How have researchers of TESOL analyzed data? How can they do it better?
What is it about?
When I wrote this article practices of inferential statistics (e.g., null hypothesis significance testing) in research of instructed second language learning were on the whole remarkably outdated. A cause of this situation, I suspect, was training based around textbooks written by and for experimental psychologists working in fields that do not involve linguistic data. An indication of this is the fact that the survey I report in this article came across hardly any uses of mixed-effects modeling. So one of the things the article is about is better methods for the kinds of data analysis that L2 researchers typically engage in.
Why is it important?
Because suboptimal statistical practices can result in misleading conclusions.
The following have contributed to this page: Mr Seth Lindstromberg