Measurement Scales

Statistics for Psychologists – Measurement Levels

In this course, Professor Dominic Dwyer (Cardiff University) explores measurement levels and scales. In the first lecture, we think about the origin of measurement scales in psychology, and the role of Stanley Smith Stevens in establishing the four measurement levels: nominal, ordinal, interval, and ratio. In the second lecture, we think about nominal scales and their prevalence in behavioural genetics, despite statistical limitations. In the third lecture, we think about ordinal scales and their regular use in survey data collection. Next, we think about ratio scales and the importance of zero meaning something (nothing) in these scales. In the fifth lecture, we think about interval scales and their lack of applicability to ratios of data, due to the arbitrary nature of the zero value. In the sixth and final lecture, we think about how these measurement scales can be summarised and discuss some of the challenges to Stevens’ understanding of measurement.

Measurement Scales

In this lecture, we think about measurement scales and their origin within the field of psychology, focusing in particular on: (i) early research into measurement in psychology in the 1940’s, which concluded that things were not being measured in the field; (ii) Stanley Smith Steven’s disagreement with this viewpoint, highlighted in his 1946 paper, which argued that measurement was being misunderstood; (iii) Stevens’ four levels of measurement: nominal, ordinal, interval and ratio; (iv) examples of each of those measurement levels in real data; (v) the book series ‘Foundations of Measurement’, which provides the most comprehensive review of how to apply statistical tools to measurements.