To adequately tackle a research problem, master students of applied linguistics should learn that the selection of (a) data elicitation technique (s) should be made in consistence with choices at three other levels: method, methodology and paradigm. Hence, this paper addresses the following question: how should the methodology course be reformulated to render it more efficient in raising students' awareness of this issue? An analysis of some research methods manuals currently in use reveals that two major obstacles hinder students 'ability to learn this issue: the pluralistic nature of applied linguistics and the rampant use of mixed methodologies. To overcome these obstacles, this paper proposes a teaching strategy consisting of focusing the initial phase on a contrastive analysis of two methods, which stand at extreme positions on the methodological continuum in applied linguistics, namely, experimental design and ethnography. Moreover, given that the presentation of the differences between these two methods is not sufficient, the paper argues that this presentation should be reinforced by a foregrounding of the essential differences in problem statement in the two research traditions in question. The paper concludes with some recommendations on the appropriate way to implement the proposed teaching strategy.