Můžeme stále věřit psychodiagnostice?

Can we still trust psychometrics?

Reflecting on the rise and fall of Paul Flowers, who it is alleged was selected largely on the basis of his performance on psychometric tests, OPP’s CEO, Penny Moyle, argues for best practice - putting psychometric tools in the right context and avoiding the facile conclusion that a single assessment can ever give you all the answers.

Over recent decades, occupational psychology and psychometric assessment have played an increasingly important role in many organisations. Whether for recruitment, staff development, or some other form oforganisational transformation, companies frequently look to psychology-based interventions for help. And given the value that occupational psychology can bring to business, the Paul Flowers enquiry and its apparently damning indictment of the role of psychometric assessments in selection processes has been disappointing to say the least. It appears from some accounts that a man who looked like God’s gift to the world of banking on the basis of his assessment results, proved by his later actions, that the assessment used was fundamentally flawed. A disastrous news week for anyone using psychometrics in the workplace, you might conclude.

There are two key areas where psychometric assessments appear to have been misrepresented: the implication that the assessments, which Flowers apparently ‘aced’, formed the basis of his selection for the job; and assumptions over what kinds of assessments were involved. One article, in fact, speculated that the personality instrument involved was the Myers-Briggs Type Indicator (MBTI). However, this instrument never was and never will be suitable for selection purposes - there are plenty of personality assessments that arehighly effective as a part of a selection process, but the MBTI is not one of them.

A key point to consider here is that different psychometric tests are developed for different purposes. Some, such as the MBTI, are type-based, and measure personality preferences rather than abilities. A preference for intuition in that particular framework, for instance, indicates that someone likes making connections between different data points and building a vision of the future. It doesn’t tell you whether this person is going to be any good at doing this in order to develop organisational strategy. Clearly then, type-based personality tools don’t address competence for a role, and anyone who tells you otherwise is misinformed. On the other hand, trait-based personality tools can predict behaviour and provide an indication of potential. There are also various ability tests, ranging from measures of specific skills relevant to a role, to general measures of intelligence. The effectiveness of the tool chosen depends on its relevance to the requirements of the job role.

Of course, factors other than personality and ability also need to be taken into account when recruiting, such as a person’s qualifications and experience. Psychometric assessments are not intended to be used in isolation, and the best selection results come from combining multiple methods. If the overall selection process is flawed, no psychometric tool is going to save you from a bad decision. A key challenge for HR, then, is to help recruiting managers synthesise the results from many sources, including the results of any psychometric assessments, to select the best candidate from those available.

The power of a psychometric assessment very much depends on the skill of the person who is interpreting its results - there’s no point buying the fastest car on the market and then criticising it because you haven’t learned to drive. What’s more, a personality assessment is completely irrelevant if the candidate doesn’t have the basic experience, sector-insights and technical expertise needed for the job. And although there are assessments designed to detect misdemeanours such as drug taking or other social deviance, it is misguided to think that all personality tests will provide such insights. If these issues are important to an organisation, recruiters should carry out a background search, and/or use one of the tools available to measure such propensities. Again, it’s all about using the right tool for the job.

Trait tools, as used in selection, have over 50 years of independent research comparing assessment scores with actual job-performance (eg Smith & Robertson). The results provide scientific evidence that these tools can indeed work, when correctly applied (and, just as importantly, that methods such as graphology and unstructured interviews are more or less useless). It is for this reason, rather than some lingering fad, that over half of all organisations use ability tests or personality questionnaires to select their staff (source: CIPD). Above all, the sorry tale of Paul Flowers highlights the need for organisations to establish valid selection processes at all levels. Recruitment for executive roles should be at least as precise, ethical and informed a process as any other. HR has a responsibility in all this to make sure that recruiting managers, whether board members or otherwise, understand the benefits and limitations of the assessment methods at their disposal, and to safeguard their effective and professional use for the benefit of everyone concerned.

Tento článek původně vyšel na webu: http://www.thehrdirector.com/features/recruitment/can-we-still-trust-psychometrics/