Human Resource Solutions

Selection Methods 

There are a wide variety of selection methods used in the UK. The following are the main ones - for each of these we have briefly described them and given an idea of their effectiveness through a 'predictive validity' score (see blue side-bar on right).

Standard "unstructured" selection interviews - where interviewers ask what they think are appropriate questions, and have no real criteria for assessing answers. Often appointments are made (subconsciously) on the basis of "gut feeling", or liking the individual. (Predictive validity = 0.2).

References - Candidates are unlikely to give someone as a referee who will give them a bad reference! Only reference of any real value is from their immediate line manager of their present employer. (Predictive validity = 0.15).

Graphology - This is the study of an individuals handwriting to assess aspects of their personality; this is then extrapolated to predict their work performance. Although quite widely used on the continent, it is seldom used in the UK. (Predictive validity = 0 to 0.1).

Biodata - This method consists of analysing certain data from individuals that are high performers. This "biodata" can include things like age, leisure interests, education, exam results etc. The fundamental idea being that we try to find a commonality of factors among high performers which we can then use to compare against applicants for the same job. For example if we find that a significant number of high performers have achieved good results in English exams at school, this could form one of our criteria for screening applicants. (Predictive validity = 0.35 when good consistent criteria are found which can be used).

Personality Questionnaires - Attempt to compare the individuals personality factors with the population (at large, or a specific subset of it e.g. non-manual female workers etc.). This allows us to predict with a reasonable degree of accuracy where an individuals particular personality traits lie in comparison to these groups. It also allows us to compare candidates against each other. We can then extrapolate to likely impact on work performance. (Predictive validity = 0.25 - 0.40 effectiveness depends on using the correct tests in the right way).

Work-Simulation Tests - for example a typing test for secretaries, or an in-tray test for management posts. These can be very effective in ascertaining an individuals ability at particular tasks. However it is vital that the test used is a reasonable simulation and that their is an effective method of scoring the candidates for comparison purposes. (Predictive validity = up to 0.55 if the test is properly validated).

Aptitude / Ability Tests - these are usually paper or computer based tests which allow us to examine an individuals particular aptitude of ability in a particular area we are interested in, for example numeracy, abstract reasoning, mechanical aptitude. Also included here are general intelligence tests, which although seldom used for selection purposes can still be a useful indicator of likely success in particular jobs. (Predictive validity = 0.4 to 0.45).

Structured Interview - once the specific competencies we are looking for have been identified, we can develop questions which elicit information from the candidate about the extent they meet these requirements. Usually the questions are "behaviourally" based and ask the candidate to give examples from their career of examples of the competencies you are looking at. Each candidate is asked the same questions, and their answers rated by the interview panel. (Predictive validity can be up to 0.6).

Selection methods as described above are often most effective when used together, and often combined with such things as a candidate presentations, observed group exercise etc. When pulled together in a coordinated way, this is called an Assessment Centre. The focus in a AC is always on behaviour defined in terms of the competencies identified which distinguish high performers. Various exercises / tests are used to capture and simulate the key behavioural aspects of the job in question. Assessors rate the candidates using a scoring system for each competence.

 

© Human Resource Solutions 2000

Predictive validity measures how effective a selection method is in predicting future job performance. Values are as follows:

greater than 0.5 = excellent

0.4 - 0.49 = good

0.3 - 0.39 = acceptable

less than 0.3 = poor