Threshold concepts were introduced nearly 10 years ago by Ray Land and Jan Meyer. This work has spawned four international conferences and hundreds of papers. Although the idea has clearly gained traction in higher education, this sub-field does not yet have a fully fledged research methodology or a strong critical discourse about methodology. This paper seeks to begin such a dialogue by analysing three projects carried out by the authors, each focused broadly on identifying and understanding threshold concepts in engineering. Each of the projects used interviews with students and academics, but differed in six main ways that seem to make a difference to the research outcomes. This paper considers the gaps in the research and why they matter, briefly outlines the methods used in each of the three case study projects, and then discusses differences in project goals, researchers' backgrounds, curricular context, participants' experiences, negotiated or independent knowledge and degree of comprehensiveness sought in the studies. The implications of these six differences are explored. The authors argue that research in this sub-field of higher education pedagogical research needs to be clearer and more explicit about the methods that are used They conclude that the field would benefit from bringing together researchers who have been developing complementary research methods to compare and contrast these approaches and to develop more rigorous protocols for research on threshold concepts.