Focus of attention

The key point, though, is that fairness to the candidates and the final appointee is increased. Candidates are assessed in more than one situation, against pre-determined, objective criteria, and by more than one assessor.

Designing and running an assessment centre is a serious undertaking, requiring considerable investment in time, resources and - in all but the largest organisations with their in-house experts - the input of a suitably experienced and qualified external adviser. The assessors, all or most of whom are usually managers from the host organisation, have to be trained and then must find the time to run the centre. The organisation is asking a great deal of the candidates themselves, as well as of its own managers, and should be prepared to offer successful and unsuccessful candidates alike in-depth feedback on their performance. Knowing that feedback will be offered can also benefit the organisation. If assessors know that their judgments will be passed on to the candidate, this can help maintain the discipline of ensuring that assessments are as objective as possible.

Attempts to short-circuit any of these steps not only undermine the effectiveness of the centre - in effect, the investment will be wasted because poor appointment decisions will be the result - but can also inflict psychological damage on the participants. An assessment centre is an intense experience for candidates, requiring a great deal of commitment and energy. If it is mishandled, the reputation of the organisation will be undermined.

Guidelines

At present, there are no best-practice recommendations or code of conduct governing the general use of assessment centres in the UK. In 1989, the USA's Task Force on Assessment Center Guidelines produced a series of standards for that country (reproduced in Charles Woodruffe's Development and assessment centres¹ - a standard text, now in its third edition). In the UK, the British Psychological Society's working party is still formulating its proposals, although the Association of Graduate Recruiters has published a briefing on the use of centres in graduate recruitment². However, most British researchers writing on the subject have set out what they see as the key issues and a synthesis of these points is shown in box 1.

Extending and evolving

Assessment centres have been in use since the Second World War; however, only in the past two decades has their use taken off.

According to Paul Iles, writing in 1991, only 7% of British employers used centres in 1973, rising to 19% by 1984³. More recently, the best evidence of their use comes from the annual recruitment survey conducted for the Chartered Institute of Personnel and Development (CIPD). Using matched samples of employers, last year's survey found that usage had risen from 29.1% of employers with 50 or more staff in 1998 to 33.9% in 2000⁴.

However, the resources required to conduct assessment centres mean that they are mainly found in larger organisations and, even then, are usually reserved for the recruitment of managers, professionals and new graduates. The CIPD's 2002 survey (to be published in May 2002) will, for the first time, analyse its findings by workforce size and is likely to confirm that most of the largest employers in the UK have adopted assessment centres.

Certainly, the use of assessment centres for the recruitment of new graduates is now well established. Our latest annual survey of graduate recruitment, conducted last September, found that more than half of all organisations that recruit new graduates use centres (57%), and that more than three-quarters of these users (79%) rate it as their most effective means of finding suitable appointees. This makes it by far the most effective selection method, according to the 108 employers that we contacted (see Graduate recruitment 2001/02). In two of the case-study employers included in this article, GCHQ and Britvic, new graduate recruitment represents a major use of these organisations' assessment centres.

Not only have assessment centres been taken up by increasing numbers of employers, but the organisations using them have been putting them to a wider range of purposes. They have expanded to include the assessment of development needs of existing staff, and in the selection of team leaders (see the Pret a Manger case study in this article), of skilled manual workers and other non-managerial grades. The Trades Union Congress has used them to recruit trainee organisers to spearhead its campaign to increase union membership⁵.

But assessment centres have changed in other ways. This has come about partly because some employers have become less willing to take the claims of assessment centres' powers of prediction at face value, and have been investigating and modifying their practices.

Following a series of research studies that have found comparatively high rates of success in centres' ability to identify the best candidates, the pendulum has swung back somewhat. Several experts in the field have argued that such studies have unintentionally "cherry-picked" the best - those that have been well-designed, conducted and evaluated - and have created an artificially high expectation that all assessment centres will be equally effective. For example, Clive Fletcher has observed that: "There is a real danger that some so-called assessment centres are less effective, and possibly less fair, than simpler assessment methods."⁶

Work focus

In one notable trend, employers are adapting their assessment centres to build in greater work relevance; they know that close linkages between the centre's design and the requirements of the job help to improve its ability to identify the most suitable applicants.

