How not to blindly measure the mind
In Psychological Testing: Theory and Practice (2019) Cooper does not offer a step-by-step guide to using statistical packages to validate scales; nor does he provided the mathematical formulae underpinning these. In that sense, he leaves the ‘how’ and the ‘what’ to other texts.
Instead, Cooper gives the ‘why’. ‘It is all too easy to run numbers through a powerful statistical package without thinking too much about the legitimacy of what one is doing’ he writes (p.236).
The text is a valuable resource for graduate students approaching their viva (myself included) or early-career researchers submitting for their first big grant. Drawing heavily on Michell’s (1997) critique of psychometrics as a pseudo-scientific discipline, Cooper opens with a word of caution. Psychological measurement has inherent epistemological limitations that need to be considered.
With this in mind, readers are then treated to 14 concise, eloquent and down-to-earth chapters detailing what psychometric testing can offer. Key tenets of Classical Measurement Theory are covered conceptually and theoretically with recourse to concrete examples and interactive spreadsheets (downloadable for free). The book covers everything needed for developing a scale with a critical eye to both construction and use, including: item generation, reliability, validity, correlations, measurement error, Exploratory Factor Analysis and emerging alternatives, and interpreting questionnaire scores.
Psychological Testing feels comprehensive. However, it would benefit from a more detailed discussion of missing data, Confirmatory Factor Analysis and Structural Equation Modelling. The latter is promoted by the author repeatedly as one of our best answers to the limitations to Classical Measurement Theory described by Michell (1997). Whilst Item Response Theory in general is given healthy contemplation, Mokken scaling is given only four pages.
Overall, Cooper delivers. The text gives readers such as myself the ammunition needed to critically select methods and justify why we develop and validate instruments of psychological measurement. For any researcher blindly running numbers through a powerful statistical package, Psychological Testing is a must-read.
- Reviewed by Jack Tomlin, PhD Researcher & Tutor in Criminology, Associate Fellow of the Higher Education Academy, Institute of Mental Health, Division of Psychiatry and Applied Psychology, University of Nottingham
BPS Members can discuss this article
Already a member? Or Create an account
Not a member? Find out about becoming a member or subscriber