Recruitment and selection involves making predictions about future behaviour so that decisions can be made as to who will be more suitable for the job (Renwick, 2001) Both approaches appear logical and rational, however in today’s world this proves not always to be the case (Buchanan and Huczynski, 2004.) Selection can be described as selecting the best candidate, the individual with all the specified essential characteristics, and selecting them for the job (Newell, 2006.) There are a number of different methods used for selection by human resources, each with a different predictive validity on the predictive accuracy model. When an organisation is deciding which method of selection to use they may consider the most cost effective (Beardwell and Holden, 2001) or rely on predictive validity.
Essentially there are two kinds of tests used for selection assessment – cognitive and personality tests. Cognitive tests, tests of general intelligence have high predictive validity across a wide range of jobs (Hunter and Hunter, 1984) There has been a well-documented trend towards greater use in the UK of testing for selecting using psychometric tests. (Shackleton and Newell, 1991.) Study by Wolf and Jenkins (2006) suggested the most important factor explaining a rise in psychometric test use is the change in the legal environment, which increased risks on making hiring decisions in appropriate ways. This method is shown to be a safeguard against accusations of unfair practices. Personality tests had typically been found to have a low validity for predicting job performance (Ghiselli, 1973) due to the ‘fakeability’ in such tests. However, more recently, evidence has been established to suggest that personality measure can be valid predictors of job performance (Day and Silverman, 1989.) One reason for the early pessimism was that there was no generally accepted model of personality. More recently the ‘Big 5’ (Wiggins, 1996) have emerged which include the following personality traits; openness to experience, conscientiousness, extraversion, agreeableness and neuroticism. Theses have been supported and agreed on by many. Salgado 2002 found that conscientiousness and emotional stability had the highest validity for job performance and ‘openness to experience’ was valid for training proficiency. Tett et al 2003 demonstrated that a broad measure of personality obscured potentially useful linkages between specific aspects of personality and performance. The best way is to focus on specific personality traits and specific measures of performance in order to improve predictive validity. Despite doubts about overreliance on personality tests with respect to their use in predicting future performance Ones et al 2007 believed them to be a valid predictor of future performance , even in complex tasks such as management. Ceci and Williams (2000) suggest the measurement of intelligence does have drawbacks if such measurement is based on the assumption of intelligence as a fixed property of individuals. They argue intelligent behaviour such as complex thinking is strongly linked to the setting, task, location and other people involved. Jaguar is one of the largest companies which use psychometric testing as a measure of independence of thought, team working and co-cooperativeness.
The interview is the most widely utilized method for the selection of human resources. (Shackleton and Newell, 1997.) Despite concerns over its validity (Huffcutt and Arthur, 1994) meta analytic research suggests that the interview can be a selection tool with substantial predictive validity. However, this is only the case when properly conducted with structure and adequately trained interviewers (Jelf, 1999) Despite having a predictive accuracy of 0.62, found in the Predictive Accuracy Model (Beardwell and Holden, 2001) the interview has received continued popularity, as people are more familiar with them. The interview is described as a controlled conversation with a purpose (Torrington and Hall, 1995.) Interviews are usually conducted face to face but also telephone interviews can be used. There are two forms of interviews; structured and unstructured. Traditionally, unstructured interviews had low validity (Mayfield, 1964) Structured interviews were then developed to improve validity so at least then all candidates were interviewed with the same questions and used the same dimensions to assess. Most firms tend to use a combination of structured and unstructured interviewing techniques as this is more effective. Unstructured interviews are more flexible and allow for probing of candidates responses further and evaluating their personality attributes such as soft skills and energy levels. This is found in line with Blackman (2002). Unstructured interviews can help in bringing out the personality of a candidate and to determine his fit within the organisation. However, the drawbacks of using completely unstructured interviews have been highlighted in literature (Anderson, 1992) Structured interviews were particularly useful for evaluating lower levels and unstructured interviews became important as the levels increased, especially in the case of middle and senior managers. The main varieties of structured interviews are behavioural and situational. Behavioural interviews are where the interviewer will ask questions about past behaviour in order to predict future behaviour. Situational interviews test thought processes and logical thinking by asking the candidate how they would respond to a hypothetical situation. Behavioural interviewing tends to be more popular than situational interviewing. Empirical data from UK firms highlights these advantages of behavioural interviews (Barclay, 2001) It is argued that selection interviews can be understood as a process that involves ‘political and power games (Bozionelos, 2005.) It is argued that important decisions on organisations are influenced by political motives rather than rationale or merit (Ferris & King, 1991) All stages in the selection process can serve as ‘arenas for political games.’ Unlike other selection methods such as references, biodata and psychometric tests, access to information from interviews is extremely limited to those who have not been present; the interview is regarded as an ‘opaque process’ to outsiders (Carlson et al. 1971) Despite all of this, and fact that interviews’ predictive accuracy is below that of assessment centres, it has been found that interviews and work samples were rated most positively over all (Kravitz et al 1996)
