Peter Saville, BA, MPhil, FBPsS, CPsychol, CSci, Academic FCIPD, FRSA
Professor Peter Saville is one of the world's foremost authorities in industrial psychology and occupational psychometric test publishing. He currently serves as the International Chairman of Saville Consulting, where he remains heavily involved in new product development and training. Founded by Peter in 2004, Saville Consulting develops measurement tools like the Saville Consulting Wave questionnaires allowing businesses to identify prospective employees who are a good fit with their workplace culture.
Peter obtained his Honours Degree in Psychology from Leicester University, earning his Master of Philosophy Degree in 1974. In 1977, he received a Doctorate from Brunel University and co-founded SHL, an innovative and wildly successful developer and publisher of psychometric tests. Peter severed all links with SHL in 2004 following a major disagreement over board policy. Peter was particularly concerned with new product development and the use of unsupervised tests delivered via the Internet, which he believed would lead to cheating.
Peter has spoken at many international conferences and has been frequently interviewed on TV and the radio. He has published over 200 academic papers, psychometric tests, and books. Saville Consulting currently operates in over 80 countries worldwide.
What inspired you to get interested in testing and make it the focus of your career?
From university I joined the National Foundation for Educational Research (NFER), which had a test division distributing virtually all of the American and British psychometric tests. I was tasked by the Director of the foundation with standardizing and adapting many of the American tests used in the UK such as the Wechsler Scales and the 16PF. I have been involved not only in testing but in talent management ever since.
What have been the most enjoyable areas of your career in testing? What do you most enjoy thinking about?
The most enjoyable part of my own career in testing was forming my own company, SHL, and now forming my second company, Saville Consulting, which enjoyed a 47% growth last year. I most enjoy thinking about innovations in psychometrics and putting them into practice, particularly working with a team of young psychologists who often know more than I do about a given area!
Please tell us about your vision and inspiration behind your founding of SHL and later Saville Consulting?
Having worked at NFER and been promoted to Chief Psychologist, I could see that many tests were old, clinical and inappropriate for the world of work. I thought that if we developed tests that were job relevant and better designed, an organisation would be able to attract more able candidates. The other issue was that at NFER there was up to three month wait for orders for tests, which I could see had the potential to lead to photocopying infringement. If I could provide top rate customer service and supply a test within 24 hours, then we would have an undoubted advantage. I was employed on a university type contract and paid to give talks to people in industry. Running courses for Shell, BP and ICI meant I already had contacts in companies that would help norm and validate the new tests I created.
How have you gone about helping to make Human Resource departments across the world more fair and objective with their assessments?
There are many jobs where applicants, especially young people, are not experienced in being interviewed but have immense talent - we could take the example of an applicant engineer who may interview badly but have excellent mechanical and spatial skills important for the role. I believe that using tests will reveal talent that might otherwise be dismissed because the interviewee lacks experience of the interview process. It used to be the case that selection decisions for Air Traffic Controllers were based on board interviews alone, however, one progressive thinking HR Director decided to pursue the possibility that psychometric tests and work simulations might improve the success rate of applicants getting through a 4 year, costly, course. We increased the success rate in training from 50% to 85%. This was a massive saving for the organisation but also meant that the best applicants were getting the jobs. I personally feel this is important not only for the organisation, but also the applicant and society in general. In some countries for example, nepotism is used for selection and I firmly believe this can lead to lack of growth of those economies. Only by getting the most able will we grow our economies, especially in current times, so objective selection methods are an immensely important process in helping HR departments to be more fair and objective.
What has been the biggest challenge of your work and how did you overcome it?
I was faced with the problem of standardizing a major personality inventory on a representative sample of 2000 people from the British general population ages 16-65. When I went to organisations which could administer questionnaires to people in their own homes across the country I was faced with enormous costs as each interview would take some 2 hours. I therefore came up with the idea of approaching a market research company, which already had a large field force of interviewers, and training them in administration of the questionnaire. This, along with a sampling method known as random location sampling would then enable the collection of such data. To cover costs I then approached a number of commercial organisations and suggested that they might be interested in knowing the personality characteristics of people that bought their products. So a large pharmaceutical company sponsored the research to discover whether people who buy health foods differed in personality from the general population and a major motoring manufacturer also contributed to research costs in return for us providing the personality profiles of people who not just bought their cars but also information relating to which styles of cars they bought. By this method we gained our representative sample of 2000 people at zero cost to ourselves.