A study of assessment centres by Sharpe Research for the Cabinet Office⁷, as part of a recent project to improve the Civil Service Fast Stream, found that centres "took place on or near the employer's own premises; it seemed to be important to all recruiters that candidates visited their offices at some point to get a feel for the organisation and its people, its atmosphere and culture". At Pret a Manger, the fast-food retailer and one of our featured case studies, each external assessment centre begins at 6.30am in one of the company's shops, where candidates "shadow" a manager or team leader for four hours.

Similarly, many employers have taken steps to ensure that the exercises and interviews in their centres more closely reflect real life in the organisation. They are checking that the centres capture the reality of the skills required: both their nature and the context (if interpersonal skills are required, whether they are required in teams or in one-to-one settings, for example). And simulations are being used more extensively and realistically. At the Halifax, a major financial services company, a report in The Times says that the company's external advisers helped it modify its assessment centre into a single simulation: a seven-hour business scenario. The newspaper quoted its group succession planning manager as saying: "We find this simulation-based approach so successful that we're adopting it for middle-manager assessment and graduate selection."⁸

Some organisations have rejected competencies - overarching definitions of the way that work is approached and performed - as the basis of their centres' design, in favour of task-based definitions, which they use to create the activities and to provide assessors with observation and assessment guides. This switch is controversial, and some experts argue that the task-based approach is not well-suited to many of the more complex and demanding types of vacancy for which centres are typically used.

Sharper design

The design of assessment centres is coming under scrutiny in broader ways, too. Centres that consist of a series of self-contained exercises lose some of their ability to provide a realistic insight into the job and candidates' potential performance in it, and can be disruptive for participants. Consequently, efforts are being made to ensure that each exercise flows into the next, and, where simulations or scenarios are used, that each exercise uses the same context - the same business challenge, or fictitious people and events.

There has been a debate about what attributes of a candidate should be assessed: should a centre take account of personality, given that this is much less changeable than an individual's personal competencies or technical skills? Some experts, such as Margaret Dale and Paul Iles, believe not. They have said: "We feel that assessment centres were never designed, nor should they be used, to measure stable personality traits, but situation-specific skills. The multiple exercises used in centres are not simply there to provide multiple opportunities for the measurement of the same stable trait."⁹

Rapid organisational change, and the likelihood that it will continue in the future, has focused some organisations' attention on core values. Given that jobs, roles and the tasks involved are likely to change significantly, it becomes more important to consider the core values underlying the organisation's activities, and to take account of them in the assessment centre. Both employer and candidate stand to gain if their values are shown to coincide.

Some employers are now administering some or all of the psychometric tests they used to include in their centres at a previous stage of the selection process. This makes the best use of the time available in the centre, and still allows the test results to form part of the information used in making assessments.

At Britvic, one of our case studies, the graduate recruitment programme administers tests in two phases. Numerical and verbal reasoning tests are administered in its phase one centre. Then, candidates who go forward to stage two complete a personality questionnaire prior to attending the second centre.

According to research conducted for the Association of Graduate Recruiters just over two years ago, the use in assessment centres of both personality questionnaires and ability tests has shown a "clear decline" - from 61% of users in 1993 to 40% in late 1999, in the case of personality inventories¹⁰.

The research, reported by Iain Ballantyne at the association's 2000 annual conference, also found that exercises with an obvious relevance to the job, and therefore more acceptability to candidates, are increasing in prevalence. Business games have increased in use from 28% to 48% of assessment centres, report-writing from 7% to 62% and role-plays from 5% to 34%.

Attention to candidates

Attention to the face validity of assessment centre exercises highlights another issue that is engaging employers' attention: the competitive environment their recruitment has experienced in recent years. Assessment centres that are seen to make unreasonable demands of candidates mean that applicants may go elsewhere.

