Abstract Empirical testing is a challenging aspect of L2 classroom-based research, especially when exploring the multifaceted nature of figurative language, such as metaphorical idioms. Typically, studies in applied linguistics involve language learners from convenience samples of intact classes. This approach can pose problems as these classes represent non-random, often small, samples of participants. Despite these challenges, appropriate precautions and considerations, such as addressing overlooked idiom-inherent variables, contemplating counterbalancing, managing time-on-task, and making well-informed treatment and data collection design choices can minimize confounding variables and enhance a study’s design and resultant validity. The author of this article offers a reflective commentary based on a previous study (Ramonda, 2022) to expound on these considerations and provide modest proposals for improving future study quality in this domain.