Independent Welch tests carried out on a random sample of Chinese speakers from this study (n=53) and the whole mixed sample (n=53) give a significant p value for a difference in mean receptive vocabulary knowledge (<.0001) with a very large effect size (cohens d=.996). The same test repeated for a second random sample of Chinese students (n=53) and the mixed sample (n=53) give a p value of the same significance (<.0001) and large effect size (Cohen’s d=.785). However, the mixed sample does not necessarily represent a population with very consistent characteristics, given its diversity so these results should be treated with some caution.