Biodata attempts to capture directly the past behaviour of a person and use this as the basis for predicting future behaviour. Behling and Eckel (1991) model assumes that managers make decisions by relying on past experiences and their general sense of the situation. Applicant forms, commonly found online collect biodata (Hill & Barber 2005) Since they are based on actual behaviour, the idea is that they are less prone to misinterpretation, resistance and distortion (Stricker and Rock, 1998) Research has shown biographical measures can produce good levels of predictive validity (Mumford and Stokes, 1992) Townley (1992) claims use of biodata is a way for an organisation to acquire a compliant, non-unionised workforce. However, there are concerns about the appropriate item content and the test construction methods (Mael, 1991) In terms of content many biographical measures contain questions, which are indistinguishable from items on an attitude scale. In terms of construction, a key problem is that because most measures are empirically keyed to predict particular criteria, they cannot be used in different settings (Furnham 1997) It can also produce a very homogenous workforce and this produces a problem with the ability to adapt to change, which is important. Nevertheless, more recently research has shown that biodata can be designed to predict performance in a variety of organisational settings. (Carlson et al., 1999)
Assessment Centres / Development Centres
The use of assessment centres is growing in the UK, 65% of firms use them (IRS, 1997) It is a ‘process which consists of a small group of participants who undertake a series of tests and exercises under observation with a view to assessment of their skills, suitability for particular roles and their potential for development.’ (Fowler, 1992) Ratings are made by a small group of trained assessors who observes the different exercises. Although there is no such things as a typical assessment centre, (Spychalski et al., 1997), exercises may include group decision making exercises, presentations, interviews, role-plays, psychometric tests etc. Mazda USA uses applicant form, aptitude tests, personal interviews and group problem solving exercises with simulated work exercises (Raymond and Stone, 2005) Evidence shows they have a high level of predictive validity (Hough and Oswald, 2000) however are not widely used in organisations which could be due to their poor construct validity (Goldstein et al. 1998) Assessment centres are costly to implement and take considerable time. The information-processing load on assessors is sometimes too high and behaviours may be misinterpreted. (Thornton, 1992) Other methods may be more useful and less costly (Robertson and Smith, 2001) Hennessy et al. (1998) attempted to resolve this problem with the use of a behavioural coding method of assessment. This has the advantage that information processing demands are reduced and also limits the extent to which personal schema will direct the observation.
A major consequence of rapid growth of the Internet and its increasing accessibility is that increasing numbers of organisations are recruiting and selecting applicants for jobs online. 18million people annually post their details on monster.com and vacancies are wide ranging across all sectors and not just in IT. Increasingly, companies using the Internet for advertising job vacancy information have the advantage of reaching a global audience. Applicants speed up the process of applying by submitting their CV online. Alliance and Leicester are a prime example of a company who accept CV’s online. Job seekers are increasingly expecting to find work through the Internet rather than by traditional means. (Bartram 1997) The internet is a simple and efficient method of storing and accessing the information. Firms are actually now using the Internet for actual selection and can conduct Internet interviews which are very useful for overseas applicants (Bartram 1997) Computer based testing can be done virtually (Anderson, 2003) Companies have the advantage of cutting costs, improving corporate image, reducing administration and shortening the recruitment cycle (Feldman and Klaus, 2002) Online assessment, referred to as E-assessment, provides organisations with the ability to test at any time and any place in the world with the added benefit of quick processing. BA are a prime example of a company who moved to internet and email recruitment where the number of applicants for their management programme was reduced from 12000 to 5000 (Mirrick, 2001) However, both companies and many candidates found that it created many problems. In terms of the company there can be numerous unsuitable applications, technical problems and not enough information about the job or the company itself shown. Candidates complain that there are not enough relevant jobs online and of the impersonal nature of the process, the slow feedback they receive and the concerns they have regarding the security of personal information. There is always the risk that they could be put off the job by people who think that the website is poor. Worryingly, the IRS Survey 2005 did not feel it led to better quality candidates and sometimes found that it reduced quality by up to 40%.
Other Selection Methods.
Less conventional methods such as physiognomy, phrenology, body language, graphology and astrology have been suggested as possible selection methods. There is little evidence to suggest that they can be used effectively. Thatcher (1997) suggests that 9% of small firms and only 5% of large firms in the UK use graphology as a selection method. However, Fowler (1990) suggested the extent of use of graphology is much higher in the UK that reports show as there is some reluctance on the part of organisations to admit that they are using graphology as selection tool. There are concerns about the quality of graphologists who can set themselves up with no training whatsoever.
References are in most selection procedures, however the predictive accuracy is only 0.13 References are used to obtain the predictive information about candidates from previous employees, academic tutors or somebody who knows them. Despite the low predictive validity score, an IRS Survey (1991) found that they were used by 97% of organisations, coupled with interviews.
The newer assessment methods appear to offer significant benefits since they can improve objectivity and criterion related validity. However, as research into actual practices suggests, in reality, selection decisions continue to be dominated by more subjective approaches. (Anderson, 2003) Although it is possible to see that assessment centres have the highest Predictive Validity of 0.68 they are not the most used technique of selection. Interviews are most commonly used, followed by references. The actual result and validity of a selection method is in the hands of the user. Assessment centres are only valid if carried out properly. It is more likely that companies will use assessment centres in selecting employees for more senior positions, as the costs of doing so will be quite high. ‘Most significant decisions are made by judgement, rather than by a defined prescriptive model.’ (Bazerman 1994) While psychometric tests can potentially add valid information to the selection decision, it is clear that this will only occur if tests are used appropriately (Newell and Shackleton, 1993) It is essential to use a variety of selection methods, with regards to the job in question, in order to make the best predictions about future behaviour, so that decisions can be made to get the best candidate for the job.