The European Association of Work and Organisational Psychology





Do the behaviours observed and recorded by a test, an interview or another technique reflect real life behaviours ? Ecological validity is a measure of how far this is the case.


To take assessment centres as an example: candidates might be taken to a hotel, observed and under some pressure, measured on their ability to take decisions, prioritise work, contribute to a team and generate ideas. To what extent can you say that they’ll show these behaviours in an office with different kinds of pressures, no observers and many more factors vying for attention ?


The more lifelike you make a measurement tool the less likely you are to be able to control other factors and get accurate scores. The less lifelike it is the more you have to question whether the accurate scores are getting at something useful in real life situations.




The European Federation of Psychological Associations.




Emotional Intelligence or EI became “ flavour of the month” when Howard Gardner published his theory of multiple intelligences starting in 1975. He suggested that rather than just one intelligence there were many, ranging from musical to emotional. More recently we’ve seen books on spiritual, social and even sexual intelligence. 


Gardner’s ideas struck a chord.


We all know of academically brilliant people who are useless at other tasks. We’ve read about people who were prodigies in music or maths but seemed somehow undeveloped in other areas of human activity. And most of us have met people who were “ good with people” – who seemed to understand where the other person was ‘coming from’ and reacted appropriately. In other words, they were emotionally intelligent. Some of us have had managers who were exactly the opposite !


Emotional intelligence became a key issue, particularly in management and leadership selection and development. Research suggested that the higher you go in an organisation the more important emotional intelligence becomes.


We’ve described EI as flavour of the month but it now seems embedded in many large organisations. Lots of EI measures have been published and there are many different definitions of what it is. In fact this is one of the basic criticisms of the area: there are so many different definitions and measurement approaches that the concept seems elusive.




An underlying “ construct” that’s being measured in a Personality test. Big Five measures ( see above ) unsurprisingly measures five of them, the Sixteen Personality Factor Questionnaire ( surprise, surprise ) measures 16 primary factors which can be translated into 5 global factors.


Factors are isolated by a complex statistical technique called factor analysis which shows whether a number of questions are addressing the same construct – which might be something like Open-Mindedness, Enthusiasm, or Team Orientation.


In essence. the concepts of factors and factor-analysis are very important for people constructing tests and the results of the process, reported in a manual, give you a good idea about how stringently the test was created.




What a test appears to measure to the test taker. How relevant a test appears to be may well affect the test takers responses: if for instance he or she sees it as irrelevant psychobabble they might fill it in without care or even randomly. That’s why a good administrator, introducing the test session carefully is so important.


Given that so much work takes place on computer screens nowadays, the use computer based testing may lead to much more face valid tests. Instead of rather old-fashioned black and white printed diagrams, we might use moving, realistic and colourful pictures to measure spatial reasoning.




This is a big issue in testing  because of the growth of testing over the internet, which tempts users not to give feedback to the candidates.  We can do no more than skim the surface. 


Feedback is the process of explaining test interpretations to the candidate after they’ve taken the test. Level A and B training stress the importance of this, for a variety of reasons:


  • in some development work, discussing test results is the start of the process


  • it’s simply fair. Candidates have committed themselves to answering fairly searching questions honestly. They deserve something out of the process


  • it gives a good impression. Research shows that organisations who give feedback even to failed job candidates are viewed as professional. Candidates value the advice and feedback given.


Recent developments are causing debates in this area:


  • the internet allows you to administer tests distantly. This sometimes makes face-to-face feedback difficult.


  • technology, on the other hand, helps make written feedback easier via e-mail ; not least because programmes can create different sorts of report available from the same scores. Some systems create technical reports for expert test users and simpler, summary reports for candidates.


  • web recruiting systems generate more applicants. The use of tests for “ pre-sifting” means that many more people might take a test. The cost of individual feedback looks prohibitive.


  • as software automatically undertakes basic testing tasks – scoring an writing expert interpretations – good feedback  becomes the core of a test users tasks.


  • video conferencing feedback seems to work quite well.


Feedback skills are generalisable to management and coaching situations – indeed, we believe they’re a core skill of any manager or supervisor.




As opposed to crystallised intelligence fluid intelligence is seen as independent of learning and includes skills like problem-solving, pattern-recognition, learning and memory. It enables you to make sense of a situation.




In testing, frequency distribution is the number of people that obtained each score on a test. If you were using a test to identify people with particularly good Maths skills in order to train them as accountants, it would be no use to you if every single candidate got the top mark. The test would seem to be too easy for the purpose and isn’t DISCRIMINATING between candidates.