Blurb:

This study investigates the effectiveness of automated feedback on causal language based on student perceptions at a higher education level. The target automated writing evaluation tool was developed based on the PRINCIPLES semantic text pattern of Mohan’s Knowledge Framework. Analyzing learners’ perceptions provides an opportunity to identify problems stemming from the pedagogical features of the tool and to determine the adjustments needed for better learning outcomes. To this end, in this qualitative study, we evaluated automated feedback on causal language from the perceptions of 27 undergraduate ESL students who used the tool for two cause-and-effect assignments in an academic writing class. Data were collected through semi-structured interviews in which students were asked questions regarding the capacity of the automated feedback features of the tool to help students focus on causal forms and causal meaning and to create opportunities for interactional modifications. Findings showed that the text-level feedback was more positively perceived by the students for focus on causal forms than the sentence-level feedback. On the other hand, the sentence-level feedback was more positively perceived for focus on causal meaning than the text-level feedback. Timing of the automated feedback appeared to be a source of negative perceptions regarding interactional modifications.