HR articles

Assessment Center: What It Is, Exercises and How to Run and Score One

SHORT ANSWER

An assessment center is a method, not a place: a group of participants completes several exercises that simulate the job, trained observers record their behavior against a fixed list of competencies, and the ratings are agreed in a joint discussion. Each competency is measured by at least two exercises. It is used for hiring, promotion and development planning.

Course on this topic: Assessment Center Course: Employee Evaluation Training

People hear “assessment center” and picture a building. It is a method. The hard part is never the day itself. It is the weeks before: deciding what exactly you want to see and in which exercise you will see it.

Below: what it is, when it is worth the effort, the exercises, and how to run and score one. To plan your own, jump to the competency × exercise matrix builder.

What is an assessment center?

An assessment center is a multiple evaluation process. The name comes from the place where AT&T assessed the managerial potential of its employees. It can last from a few hours to several days.

Five features make an event an assessment center:
  1. A group of participants. Most often four to six people, twelve at most. A center for one person is possible as a final check before hiring.
  2. Varied exercises. Simulations built on real work material, usually with an interview and sometimes with tests.
  3. A team of trained observers. A mix of HR people and line managers, all trained for the role.
  4. Predefined behavior. Observers assess against a fixed list of job-related competencies, not against a general impression.
  5. A joint decision. Ratings are agreed when the observers sit down together and go through all the data.
Everything in it is about behavior: what a person says and does in a situation close to the job.

Assessment center vs development center

The exercises can be identical. The purpose is different.
  • Assessment center — selection. It ends with a decision: hire or not, promote or not. Participants are compared with a standard set in advance, the assessor acts as a judge, and feedback comes later.
  • Development center — diagnosis. There is no pass mark. It looks for strengths and development needs, the facilitator gives feedback during or at the end of the event, and the result is an individual development plan.
Decide which of the two you are building before you design anything, and tell participants which one it is.

When do you need an assessment center?

The four classic uses: selecting graduates, hiring outside candidates for senior roles, choosing people for promotion, and career development.

From practice I would add a fifth: when you cannot judge people by results. A new manager once took over a unit of 50 people at the TV channel where I worked. There were problems with turnover, efficiency and pay. He asked HR for a way to assess competencies, behavior and knowledge, not only results. These employees hardly worked with colleagues — their job was to find guests for the shows and people for the studio audience — so a 360 review had nobody to ask. We ran an assessment center instead: a role play where you had to persuade a person to come on a show as its hero, a list of the competencies that mattered, and a basic computer test, because they prepared reports.

When is it not worth it? When the purpose is vague. Launching a center from the first idea takes about four months, and the cost has two parts: development and running the sessions. If nobody can say what decision the results will feed, do not start.
FROM MY PRACTICE · STARLIGHTMEDIA

It started with the casting team for studio audiences: we assessed them, split people into categories, and people got promotions and pay rises. Then the head of the camera operators came to me: “I need an assessment too. I have many operators, I need to split them into categories — who earns more, who earns less — and understand whom to put on which shoot. We do not know who can handle a difficult shoot and who can cope with a simple one.”

The problem: an operator is a person with a camera, with no goals or KPIs as such. The quality of the footage did not depend on him — it depended on the director and on the light. It was a rare case where the result could not be measured, and what saved us was an assessment center and an assessment of hard and soft skills.

Together with the chief operator we described the competencies in four categories: light (a hard skill), creativity (a soft skill), camera work and editing (a hard skill), self-sufficiency and proactivity (a soft skill). One of the tasks was to recreate a shot. A person walked into a room with lights, an actor, props and a camera, drew a card for his category, and had to set the light, set the camera and record the shot. We assessed how quickly and how well he recreated it and how many takes he needed.

When a job has no measurable result, do not invent a KPI. Build the job in one room and watch the person do it.

— Mike Pritula

Assessment center exercises: the main types

The exercises are the core: they let a person show the behavior the job needs.
  • Cooperative group discussion. Everyone gets the same information and no roles; the group analyzes a problem and gives recommendations within a time limit.
  • Competitive group discussion. Each person also holds information only they know and competes for limited resources. The task is double: win the most for yourself and help the group reach the best solution.
  • Presentation. Prepare and deliver. The simplest form is a self-presentation: five minutes to prepare, five to speak.
  • Role play. A one-to-one meeting with a role player: a client, a colleague, a subordinate. In my example a manager talks to a “difficult” employee with ten minutes to prepare and twenty for the conversation.
  • Fact-finding. A short description of a situation and an expert who answers only specific questions. Five minutes to read, fifteen to ask, five to present the decision, ten to defend it.
  • In-basket. A folder of 10 to 40 letters, reports and notes. The participant sets priorities, delegates and writes instructions for each item.
  • Written analysis of information. Two or three pages of text with tables or charts and a report to write.
Next to the exercises usually sits a competency-based interview: it asks for behavioral examples in the STAR format — situation, task, action, result — and is sometimes the best source of data for a criterion. Specialist jobs may add a knowledge test.

One rule if you want to see leadership in a group discussion: give the group several small problems and appoint a different leader for each. Otherwise one person dominates and the rest give you nothing to record.

How to match exercises to competencies

You build a table: competencies in rows, exercises in columns, a mark where the competency can really be observed.

