Abstract
This study examined the effect of QuillBot on vocabulary and writing style in Grade 8 learners' English essays at a public high school in South Africa. The research was guided by two conceptual frameworks: the Technology Acceptance Model and the Theory of Planned Behaviour. The aim was to evaluate how QuillBot can enhance learners' grammar accuracy, vocabulary, fluency, coherence, and writing style. It also sought to understand learners' perceptions of QuillBot’s usability and usefulness.
A mixed-method approach was adopted for this study. For the quantitative aspect, data were collected through pre- and post-tests, as well as questionnaires. Qualitative data were collected through focus group interviews and the teacher-blinded comparison tool. The study had a sample of 64 grade 8 learners through purposive sampling, representing diverse backgrounds in home language. Statistical and thematic methods guided the analysis of the data. The quantitative data were analysed using descriptive statistics to show any changes in the learners' writing skills. In contrast, the qualitative data were analysed using themes to gain insight into the learners' experiences.
The findings indicate that QuillBot was effective in improving grammar accuracy, vocabulary, as well as fluency and coherence in learners’ writing. The data revealed that learners reported QuillBot as an easy-to-use tool that they found helpful in editing their work. However, there were drawbacks regarding their writing style post-intervention, where learners mentioned that the tool changed their tone to become too formal for grade 8 learners, and there was a risk of over-reliance on the tool. The results highlight both the benefits and limitations of integrating AI tools, such as QuillBot, in classrooms.