GUI with practice. Experiment 1 had separate groups of subjects make speeded choice responses to words on a 20-item list either by clicking on a tab in a dropdown menu (GUI version) or by entering a preassigned keystroke combination (CLI version). The predicted crossover was observed after approximately 200 responses. Experiment 2 showed that following training all but 1 subject in the CLI-trained group chose to continue using shortcuts. These results suggest that frequency of shortcut use is a function of ease of retrieval, which develops over the course of multiple repetitions of the command. We discuss possible methods for promoting shortcut learning and the practical implications of our results.