Assessment Centre

A manager interviews several candidates for a role that involves difficult customer conversations. Each gives confident answers, but the employer wants to see how they handle a realistic situation.

An assessment centre can provide several observations rather than one conversation. Define the decision first: is the centre used to hire, promote or plan development?

The exercises and standards should match that purpose, so analyse the job and identify tasks and capabilities that matter to successful performance, because unrelated role-plays can mislead. Choose a small set of complementary exercises, since a case analysis, group task, presentation or simulated customer call can show different behaviour and more exercises are not automatically better.

Tell candidates what to expect, with clear timing, format, accessibility arrangements and assessment criteria as far as appropriate, because surprising people with avoidable logistics does not measure their job skill. Use trained assessors who know the scoring guide, record evidence and avoid discussing a candidate before independent notes are made, since a familiar manager's impression is not a substitute for observed facts.

Train assessors on bias too, as similarity, halo effects and early impressions can influence notes, while independent scoring and evidence review help challenge unsupported judgments. CIPD describes assessment centres as multiple observed activities and notes evidence that overall scores can predict performance, but it also cautions that individual competency ratings may be unreliable, so do not market every sub-score as a precise measurement.

Build a consistent rubric that scores the same exercise using the same anchors for comparable candidates, recording behaviour and results rather than adjectives such as confident or polished alone. Combine evidence deliberately: a numerical total can help summarise, but decide how much each exercise counts before seeing names or results, and avoid changing weights to favour a preferred person.

Consider conditions and fairness. Exercise order, fatigue or a long wait between tasks can affect performance, so give people comparable conditions, breaks and instructions, allow questions without penalty unless handling ambiguity is itself a stated and job-related skill, and review accessibility, since a disability or language need may call for reasonable changes that preserve the job-related construct under local discrimination law.

Avoid hidden tests, because candidates should know if an activity is part of assessment and secretly judging casual conversation in a waiting area can be unfair and difficult to validate; use work samples where possible, but do not ask candidates to produce commercially usable unpaid work disguised as a test, and observe group work carefully, because a loud speaker is not automatically an effective leader. Protect personal information, since assessment notes may contain sensitive impressions and should be shared only with authorised decision-makers under appropriate retention rules, and keep a decision record that explains how observed evidence, interviews and references contributed, because a ranking table alone may not tell the story.

An illustrative cost per candidate is total centre cost divided by candidates assessed: if a centre costs $24,000 and assesses 24 people, the arithmetic is $1,000 each, which does not measure selection quality or later job performance, and budgets should include assessor time, venues, technology and candidate travel, with online versions having different access needs. Check sample size and validate over time, because one exercise on one day can be affected by unusual nerves or ambiguity, and comparing decisions with later job outcomes lets you revise exercises that add no useful evidence; for owners, an assessment centre can widen the evidence for a consequential decision, and its value comes from job relevance, fair conditions and careful interpretation, not an impressive event schedule.