Which new testing technologies have you most excited lately? Which testing patents have the most influence?
What has excited me most is the realization that although certain personality characteristics may go together in the population as a whole, at the individual level you can often find wide disparities. For example although being conceptual and innovative in approach normally correlate highly, an individual like Sir Richard Branson, when tested, had a low score on the former and high sore on the latter. This unique feature to show where and individual goes against the general trend was introduced by Rab MacIver in our new suite of questionnaires we have called Wave.
What are the most important elements to consider when designing and developing secure high stakes tests and their supporting administration, scoring and other systems?
One of the most important elements to consider when designing and developing tests for selection is of course the matter of security. It is for that reason that we have designed internet based tests, openly accessible to people in their offices and homes, but have also produced numerous parallel forms, which candidates can be asked to take if selected in to check the accuracy of their scores. Even Mensa, when sending out a test by post, would insist that a high scoring applicant came to a controlled testing centre to complete a second test to ensure the fist test hadn’t been taken by someone else! Cheating on the internet will only get bigger and we have ensured that our online tests have far greater security than are true of others.
Reflecting on the lessons learned during your career, what do you now believe is the primary value of testing and credentialing in our society? How has your viewpoint evolved though the years?
In his books Sovereign of the Seas, which discusses the history of the British Navy, Howarth points out that Samuel Pepys, when Clarke of the Acts, introduced a test of navigation ability when selecting individuals for promotion to Ship’s Captain. Howarth believes one of the most important factors in the success of the Navy over next 200 years was that whatever a person’s background, only able people made it to Captain. People of quite humble birth but with immense talent were promoted to captain a Navy ship in their early twenties. What’s more, able Captains tend to appoint other able Captains, so one gets a spiral of competence. When those in charge are of low ability, they then appoint people of even lower ability, creating a spiral of incompetence. So I believe that testing, with clear competency analysis, a good structured interview and good feedback can make an incredible impact upon society in terms of fairness and economic advances.
Psychometric Education and Careers
For people considering a career in psychometrics, what are important qualities, skills or talents they should possess to be successful?
For those considering a career in psychometrics there are many types of people who can succeed. Some, like R. B. Cattell, will be in love with data in the back room. Others will wish to use psychometric tests as part of talent management programme and, therefore, will be much more interested in influencing people rather than manipulating data. It is difficult to generalize. Certainly, meta-analysis, when one combines samples from different studies into one large data set, has shown that intelligence is an important characteristic, as is conscientiousness. Focus, determination, achievement and drive are important qualities for success in psychometrics.
To help people considering the field of testing and measurement, what are the specific fields of testing and testing industry sectors?
The main field of testing and measurement sectors are Industrial and Organisational Psychology, Clinical Psychology, Educational and Child Psychology, Sport and Exercise Psychology, Forensic Psychology and Health Psychology all of which benefit from the addition of more objective psychometric questionnaires.
The History and Future of Testing
What do you estimate is the current size of the testing industry in the UK? What growth rate do you estimate over the next 10 years?
Of course we do not merely work in the UK but operate in some 80 countries worldwide through a network of over 40 partners. So this year the worldwide usage of Saville Consulting’s tests will be greater than that of just the UK. To estimate the size of the UK testing industry would have been simple 10 -- 20 years ago but it is a lot more difficult now. In an increasingly global market it is really quite difficult to estimate exactly which companies buy tests purely for UK usage. Most of our clients are those like Jaguar Land Rover, which we increasingly work with on a global basis in as many as 40 countries. All I can say in relation to the UK is that testing is popular and very well used. In one survey of the UK FTSE 100 companies, 80% indicated that they used psychometric tests somewhere within an organisation.
How many tests does the average organization require of employee candidates prior to hiring in the UK? What are the most popular of these tests? How many and which tests would you recommend?
There used to be a time when, at senior management levels, a whole day of testing would take place before a person was offered a post. There is now, however, a trend towards quicker, slicker testing. A combination of two cognitive tests of verbal and numerical critical reasoning and a personality questionnaire of relatively short duration is much more likely today. Thanks to advances in psychometric theory and, of course, internet delivery, the speed to hire has been greatly increased. I think the movement towards shorter but still highly valid assessments also increases candidate acceptability, something about which we should, quite rightly, be concerned. Good candidates are likely to have more than one choice, so in ensure you get to the best candidates, acceptability is important. We are seeing enormous growth in our Swift Aptitude range, which can take under 10 minutes to complete with no loss but indeed often improvement in reliability and validity.
