The Usage of Predictive Text Technology and Its Effects on Written Language Accuracy

Abstract

With predictive text technology increasingly prevalent in digital communications, its impact on users' language accuracy warrants investigation. This study examined how different aspects of predictive typing usage affect the written linguistic precision of tertiary-level students in Bangladesh. A mixed-methods approach was employed, in that the study dealt with both quantitative and qualitative sets of data. 51 students from five universities in Dhaka participated in the study. The researchers asked participants to answer a 12-question survey with scaled and open-ended questions regarding their predictive text use and perceived writing quality. The findings revealed that while 90% of participants use predictive text technology, most report reduced attention to spelling and grammar when writing without access to such tools. However, the participants showed awareness of the cognitive cost of overusing the technology, and some reported regular proofreading habits. The authors conclude that predictive text is neither entirely harmful nor purely beneficial; rather, it is used as a means to an end. Suggestions include incorporating AI literacy strategies into EFL writing teaching, supervised tasks with AI support, and adjustable predictive text features. These findings provide insights for educators, linguists, and language tool developers in Bangladesh and similar multilingual contexts.