ABSTRACT
While translation and post-editing are often seen as a problem solving activity, it is difficult to subjectively assess when individual translators have problems during translation. Especially when teaching translation classes, the teacher has to objectively decide which problems the students faced when they prepared the translation. I first proposed a theoretical approach to extracting translation problems in Nitzke (2019). The approach builds on the idea that potential translation problems in a text can be identified by looking at translation process information found in eye-tracking and keylogging data. It combines data on temporal, technical, and cognitive effort. In this paper, I test the proposed approach by applying it to identify potential problems, here potentially problematic verbs, in process data from 24 translators (12 students and 12 professionals) taken from the CRITT TPR-database. The paper contains the calculations of the respective formulas used to identify translation problems and a review of the results. Finally, I discuss the advantages and shortcomings of this approach, how it could be applied in teaching translation, and what further research might be necessary.