In addition, our portfolio of Wave personality questionnaires, which can reliably measure 36 variables of personality relevant to the world of work in as little as 15 minutes, is also ever growing in popularity. In fact, in a recent major research project which has been independently reviewed, the Wave Professional Styles questionnaire was shown to be more valid than well- known instruments such as the 16PF and the OPQ. We do, however, have the advantage that some of these other instruments are up to 50 years old. Whilst Wave was designed specifically for internet rather than being an original pencil and paper form put up on the internet. We also know a great deal more now about the criterion space that is predictive of job performance than I knew in 1984 when I developed the OPQ. We know both more about what should be measuring and what we should be predicting.
How would you rank the top five countries in the world based on their use of quality assessments? Where are some of the best opportunities to help?
I would find it difficult to rank the top five countries in the world. Certainly, the most important areas are the USA and the UK, who have led the field in psychometrics since early days and still do today. Rapidly expanding economies and also poorer parts of world, which need to grow their economies, could benefit hugely from objective testing rather than using corruption and nepotism as is currently the case in some countries.
What are some of the most important tests in history?
The Woodworth Personal Data Sheet: This was the first personality questionnaire intended to screen out potentially neurotic soldiers from fighting on the front line in the First World War because of the enormous potential for disruption to those fighting around them an individual with such affliction may present. Peculiarly enough, however, this was never used in practice. Sir Francis Galton was actually thought to have developed the first questionnaire method in his laboratory in Kensington in London in the 1880s.
Stanford-Binet Intelligence Scales: Binet was instructed by the French government to develop a test to help screen out children who were unable to cope with the standard school curriculum in France. He developed the first test of mental ability. The test was translated into US English at Stanford University and was renamed Stanford-Binet Intelligence Test.
Cattell's 16PF: This was designed around a factor structure proposed to be the building blocks of all personality. It was extensively used in educational, clinical and occupational psychology for many years, though no-one has managed to replicate Cattell’s factor structure and for this reason the belief that he had discovered the basic variables of personality is untenable.
Minnesota Multiphasic Personality Inventory (MMPI): This was developed for clinical use and became the world’s most widely used psychiatric diagnoses tools. Unfortunately it was also adopted by practitioners in the industrial field and has since come under tremendous attack by lawyers due to some of its question content being regarded as a invasion of privacy -- as a result it was heavily sued. The MMPI was a very important tool designed for the clinical setting and should have stayed within the field for whose use it was intended.
Who are five of the most influential test authors in history?
Sir Francis Galton
Robert S. Woodworth
If you had ultimate power to change the world of assessment, what would be the first three things you would do?
I would ensure very full and adequate training in the basic statistics required in testing, which are now overlooked on many courses. I would also encourage people to have a clear objective of what the purpose of testing is -- in terms of what you’re trying to measure. And finally I would make sure that candidates are given meaningful feedback on tests scores and their performance, which will help our reputation as psychometricians and testers within the general society.
How do you envision the future of testing?
Predicting the future, especially in current times can be a pretty scary thing. A great deal has been written of course in academic literature about the Big Five, developed through the technique of factor analysis, which seeks to find the most parsimonious solution in a set of data. It would seem that the variables of Openness, Conscientiousness, Extraversion, Agreeableness and Emotional Stability have wide applicability across most cultures so far studied. However, there is increasing evidence that whilst they may be parsimonious, they are not necessarily the most predictive taxonomy for measuring performance at work. So I would envisage more work in the future on personality assessment looking at particular facets which make up the Big Five to improve our understanding of people at work. My colleague Rab MacIver has made tremendous headway with this by breaking down the facets of personality into two further concepts -- motive and talent. By measuring both we are able to observe disparities when they occur, providing us with a greater depth of understanding of the way people behave at work.
We also have the problem of exactly how we measure success -- what exactly is it? The work of my colleague Dr Rainer Kurz has found there to be three factors of effectiveness at work -- Demonstrating Capability, Working Together and Promoting Change. The same pattern can be observed in different types of leadership -- Task, People and Growth. My colleague Tom Hopton has found that leadership relating to Growth (that promoting change, innovation and entrepreneurialism) is the most powerful.
I hope to see much more work like this in the future.