Drawing on the perspective of decolonizing methodology, this paper problematises the conventional interview technique of avoiding yes-or-no questions introduced by handbooks of research methods in applied linguistics. By demonstrating how this interview technique was found unsuitable in the first author's ethnographic study of multilingualism with the Blang community, an indigenous group inhabiting the Chinese border with Myanmar, we argue that this oversimplification of avoiding yes-or-no questions implies an ideology that long responses are the desired product of interviews and is underpinned by a colonised research paradigm and mindset from an elite perspective. It overgeneralises the complex process of co-constructing knowledge with participants in multilingual contexts and results in neglecting the diverse lived experiences of grassroots and the sociohistorical contexts they live in. As a result, voices of these under-represented groups, especially those who are under-communicative and under-educated are further marginalised. We call to decolonise our epistemological thinking and practices influenced by ideologies as such. We also call for researchers to be more reflexive in their research practices and to advance research methodological decolonisation with insights generated from constant reflection.