The Civil Service investigation into its fast stream programme, mentioned above, showed that many users of assessment centres have been trying to reduce the length of their recruitment process, so that sought-after applicants do not lose interest and take up a job offer from a competitor. The government report recommended that, while there are constraints on greatly reducing the period involved in the Civil Service process, efforts should at least be made to keep in contact with candidates on the progress of their application.

Selection of applicants to be called to a centre also provides opportunities to improve both the centre's effectiveness and the candidate-friendly nature of the recruitment system. Centres are expensive to run, and it is rarely possible or desirable to invite all applicants to attend. As a result, some employers have improved their selection techniques to ensure that this sifting is as objective and effective as possible. Some employers have enhanced their application forms, for example, by including more questions directly linked to the dimensions used in their assessment centres. They have found that such forms not only help selectors to draw up a shortlist of centre participants, but also provide insights that assist candidates to self-select. The information can also be used as part of the centre process in respect of candidates who go forward to this stage of selection.

Pre-selection is also being used by some employers to ensure that the chosen candidates are suited to the assessment centre method. This minimises the risk of a strongly negative reaction by participants.

Applicant selection takes on an added dimension where a centre involves both external and internal applicants. The two groups may not be compatible. Internal candidates have an inside knowledge of the organisation, often of the job involved. They may know how to perform so as to make a favourable impression on the assessors, many of whom will be managers drawn from within the organisation, and could well already be known to the candidates.

Conversely, the organisation will know much more about internal applicants than external ones, and this may not be to their advantage. There is no guarantee that these advantages and disadvantages will cancel themselves out, and create a level playing field across the group of candidates as a whole. This has led some employers, including Pret a Manger, to make special provision for internal applicants - in its case, through a parallel centre separate from the one used for external applicants.

BRITVIC'S GRADUATE RECRUITMENT PROGRAMME

Britvic manufactures soft drinks and employs around 2,700 staff in a number of UK locations. The company runs assessment centres predominantly for recruitment to its young manager graduate and mature graduate training programmes. Recruits to the mature programme must have at least 18 months' relevant work experience and are employed in the operations side of the business - engineering, for example. Graduates on the young manager programme are recruited to business functions, such as sales and marketing, finance and purchasing.

Britvic is a well-known brand; its recruitment programmes attract a great deal of interest, and the company aims to select the highest-calibre graduates to participate in them. Selection is a rigorous and intensive process, with more than 70 candidates being assessed for around six vacancies on its young manager graduate programme annually. The assessment centre process for both programmes involves two stages, with typically half the candidates successfully making it to the second stage.

Young manager assessment centres

Up to 24 candidates a day attend the first stage of Britvic's assessment centre for its young manager programme. This takes place in the attractive setting of an old abbey situated in the grounds of Britvic's Norwich site. HR takes responsibility for assessing this first assessment event, with a ratio of one assessor to two candidates. The exercises are timetabled to ensure that as many different assessors as possible observe the candidates throughout the course of the day.

Candidates attending all assessment centre events are closely observed against Britvic's competency framework. Each exercise is carefully selected to measure candidates against at least three of the company's six competencies. Behaviour relating to each competency is observed on three different occasions during the event.

Throughout the event, candidates tackle verbal and numerical reasoning tests. There are also one-to-one competency-based interviews and two group or "discussion" exercises. Typically, the group exercises require candidates to plan tasks and projects within a clearly defined framework and make decisions affecting specific issues or projects.

An "information event" is also included in the day. Candidates have the opportunity to meet young Britvic managers who are currently employed on the graduate programme. The assessors make a point of not being present for this part of the day. As recruitment manager Alison Cowen explains: "We feel that it is very important that would-be recruits are able to talk openly with existing employees, so that they can ask direct questions about the job and what it is like to work for Britvic."

The final "activity" exercise has a strong fun element to it. Candidates take part in a challenging practical task such as building a structure with limited materials that will support an object. All candidates, whether successful or not, are given feedback from the ability testing and on how they scored against set competencies.

