CAT Tools in Translator Education: An Experimental Investigation of Efficiency, Quality, and Mental Workload

Abstract

Computer-assisted translation (CAT) tools have revolutionised translation practices by automating repetitive tasks, promoting consistency in terminology usage, and providing translation suggestions through generative artificial intelligence. This study investigates the impact of CAT tools on student translators’ performance in English-to-Chinese translation tasks, focusing on translation efficiency, quality, and mental workload. A within-subjects crossover design was employed with 30 undergraduate translation students, who completed translation tasks under both CAT-assisted and non-CAT conditions. Translation efficiency was measured via recorded completion times, translation quality was evaluated using a structured rating rubric emphasising accuracy and consistency, and mental workload was assessed using the NASA-TLX scale. The findings indicate that CAT-assisted translations were completed significantly faster and achieved higher overall quality scores. However, a higher incidence of core semantic errors was observed in CAT-assisted outputs, suggesting that reliance on machine-generated suggestions can lead to overly literal translations. In contrast, no significant difference in perceived mental workload was detected between the two conditions. These results suggest that while CAT tools enhance efficiency and overall translation quality, they do not necessarily reduce mental strain during translation tasks. The present study underscores the need to integrate CAT tool training into translator education curricula. Emphasising digital literacy and technology acceptance in translator training can help students maximise the efficiency and quality gains of CAT tools while maintaining high translation standards and strengthening their professional competence.