Three rules from the course:
  • Start from the job. Job analysis shows what separates successful people in this role from less successful ones. Keep up to five competencies and describe each with positive and negative behavioral indicators.
  • Every competency is measured by at least two tools. The most important ones — by three, the rest — by one or two.
  • Mix the formats: at work people act in groups, one-to-one and alone.
The builder below uses the matrix from my course: nine typical management criteria and six exercise types.
FREE TOOL

Competency × exercise matrix builder

Tick the competencies you want to assess and the exercises you plan to run. The table shows where each competency can be observed and whether you have enough observations.

Competencies
Exercises
Five competencies and four exercises are selected: 14 observation points, every competency is seen in at least two exercises.

The full version — the specification table for a real position, ready exercises, observer forms, the observer–participant schedule and the report template — is in the Assessment Center Course: Employee Evaluation Training.

How to run an assessment center: step by step

  1. Write a memorandum. A short document for the organization: the goals, who it is for and how participants are selected, who the observers are and how they are trained, how the results will be used and stored, who gives feedback and when, and whether reassessment is possible. Competencies come from job analysis.
  2. Build the matrix and develop the exercises. Form a team with line managers who know the job, write a first draft, run a pilot on people of the same level, refine, and write the observer guide with examples of positive and negative behavior for each criterion.
  3. Select and train observers. Training usually takes three to five days. Cut it short and you pay later with long arguments about what a criterion means.
  4. Plan the day. Six people is the best size for a group discussion. The number of observers depends on the exercises and criteria: from 0.5 to 2 per participant, and for a selection center one assessor for every three to five participants. Rotate observers between participants.
  5. Brief the participants. Observers will record behavior, nobody needs to play a role, most tasks are timed.
  6. Run the exercises. Observers stay neutral: no nods, no smiles, no comments on results, no discussion with each other until the end.
A one-day center for six people in my materials runs from 10:00 to 17:30: self-presentation, in-basket, role play, group discussion and interview. The observers then stay for the score consolidation, planned from 17:45 to 22:00.

Should you bring in consultants? You can at any stage. At Wargaming the company was growing fast and managers were appointed quickly and without common criteria; many great specialists turned out to be poor managers. A project team of about ten HR people, me included, chose DDI as the provider, defined an assessor profile, selected assessors from our own employees and trained them. They were certified and could then run the assessment themselves.

How to score an assessment center

Observers work in four steps:
  1. Observe. Watch and listen. Facts, not interpretations.
  2. Record. Write down what is said, verbatim where possible. “He was not interested” is a guess. “He did not speak for ten minutes” is a record.
  3. Classify. Right after the exercise, sort each piece of behavior by competency and mark it as a positive or negative example. Silence after a direct question is also data.
  4. Evaluate. Only now give a rating for each competency in this exercise, on a scale all observers read the same way.
Rating scale
5 — very high level of ability
4 — more than acceptable
3 — acceptable
2 — less than acceptable
1 — unacceptable
At the start of training the same behavior often gets a 2 from one observer and a 4 from another — the reason to practice before the real day.

After the last exercise the ratings go into a results matrix for each participant: competencies in rows, exercises in columns. Then comes the observers’ discussion. Participants are taken one by one, and for each competency the group looks at the evidence from all exercises and agrees on a final rating.

The chair has three jobs:
  • ask for the behavior behind every rating;
  • let senior observers speak last, so their status does not set the score;
  • keep out anything learned outside the center. The rating rests only on what was seen that day.

What to do after the assessment center

The report. For a hiring decision it is short; for a development center it is detailed — from 1 to 15 pages. It gives the final rating for each competency with the behavior that supports it, two or three main strengths and weaknesses, and development recommendations.

Feedback. Give it to everyone, including candidates who did not get the job, and as soon as possible. A personal meeting of one to two hours works best. The direct manager should not be the one who delivers it.

The check. Later, compare the ratings with how people actually perform, for example with their 360 feedback. If the results go into a talent review, see also the guide to the 9-box grid.

Common assessment center mistakes

  1. No clear purpose. Selection and development need different designs.
  2. One exercise per competency. A single observation is an episode, not evidence.
  3. Untrained observers. They write impressions instead of behavior, and the final discussion turns into a contest of opinions.
  4. Outside knowledge in the room. A favorite who performed poorly gets rescued by reputation.
  5. No feedback. People gave you a day of hard work and heard nothing back.
The full path — job analysis, ready exercises, observer training, the forms for the day and the report template — is in my Assessment Center Course: Employee Evaluation Training.
FAQ

Frequently asked questions

What is an assessment center?

An assessment center is an evaluation method in which a group of participants completes several job-like exercises while trained observers record their behavior against predefined competencies. The final ratings are agreed in a joint discussion of all observers.

What exercises are used in an assessment center?

The main types are group discussions with or without assigned roles, role plays, presentations, fact-finding exercises, in-basket exercises and written analysis of information. They are usually combined with a competency-based interview and sometimes with ability tests or personality questionnaires.

How many participants and assessors do you need?

Most centers are run for four to six participants, twelve at most. In my course the guideline is one assessor for every three to five participants in a selection center, and six people as the best size for a group discussion.

What is the difference between an assessment center and a development center?

An assessment center ends with a decision: hire or not, promote or not. A development center uses the same exercises to find strengths and development needs, gives feedback during or right after the event and ends with a development plan.

How is an assessment center scored?

Observers record behavior, sort it by competency and rate each competency in each exercise on a common scale, for example from 1 to 5. The final rating for every competency is agreed in a joint discussion, participant by participant.