Machine translation (MT) use is widespread among language students in higher education, including in the undergraduate translation classroom. This highlights the need to develop MT literacy in a manner that improves translation performance as well as translator self-efficacy beliefs to prepare students for a professional future that will require them to be critical and ethical users of MT and other artificial intelligence (AI) tools. Previous research has shown that post-editing (PE) training can have a positive impact on MT literacy, translation performance, and self-efficacy beliefs. However, the fluency of MT output makes it difficult to spot errors, particularly among L2 learners, as is the case in undergraduate L2 translator training. This action research study therefore investigated two different didactic approaches to enhancing MT literacy delivered to two groups of students in an undergraduate L2 translation course: one based on post-editing, which is common in translator training (the post-editing group), and the other based on using MT as one linguistic tool amongst several, as is common in second language learning (the use-as-a-tool group). Using a convergent parallel mixed methods design, quantitative and qualitative data such as questionnaires, translation products and commentaries, screen recordings, and focus group discussions were collected pre-, post-, and delayed post-intervention. The results reveal that both didactic approaches positively impacted MT literacy, translation performance, and self-efficacy beliefs related to the use of online resources because they made participants aware of the limitations of MT and of the need to use other external resources to solve translation problems. However, the positive effect on MT literacy was stronger in the use-as-a-tool group, likely because it afforded participants in that group with the opportunity to experiment more with MT and learn from their mistakes. These results can inform the design of future MT literacy training in the undergraduate L2 translation classroom.
Building similarity graph...
Analyzing shared references across papers
Loading...
Elana Summers
Institute of Translation and Interpreting
Building similarity graph...
Analyzing shared references across papers
Loading...
Elana Summers (Thu,) studied this question.
synapsesocial.com/papers/6a0ff1dbd674f7c03778b184 — DOI: https://doi.org/10.21954/ou.ro.00109984