Ellen Yezierski, Justin Carmel, Brittany Christian (Miami University, USA)
It is critical to ensure that statistical procedures employed for data analysis align with the research questions and the features of the data. Frequently the sample size, type of data, distributions, and questions of interest warrant the use of non-parametric procedures. Some non-parametric techniques have a parametric analog while others simply test questions that simply cannot be addressed with parametric techniques (i.e. comparing frequencies). This talk features non-parametric techniques employed in two chemistry education research studies. The rationale for the techniques, their assumptions, tools for analysis, and authors’ considerations for the narrative and graphical presentation of data and results will be emphasized.