Diagnosing and coping with mode errors in using a dual-language keyboard

Abstract

In countries that do not use Latin characters, users use a keyboard with two language modes: native language and English, and often experience input mode errors. To diagnose the mode errors, we conducted a field study and found that 60% of the mode errors occurred immediately after application switching. Based on this finding, we implemented four methods (Auto-switch, Prevent, Recover and Prevent & Recover) to cope with the mode errors, and conducted field studies to verify their effectiveness. In Korean-English dual input environments, Auto-switch was not effective. On the other hand, Prevent significantly reduced mode errors (from 75.1% to 41.3%), and Recover showed a high adoption rate (70.2%). In Prevent & Recover method, Prevent reduced mode errors (from 75.1% to 43.8%) and Recover handled 86.23% of the mode errors that slipped past Prevent. We hope that the findings of this study will motivate further studies coping with mode errors in other dual input environments.