The next round

The next assessment phase takes place about two weeks after the initial assessment centre. At this stage, candidates are grouped together according to the area of the business they would work in. Three different assessment days are organised with between 10 and 12 candidates attending a "sales and marketing", a "purchasing" or a "finance" assessment day.

The venue for the second stage is a management training centre. Prior to this, candidates will have completed an Occupational Personality Questionnaire. The format for the day is similar to that of the first assessment centre, with a variation in some of the exercises. For instance, on this occasion, candidates must give a 15-minute presentation. The topic for the talk, which is related to the chosen area of the business, is given to candidates in advance so that they have plenty of time to prepare.

Candidates at this selection stage are assessed by senior managers from the recruiting function, such as sales and marketing. Following a "wrap-up" session by the assessors, where they discuss their assessments of the participants, candidates are informed that evening whether they will be taken on by the company.

Mature entry

Although consisting of a similar battery of selection tests, Britvic's assessment centre for mature graduates is organised slightly differently. Fewer candidates attend one session and senior managers support the process from the outset.

"We feel that we need to invest more time in applicants who already have experience of the world of work, as this could be a serious career decision for them," explains Alison Cowen. "Therefore, the second-stage assessment centre involves an overnight stay to allow them to seek information from not only the key stakeholders of the programme but also from current programmers."

This final selection event is held at a commercial management training centre, where candidates have the opportunity to socialise in the evening with Britvic employees and may even have the chance to show off their rowing or bowling prowess.

A cost-effective option

Britvic is making a big investment in its graduate recruits. The "cost per hire" is estimated at around £2,800 per candidate. This is considered very reasonable, particularly in view of the 300% saving on previous years that this figure represents.

"The assessment centre route for selecting graduates is well worth the expenditure, and over the last year we have managed to make it even more cost-effective by streamlining the process and by the use of technology," says Alison Cowen. "We also involve a range of Britvic employees by keeping the process in-house."

PRET A MANGER ASSESSES MANAGERS ON THE JOB

Pret A Manger prides itself on selling fresh, natural food in its 120 fast-food outlets throughout the UK. Pret employs around 230 team leaders and 200 managers to run its shops, together with counter and support staff. The company aims to recruit to 60% of its management positions internally, while the remaining vacancies are filled externally.

Assessing managers

In recognition that internal and external candidates applying for a management position may not be on an equal footing, there are two distinct assessment centre processes for each group.

As Esther O'Halloran, retail recruitment manager, explains: "There is the obvious advantage to an internal candidate of being familiar with Pret's culture and modus operandi, but, conversely, our team members may not have had the opportunity to gain experience in other skills, such as giving presentations."

Pret aims to run one internal and one external assessment centre each month. The first stage of the process is the same for all candidates and consists of an application form, a psychometric test and an initial screening interview. The results of all three are taken into account before candidates undergo the assessment centre process.

The psychometric test focuses on numerical and verbal reasoning, and includes a personality questionnaire. The personality section of the questionnaire was developed with the input of around 60 Pret managers to ensure that it measures "real Pret behaviours". A software package produces a report based on the test answers that suggests specific questions to follow up with the candidate.

Back to the floor

The assessment day for external candidates kicks off at 6.30am at one of Pret's shops. The would-be manager shadows a Pret manager or team leader for four hours, learning the skills of the trade, such as making coffee and dealing with customers. In Esther O'Halloran's view, the morning session is an excellent way of getting across the reality of the job and the ethos of the organisation, which could not be achieved as effectively in an artificial setting. It also gives candidates an opportunity to assess whether the Pret environment is one they would feel comfortable working in.

Following this initiation into Pret life, the general manager responsible for the store completes a feedback sheet on the candidate's performance against set competencies. "By the time the assessment tests start, we have a substantial set of information to inform the selection process," explains Esther O'Halloran. Candidates will have been supplied with details of what to expect at the assessment centre well in advance of the day itself.

