How we change what others think, feel, believe and do
The Assessment Center is an approach to selection whereby a battery of tests and exercises are administered to a person or a group of people across a number of hours (usually within a single day).
Assessment centers are particularly useful where:
Individual exercises provide information on how the person works by themselves. The classic exercise is the in-tray, of which there are many variants, but which have a common theme of giving the person an unstructured large pile of work and then see how they go about doing it.
Individual exercises (and especially the 'in tray') are very common and have a correlation with cognitive ability. Other variants include planning exercises (here’s problems, how will you address them) and case analysis (here’s a scenario, what wrong? How would you fix it?).
In one-to-one exercises, the candidate interacts in various ways with another person, being observed (as with other exercises) by the assessor(s). They are often used to assess listening, communication and interpersonal skills, as well as other job-related knowledge and skills.
In role-play exercises, the person takes on a role (possibly the job being applied for) and interacts with someone who is acting (possibly one of the assessors) in a defined scenario. This may range from dealing with a disaffected employee to putting a persuasive argument to conducting a fact-finding interview.
Other exercises may have elements of role-play but are in more 'normal' positions, such as making a presentation or doing an interview (interesting reversal!).
Group exercises test how people interact in a group, for example showing in practice the Belbin Team Roles that they take.
Leaderless group discussions (often of a group of candidates) start with everyone on a relatively equal position (although this may be affected by such as the shape of the table).
A typical variant is to assign roles to each candidate and give them a brief of which others are unaware. These groups can be used to assess such skills as negotiation, persuasion, teamwork, planning and organization, decision-making and, leadership.
Another variant is simply to give a give topic for group to discuss (has less face validity).
Business simulations may be used, sometimes with computers being used to add information and determine outcomes of decisions. These often work with 'turns' that are made of data given to the group, followed by a discussion and decision which is entered into the computer to give the results for the next round.
Relevant topics increases face validity. Studies (Bass, 1954) have shown high inter-rater reliability (.82) and test-re-test results (.72).
A neat trick is to ask candidates to assess themselves, for example by asking them to rate themselves after each exercise. There is usually a high correlation between candidate and assessor ratings (indicating honesty).
Ways of improving these exercises include:
Those with low self-assessment accuracy are likely to find behavioral modification and adaptation difficult (perhaps as they have low emotional intelligence).
Developing assessment centers involves much test development, although much can be selected 'off the shelf'. A key area of preparation is with assessors, on whose judgment candidates will be rejected and selected.
Identify the criteria by which you will assess the candidates. Derive these from a sound job analysis.
Keep the number of criteria low -- less than six is good -- in order to help assessors remember and focus. This also helps simplify the final judgment process.
Make exercises as realistic as possible. This will help both candidates and assessors and will give a good idea what the candidate is like in real situations.
Design the exercises around the criteria so they can be identified rather than find a nice exercise and see if you can spot any useful criteria. Allow for confirmation and for disconfirmation of criteria.
Include clear guidelines for player so they can get 'into' the exercises as easily as possible. You should be assessing them on the exercise, not on their memory.
Include guidelines also for role-players, assessors and also for those who will set up the exercises (eg. what parts to include in exercise packs, how to set them up ready for use, etc.).
Triangulate for results across multiple exercises so each exercise supports others, showing different facets of the person and their behavior against the criteria.
Select assessors based on their ability to make effective judgments. Gender is not important, but age and rank are.
There are two approaches to selecting assessors. You can use a small pool of assessors who become better at the job, or you can use many people to help diffuse acceptance of the candidates and the selection method.
Do use assessors who are aware of organizational norms and values (this militates against using external assessors), but do also include specialists, e.g. organizational psychologists (who may well be external, unless you are in a large company).
Develop tools for assessors
Asking assessors to make personal judgments is likely to result in bias. Tools can be developed to help them score candidates accurately and consistently.
Include behavioral checklists (lists of behaviors that display criteria) and behavioral coding that uses prepared data-gathering sheets (this standardizes between-gatherers data).
Traditional assessment has a process of observe, record, classify, evaluate. Schema-based assessment has examples of poor, average and good behavior (there is no separation of evaluation and observation).
Prepare assessors and others
Ensure the people who will be assessing, role-playing, etc. are ready beforehand. The assessment center should not be a learning exercise for assessors.
Two days of training are better than one. Include theory of social information processing, interpersonal judgment, social cognition and decision-making theory.
Make assessors responsible for giving feedback to candidates and accountable to organization for their decisions. This encourages them to be careful with their assessments.
Run the assessment center
If you have planned everything well, it will go well. Things to remember include:
After the center, follow up with candidates and assessors as appropriate. A good practice is to give helpful feedback to candidates who are unsuccessful so they can understand their strengths and weaknesses.
Assessments have grown hugely in popularity. In 1973 only about 7% of companies were using them. By the mid-1980s, this had grown to 20%, and by the end of the 1990s it had leapt again to 65%.
Assessment centers allow assessment of potential skill and so are good when seeking new recruits. They allows a wide range of criteria to be assessed, including group activity and aggregations of higher-level, managerial competences.
Assessment centers are not cheap to put on and require multiple assessors who must be available. Organizational psychologists can be of particular value to assess and identify the subtler aspects of behavior.
The assessment center was originated by AT&T, who included the following nine components:
Reliability and validity is difficult, as there are so many parts and so much variation. A 1966 study showed high validity in identifying middle managers. There is a lower adverse effect on individuals than separate tests (eg. psychometrics).
The outcome of assessment centers are based on the judgments of the assessors and hence the quality of those judgments. Not only are judgments subject to human bias but they also are affected by the group psychology effects of assessors interacting.
Assessors often deviate from marking schemes, often collapsing multiple criteria into a generic ‘performance’ criterion. This is often due to overburdening of assessors with more than 4-5 criteria (so use less). More attention is often given to direct observation than other data (eg. psychometric tests). Assessors even use their own private criteria – especially organizational fit.