How to Choose a Psychometrics and Assessment Provider (UK)

UK employers use psychometric tests and structured assessments to inform hiring, promotion, succession, coaching and team development. Used well, they add evidence an interview cannot. Used badly, they produce confident reports that predict nothing and leave the employer open to a discrimination claim. This guide is for HR, People, Learning and Development (L&D), talent and procurement teams choosing a provider. It covers the types of assessment, the standards and qualifications that count in the UK, what the research says about predictive power, the law on fairness and data, what training and reports cost, and what to put in the contract.

Who sells what

The market has four kinds of business, and many combine more than one role. Test publishers develop and own instruments and license them: SHL, Hogan, Saville and Thomas are examples of publishers whose instruments you will meet in UK selection work, and Insights and Lumina of publishers whose instruments are used mainly for development and team work. Distributors and accredited practitioners deliver a publisher's instruments and interpret the results. Consultancies design and run assessment or development centres, 360-degree feedback and bespoke processes. Platforms deliver online testing at volume, increasingly with automated scoring or artificial intelligence (AI). Knowing which you are dealing with tells you where the expertise sits and who answers if something goes wrong.

Types of assessment

Selection use and development use

Settle this before you talk to anyone. Selection decisions, whether to hire or promote, need instruments with published evidence that scores predict performance in the role, applied consistently, with fairness monitored. Development uses, such as coaching, team sessions or career conversations, can draw on a wider range of tools, because the output is a conversation rather than a decision about someone's livelihood. Common practice among occupational psychologists is to use trait-based personality measures for selection and to keep type-based instruments, which sort people into categories, for development. A provider who proposes the same instrument for both without explaining the difference has not thought about your decision.

Standards and qualifications

No UK law restricts who may buy or use a psychometric test. Standards come from the British Psychological Society (BPS) and from the regulation of psychologists.

• BPS test user qualifications. The BPS recognises Test User: Occupational, Ability (formerly Level A), Test User: Occupational, Personality (formerly Level B) and Assistant Test User: Occupational Test Administration. The BPS does not run the training. Independent providers deliver and assess it, then the BPS charges a registration fee: £126 for Ability and £66 for Personality at the time of writing. Anyone who interprets results for your organisation should hold the relevant qualification.
• The Register of Qualifications in Test Use (RQTU). A public BPS register of people who hold its test user qualifications. Check the named practitioner before you sign.
• Publisher accreditation. Most publishers require users to complete training in each instrument before they can buy it. A certificate in one product does not qualify someone to use another.
• Regulated psychologists. "Occupational psychologist" and "practitioner psychologist" are protected titles. Only people registered with the Health and Care Professions Council (HCPC) may use them. The HCPC register is public. Chartered membership of the BPS (CPsychol) is a separate professional standing. The Association for Business Psychology also certifies practitioners, though its titles are not protected by law.

Registered tests

The BPS runs a voluntary test registration scheme. Publishers submit an instrument, and reviewers assess it against the European Federation of Psychologists' Associations (EFPA) review model, which covers the quality of the documentation, validity, reliability and the norms needed to interpret scores. A test that meets the standard becomes a BPS Registered Test for five years and may carry the logo. Registration is not an endorsement of the instrument for your purpose. It does show the publisher was willing to be examined, and the review itself is useful reading. Full reviews are free to RQTU members and to Chartered and Graduate members of the BPS, and non-members can buy one for £16.48 at the time of writing. If an instrument is not registered, ask the publisher why.

Validity, reliability and fairness

Three terms do most of the work when you compare instruments. Reliability is whether the instrument gives consistent results, for example if the same person takes it twice. Validity is whether it measures what it claims to and, for selection, whether scores relate to performance in the job. Fairness is whether it disadvantages groups who share a protected characteristic, usually examined through adverse impact analysis. A credible provider publishes all three in a technical manual and sends it on request. Ask which groups the norms represent and when they were last updated. Norms built on a UK working population are more useful than international student samples. The Chartered Institute of Personnel and Development (CIPD) makes the same point in its selection factsheet: tests should be backed by statistical evidence of validity and reliability, and used alongside other methods rather than alone.

What the evidence says

The research on which selection methods predict job performance has shifted. A widely cited 1998 review by Schmidt and Hunter, covering 85 years of studies, put general mental ability tests among the strongest single predictors. A 2022 re-analysis by Sackett and colleagues in the Journal of Applied Psychology argued that earlier estimates had been inflated by a statistical correction, produced lower figures for ability tests, and placed structured interviews among the strongest methods under the revised estimates. The authors cautioned against reading the results as a reason to prefer one method regardless of context.

The lesson for a buyer is that no single instrument should carry a hiring decision. Well-chosen tests add evidence, particularly alongside a structured interview and a work-sample task, and the combination is easier to defend on fairness grounds than any one method alone.