Up to eight candidates attend the centre in any one session. The ratio of assessors to candidates is high: three assessors to four or five candidates, rising to four assessors if there are six, seven or eight candidates. All assessors are Pret personnel, ranging from store managers to HR staff, who are trained to recognise the behaviours outlined in Pret's competencies.

"We like to involve Pret employees in recruitment," says Esther O'Halloran. "As well as giving them a sense of ownership of the process, they are the best people to judge whether candidates match up to our expectations."

The tests

Following an initial icebreaker, mysteriously involving a toilet roll, the first test is a group exercise involving a "leaderless project". The group is presented with an issue or problem that needs solving by consensus, and each candidate is measured against three key competencies: leadership, communication and tenacity/ resilience. Half-an-hour's preparation time is allowed beforehand.

The second exercise follows a communal lunch and is a one-to-one interview. Situational and behavioural questions based on Pret's competency framework follow up the results of the psychometric test and feedback form.

The final test is a presentation, made in the presence of candidates and assessors. Candidates are expected to talk on a subject that they have researched that day, using the internet. If possible, the topic is related to the food industry, and past presentations have ranged from "the history of English tea parties" to "the coffee bean". The brief is to make the presentation "interactive, visual (a flipchart is provided) and interesting". Audience participation is welcomed, as is a sense of humour.

Following the assessment day, assessors have a "wash-up" where they compare notes on the suitability of candidates. A summary sheet is completed outlining people's development needs, regardless of whether they were successful. This is handed to candidates before they leave. As Pret has a rolling management recruitment programme, it will take on all the candidates attending the session who are considered suitable.

Internal candidates

Pret found that existing employees attending the assessment centre were noticeably more nervous than their external counterparts. The company has therefore developed an alternative assessment day that mirrors the assessment centre process but, instead, takes place on the job.

Candidates undertake the same exercises as external candidates, but in a work environment - such as a presentation to the shop team and an exercise that tests leadership - and are measured against the same competencies. In one sense, the management recruitment process is even more rigorous for internal candidates, as their applications must first have been endorsed by at least three Pret managers.

Team member recruitment

A scaled-down version of the assessment centre has recently been developed for candidates applying to work as team members in a new store. Around 16 candidates attend a four-hour session, where they take part in games designed to assess competencies such as teamwork, communication and listening skills. The process is very useful, in Esther O'Halloran's view, as it begins to build the team spirit considered so vital to Pret's culture.

GCHQ SELECTS SPECIALIST STAFF

Government Communications HQ (GCHQ) employs around 4,000 staff, who study international communications for intelligence purposes and protect UK communications from hostile attack. GCHQ recruits between 200 and 300 people annually to its civil service operation. This involves running up to 30 different recruitment campaigns that could take on between one and 60 employees in any one round.

Recruitment at GCHQ is governed by the Civil Service Commissioner's "free, fair and open competition" code, which promotes equality of opportunity. All vacant posts are widely advertised to encourage applications from a broad range of people.

Assessment centres are the preferred selection method for around 40% of new employees and are used at two levels: for graduate-level recruitment (which includes candidates who may not have a degree but have compensatory work experience), and for internal promotion to some management grades.

Head of recruitment Geoff Trett explains GCHQ's approach to graduate-level recruitment: "Our external recruitment programme reflects the high intellectual and skills base needed for operations at GCHQ. Our main two assessment centres are therefore used for our technical campaigns and to select intelligence analysts."

The technical campaign

The assessment centre for graduate-level entry to a technical post at GCHQ is a 24-hour event and includes an overnight stay. The venue is a local Cheltenham hotel, where eight candidates assemble at 3pm to embark on a series of tests. The first challenge is a psychometric questionnaire involving numerical and written exercises. An opportunity to relax and socialise follows, with a senior technical representative of the company arriving for dinner and an informal chat.

