This was presented at the University of Westminster on 21st April 2015 at a session arranged by the ABP.
- Unease that, with the dramatic internet fuelled increase in the use of personality assessment and the proliferation of questionnaires, the good practice standards painstakingly developed by the profession are proving largely ineffective. An earlier generation of test publishers were very aware of the importance of reputation for the credibility of the process. At TPC (archetypal test publisher established by Cattell and colleagues around 1920 – world’s biggest test publisher when I worked for them in the 1980s) the psychologists and psychometricians called the shots – not the accountants or the sales team, or current vogues, fads and fashions. Potentially money making propositions were abandoned if stringent research evidence didn’t stack up.
- Concern that the pendulum has swung away from understanding (appreciating the benefits and limitations imposed by personality characteristics and what they mean in terms of job challenges and success) and towards a simplistic focus on test scores (with an emphasis on prediction that over simplifies the relationship and stretches the credibility of the process).
- There are two clear strands to psychometric personality assessment; the maths and the psychology. Cattell argued for both ‘psychological understanding … and statistical understanding’. Former APA president Frank Landy argued for a “focus attention on understanding rather than on the simpler and less meaningful search for predictability”.
- Psychometric measurement is challenged by the complexity of cause and effect of behaviour, the reliance on inference and estimation and the ambiguity of the language that we need to use to think and communicate about it.
- In spite of these hurdles, the current understanding of personality assessment and its descriptive output is remarkable. We know how to capture the major features that characterise an individual’s social behaviour with impressive reliability. These descriptions are implicitly predictive in that they tell us ‘what someone is like’. This aspect of personality assessment maximises the essentially lexical nature of the Five Factor Model. Put simply, FFM assessments tell us which words best describe the person assessed. This is potentially very useful, it is the best that has ever been available and, to my mind, it is under valued and under utilised.
- Language is fluid and ambiguous. Most words have numerous synonyms. ‘Conscientious’, for example, has 21 synonyms in my thesaurus. That’s potentially 21 different nuances. Words can imply different things to different people. In translating test scores to narrative speech or text, beware the ‘Semantic merry-go-round’ in which meaning slides from synonym to synonym, from one meaning to another (exploited by the Barnum effect). FFM identifies five ‘hot spots’ of meaning in this sea of personality focused vocabulary, imposes some lexical discipline and provides pegs against which to align behavioural observations and other research findings.
- It is questionable how far it is possible to take the implicit prediction already mentioned (understanding what someone is like) in the direction of explicit prediction relating to specific work behaviours. The problem for prediction is that we are sentient beings that have free will. Most people are able to perform outside their own comfort zone in many different ways for significant parts of their day. Belief and motivation will usually trump temperament, at least in the short term.
- Prediction equations based on personality facets, HICs or small numbers of items are inherently unstable. They may very accurately describe current incumbents in a role, but are poor predictors of performance of fresh job applicants. Using complete and coherent scales of reliable personality tests is a much better bet.
- The over bearing professionalism and science presumed in personality profiling predictions can undermine confidence and personal responsibility in decision makers. Personality assessments cannot make decisions for people. On the other hand, if decision makers can know ‘what people are like’, they can make their own informed selection and team building decisions. They can learn from their successes and from their mistakes and develop their own insights.
- My argument is that we have become too numbers and prediction oriented and undervalue understanding. Finely focused behavioural predictions of performance are unrealistic for many reasons. The best we can do at the moment is to make inferences about the potential effects of behavioural dispositions associated with well researched personality dimensions of the Five Factor Model and we can develop our understanding of the ways these ‘primary colours of personality’ can be mixed to better reflect the complexity of personality as we encounter it.
For professional psychologists, the quest is to understand human nature. There are advances almost daily in the sister professions of neuroscience, evolutionary psychology, anthropology, genetics, philosophy and the social sciences that cast new light on human nature and contribute to the pool of knowledge. All this places psychology in an axial position;“Personality psychology is about the nature of human nature. How you think about human nature frames all subsequent discussions“, Bob Hogan, 2014, ‘The Science of Personality” documentary. Rather than reifying the spurious precision of numbers, why not be the brokers of insightful psychology?