Statistical Literacy: Ratio-Based Grammar Research

Confounding is a major problem in dealing with statistics obtained from observational studies. The influence of confounders can be blocked in two ways: physical (effect size and study design), and mental (selection and ratios). Ordinary English is typically used to describe and compare ratios. Today's students need guidance on the proper use of English to form descriptions and comparison that are accurate. They have no idea of whether "the percentage of men who smoke" is the same as "the percentage of men among smokers". They have no idea of whether "men are more likely to smoke than [are] women" is the same as "smokers are more likely [to be found] among men than [among] women." or "men are more prevalent among smokers than [are] women." In 2023, Kendall-Hunt published a statistical literacy textbook that provides provisional classifications and grammatical structures for describing and comparing ratios. In order to provide guidance that is generally accepted, it must be empirically based on common usage. This requires research using a large corpus of text based on written and spoken English. This paper identifies the data sources used in generating these classifications and structures. And it summarizes the provisional classifications and grammatical rules obtained from this investigation. While, this paper indicates the connection between these data sources and the results, it does not – and cannot – detail the thinking and choices involved in obtaining the results from these sources.