The law on testing and data

• Equality Act 2010. A test that disadvantages people who share a protected characteristic can be indirect discrimination unless it is objectively justified. Employers must also make reasonable adjustments for disabled applicants: extra time, an alternative format or a different method of assessment. The provider should have a documented adjustments process and be able to explain how adjusted results are interpreted.
• UK GDPR and the Data Protection Act 2018. Assessment results are personal data, and some reports touch on health or other special category data. Agree in writing what is collected, where it is stored, who can see it, how long it is kept and whether the provider uses candidate data to train or improve its tools. A data protection impact assessment (DPIA) is likely to be needed for large-scale profiling.
• Automated decisions. UK GDPR gives individuals rights where a decision with legal or similarly significant effects is based solely on automated processing. If a platform sifts candidates on scores with no human review, take advice on how those rights apply.
• AI in recruitment. On 6 November 2024 the Information Commissioner's Office (ICO) published the outcomes of voluntary audits of AI recruitment tool providers carried out between August 2023 and May 2024. The tools could comply with data protection law, but the ICO found gaps: some providers had not tested accuracy, some tools let recruiters filter candidates by protected characteristics, and some inferred gender or ethnicity from names. The ICO made 296 recommendations to the providers it audited. Ask any provider using AI how its tool was tested for accuracy and bias, and where the human sits in each decision.

Delivery and candidate experience

Most testing is now online and unsupervised, which raises the question of who sat the test. For high-stakes selection, common practice is to follow an unsupervised test with a shorter supervised verification test at interview. Check the platform's accessibility, availability in other languages if you recruit abroad, how reports are presented to managers, and whether it integrates with your applicant tracking system. Candidates should receive feedback on their results as a matter of course. It is long-standing good practice in the profession, and it protects your reputation as an employer.

What it costs

Most publishers price per report, per candidate or per block of credits, and the large selection publishers do not publish those prices online. We could not find a published per-report price for the main selection instruments in October 2026, so treat any figure a provider gives you as a quote to be checked against the scenario you describe. Training prices are published, and they tell you what it costs to bring the skill in-house:

• A combined BPS Test User: Occupational, Ability and Personality course was listed at £1,100 plus VAT by Quest Partnership and £1,350 plus VAT by Hogrefe UK in October 2026. ERAS listed a two-day Ability course on its own at £999 plus VAT. The BPS registration fees come on top.
• Publisher accreditation costs more. Hogan certification was listed from £2,330 plus VAT by its UK distributor, with one-day advanced courses from £1,430 plus VAT. Insights Discovery accreditation was listed at £3,817 plus VAT virtual or £4,333 plus VAT in person by a licensed provider, with an annual practitioner fee of £444 plus VAT after the first year.

The other variables that move a quote are volume, since per-unit prices fall with numbers; assessment centre design, assessor days and venues; set-up for 360-degree feedback, including bespoke questionnaires; bespoke norms or local validation studies; and feedback sessions for candidates or participants. Ask every provider to price the same scenario: the same number of people, the same instruments, the same feedback.

What to put in the contract

• A data processing agreement naming the provider as processor, with retention periods in months
• A clause that candidate data will not be used to train or improve the provider's products
• Where data is hosted, and notice of any transfer outside the UK
• The reasonable adjustments process and who pays for it
• Feedback to every candidate, and who delivers it
• Who in your organisation may see which reports
• The qualifications of the people interpreting results, by name or by standard
• For AI tools, the accuracy and bias testing to be repeated and reported, and how often
• Exit terms: return or deletion of data, and what you keep if you leave

What a bad answer sounds like

• "The technical manual is proprietary." A manual exists to be read.
• "It's a type tool, but our clients use it for hiring all the time."
• "Our algorithm removes bias." With no adverse impact figures.
• "You don't need the BPS qualification, our two-hour webinar covers it."
• "Adjustments are handled case by case." With nothing written down.
• "Candidate data helps us improve the product." Said as a feature.
• "Feedback is available on request." Candidates rarely know to ask.

How to evaluate a provider

Questions to ask providers

FAQs

Sources and further reading

• About qualifications in test use, British Psychological Society
• Qualifications and fees, British Psychological Society (checked October 2026)
• Register of Qualifications in Test Use (RQTU), British Psychological Society
• Test reviews and registration, British Psychological Society
• The professions we regulate, Health and Care Professions Council
• Selection methods, CIPD
• The validity and utility of selection methods in personnel psychology: practical and theoretical implications of 85 years of research findings, Schmidt and Hunter, Psychological Bulletin (1998)
• Revisiting meta-analytic estimates of validity in personnel selection: addressing systematic overcorrection for restriction of range, Sackett, Zhang, Berry and Lievens, Journal of Applied Psychology (2022)
• Reasonable adjustments for workers with disabilities or health conditions, GOV.UK
• Rights related to automated decision making including profiling, Information Commissioner's Office
• AI tools in recruitment: audit outcomes report, Information Commissioner's Office (6 November 2024)
• BPS Test User: Occupational, Ability and Personality course, Quest Partnership (price checked October 2026)
• Test User: Occupational, Ability and Personality combined course, Hogrefe UK (price checked October 2026)
• BPS Certified Occupational Test User: Ability, ERAS (price checked October 2026)
• Hogan training and certification workshops, Psychological Consultancy Ltd, UK distributor (price checked October 2026)
• Insights Discovery accreditation, Performance Catalyst, licensed provider (price checked October 2026)

Named publishers and training providers appear as factual examples of a type, with prices taken from their own published pages on the dates given, and are not recommendations. This guide is general information, not legal advice.

To see every provider listed, you can compare psychometrics and assessment providers on CompareHQ.