ABSTRACT
In quantitative research, the most effective way to avoid dealing with this problem and introducing a subject selection bias is to work with learners who have developed a high level of proficiency in their background languages. While non-native languages have been regarded as a potential source of crosslinguistic influence (CLI) for a long time, their treatment in quantitative research has been inconsistent and uneven at best. In this chapter, the author discusses why the presence of non-native languages in learners’ minds has not been examined more consistently in the past and the type of problems associated with non-native language knowledge, particularly in relation to the possible introduction of a subject selection bias in research. She also discusses how students can conduct research with multilingual individuals without introducing a subject selection bias in the design. By referring to a real example of quantitative research on CLI and multilingualism, the author outlines how she dealt with proficiency assessment.
