The study described in this article was prompted by the poor performance of students in an "at risk subject" in a science faculty at a university in South Africa. Teacher performance could contribute to poor performance among students, therefore the performance of one of the science teachers whose students were performing poorly was evaluated by his students and through peer observation of his teaching. The article draws on a merger between Bernstein's ideas on framing as well as deep and surface approaches to learning to form a theoretical framework that underpinned the study. Peer observation showed that the teacher employed predominantly teacher-centred, passive approaches to teaching, and the facilitation of active learning was minimal, that is, framing was strong. Students, however, evaluated their teacher positively, indicating that he was an "effective" teacher. Therefore, the perception of what constitutes "quality teaching" is viewed differently by peers and students. This is most likely due to the incompatibility between peers' conception of teaching and students' conception of learning. Therefore, students' feedback on teaching is not necessarily accurate or useful.