“…So we guess that the significant positive predication of perceived autonomy support on agentic engagement might occur largely through the mediation of students' need satisfaction (Jang et al, 2012 ; Cheon and Reeve, 2013 ; Reeve, 2013 ; Reeve and Shin, 2020 ) and students' perceived autonomous reasons for action (Benita et al, 2014 ) from an SDT perspective. In the context of EFL teaching, take writing class for instance, various sources of feedback, either technology-assisted autonomatic feedback, teachers' feedback, or peers' feedback, which are intended to support students' autonomy, can promote students' agent engagement in teacher-student and student-student interactions, largely because these sources satisfy students' needs for competence, i.e., to improve their language proficiency (see Tian and Zhou, 2020 ). However, the result that autonomy support was negatively associated with master-approach goals is a bit beyond expectation, because theoretically as previous elaborated, teachers' autonomy supportive teaching contributes to building an autonomy supportive classroom goal structure, which in turn, should promote students' master-approach goal orientation (Meece et al, 2006 ).…”