An Introduction to Statistics in Linguistics (2): Statistical Scales and Linguistic Data

Abstract

Linguistic data are typically classified based on the nature of the values into one of four categories known as the nominal scales, ordinal scales, interval scales, and ratio scales. The nominal scales and ordinal scales are useful for qualitative data and the interval scales and ratio scales are useful for quantitative data. Since each of these statistical scales allows different calculations, there are different statistical and data visualization methods suitable to each scale. This paper gives an overview of these methods. In addition, given that the majority of linguistic data are qualitative, in this paper, we show that researchers must be careful when taking a statistical approach to such data.