A passion for psychometrics: what it is and what I do

By Chris Hope

Monday 8 April 2019

Psychology is an intrinsically interesting subject. Human beings are fundamentally social creatures so a field of study that focuses on human thought, behaviour, motivations and attitudes will always be appealing. As a relatively new science, some 140 years old, it is a new and rapidly evolving discipline and one that blends both the humanities and the sciences together so that people with a broad range of talents and backgrounds can enjoy it and contribute.

Despite my enthusiasm as an undergraduate student, I was often met with scepticism about how useful my degree would be once I started looking for a job. People seem to get the impression from TV personalities and pop-psychology books and magazines that Psychology is a soft, fluffy pseudoscience. I’m pleased to say however that my undergraduate and postgraduate education had enough rigor to prepare me well for the world of work and that there are sufficient employment opportunities that make direct use of ideas and methodology from the world of academic Psychology.

Psychometrics, my current area of employment, is a good example of an occupation that draws heavily on Psychology. People who practice psychometrics (Psychometricians) often make use of psychological theories of personality, intelligence and attitudes when designing questionnaires or assessments and use the same software and statistical techniques learnt at university. We aim to develop precise measurements of complex and abstract psychological traits in order to assess people fairly and to facilitate scientific understanding.

Whether it’s personality attributes such as extraversion or phenomenon that involve an element of human perception such as the tastiness of wine, psychometrics enable measurement in the same sense that height can be measured with a tape measure. It does this by estimating the relative difficulty or location of each “item” used to assess a particular trait, such as a question in a questionnaire, on a scale. For example, taking the measuring tape analogy, extraversion questionnaire items may be arranged at irregular intervals along the extraversion scale. To agree with the question “Do you like parties?” might require 2.5 units of extraversion, whereas to agree with “Do you enjoy talking in front of large groups of people?” could require 7 units and “Do you enjoy singing at karaoke in front of strangers?” may need 10 units. In order to accurately measure a person’s extraversion, you need to account for the difficulty or location of the items on the scale. Psychometric methods mathematically model questionnaire or assessment data to take account of question location on a trait of interest and use this information to optimally measure a person.

In my role as a Psychometrician at NFER, I use my skills to statistically analyse the educational assessments we produce during the development phase to check that they work in a reliable fashion and that the difficulty level is correct etc. Part of this work also entails running mathematical models to generate psychometrically valid measures and generating age standardisation tables to allow teachers to make sense of their pupil’s performance in comparison with the national average. In terms of research work, I regularly work on analysing international survey data such as PISA and TIMSS, which often involve a substantial psychometric component. Development work includes an exploration of Computer Adaptive Testing (CAT), an area that involves an in-depth knowledge of psychometrics to produce computerised tests that adapt to the person taking them to deliver questions that are appropriate for their ability level.

NFER is a great organisation to work for. When I first started at NFER, I had no formal education in psychometrics which they remedied by sending me on both an introductory and an advanced course in the area run by David Andrich, a world renowned Psychometrician. I have also been lucky enough to attend several conferences and workshops across Europe. NFER is a very friendly and supportive organisation where people are always willing to help and problems are usually solved through collaborative effort. The work-life balance is better than other places I have worked before. Our organisation is growing and we’re always on the lookout for psychometricians to join the team, so feel free to get in contact if you are curious!

Chris Hope is a Research Manager/Psychometrician at NFER and can be contacted at [email protected]

If you are interested in a statistician role at NFER, sign up to our job alerts and we will let you know when we are hiring.