The next day, the assessment process begins in earnest with one-to-one interviews for all candidates, conducted by HR and technical specialists. These are followed by a group exercise where would-be recruits are presented with a problem-solving scenario. The final hurdle is a psychological test that involves a self-assessment questionnaire and an interview with a psychologist. The group is released at 3pm, when the whole process starts again for the next set of hopefuls.

As GCHQ operates a rolling recruitment programme for technical posts, each assessment centre is likely to be repeated several times over a period of two or three weeks. This means that as many as 240 candidates could be assessed for around 50 to 60 posts, making it "a major logistical exercise". However, as Geoff Trett points out, a carefully planned matrix with clearly allocated responsibilities for administrative tasks helps the process to run smoothly. The event takes place three or four times a year.

A pool of specially trained assessors, drawn from relevant areas of the organisation, is available to assist on the day. The ratio is five assessors to eight candidates, made up of two technical assessors, two HR representatives and one psychologist. Although there is some variation in the precise make-up of the team on a day-to-day basis due to the intensity of the programme, it is felt that continuity is vital in order to preserve fairness. To achieve this aim, the organisers always ensure that at least some assessors are "carried over" to the next day.

Variation on a theme

The assessment centre process used to select intelligence analysts is a shorter event and starts at 8.30am with a lunchtime finish. As relatively few intelligence analysts are taken on by the organisation, assessment centres are organised on an ad-hoc basis. Six people attend at any one time, and the event is held at the GCHQ offices.

The process begins with two written tests, both of which are job-related. The first one is a "sizeable analytical exercise" that involves assimilating a significant amount of information, drawing conclusions and writing up the results. The second exercise requires the candidate to advise on a specific course of action following consideration of related data. A one-to-one interview follows.

The final selection test is a group exercise, with candidates presented with a logistical problem to solve. A number of possible solutions are offered and the group is invited to reach a consensus on the most suitable one. As Geoff Trett explains: "We are not just looking for who steers and directs the group. We are also observing how candidates contribute in other positive ways, for example, by facilitating the discussion."

A competency-based approach

GCHQ's competency framework is the bedrock of its selection process. It consists of 14 competencies (which will be reduced to seven during 2002), such as:

Five or six key competencies are chosen that are considered vital to a particular role and evidence of these is carefully assessed at the assessment centre.

A means to an end

In Geoff Trett's view, using assessment centres for some employee groups delivers value for money and yields high-validity results. However, he does not believe that they would be an effective selection method for all staff: "Assessment centres are very good at producing all-rounders, but could run the risk of not recognising some rare and specialist skills. For example, we employ some of the most brilliant mathematicians in the country, who excel in their field but may not perform so outstandingly in a group."

This research was conducted by journal staff, and by Rachel Suff (case studies) and Pauline Grant, author of Careering upwards (published by Right Way, price £9.99, www.right-way.co.uk) and a business psychologist with YSC, a consultancy assisting organisations to develop assessment centres and providing other human resources-related services (www.yscltd.co.uk ).

1Development and assessment centres: identifying and assessing competence, Charles Woodruffe, Chartered Institute of Personnel and Development, 3rd edn, 2000, www.cipd.co.uk .

²Assessment centres, Association of Graduate Recruiters, 2001, www.agr.org.uk , price £15 (free to AGR members).

³"Centres of excellence?", Paul Iles, British Journal of Management, vol. 3, 1992, pp.79-90.

⁴Recruitment survey report, May 2001, Chartered Institute of Personnel and Development, www.cipd.co.uk .

⁵"The TUC converts to competencies", Katherine Adams, Competency & Emotional Intelligence, vol. 6 no.3, spring 1999, published by IRS.

⁶"It pays to be aware of the subtle inequalities", Clive Fletcher, People Management, 7 November 1996.

⁷Redefining the fast stream, Cabinet Office, September 2001, www.civil-service.gov.uk .

⁸"Assess your staff with a modern business plan", Widget Finn, The Times, 18 January 2001.

⁹Assessing management skills, Margaret Dale and Paul Iles, Kogan Page, 1992.

¹⁰Janus, no.54, Association of Graduate Recruiters, September 2000.