Only 28% of hiring managers trust AI CV screening to predict job performance. Candidates trust it exactly the same, at 28% too.
Meanwhile the top reason hiring managers lose trust in their own hiring process is candidates who interviewed well and then couldn't do the job, at 39%.
There's a wave of new tools promising to fix high-volume screening. Very few of them are being asked the questions the psychometric world has been asking for decades: is it valid, is it consistent, and is it fair?
Join host Robert Newry as Dr Charlie Eyre sets out how to answer those three questions about any screening tool, whether it's a psychometric assessment or an AI matching engine. Charlie is Director at Workspheres Ltd, a former Principal Psychologist at the College of Policing, and a former Senior Editor at the British Psychological Society, where he oversaw the certification of psychometric assessments.
🎯 Does it do what it claims to do? Charlie explains what validity evidence actually looks like, from predictive validity studies to running a new measure side by side with an established one. Plus his honest caveat: you never see a fully valid measure, so it's about piecing together a jigsaw.
🚪 The excuse to stop accepting: "we couldn't have that evidence yet, it's a new tool." No tool is born with a complete evidence base. But if the research was done properly, it was done in a testing environment first.
🔁 Why consistency matters more now than it used to. Test-retest reliability was easy when assessments were fixed and identical for everyone. As tools adapt to each candidate, proving you're still measuring the same attribute the same way gets harder, and more important.
🚦 Why "ethical AI" isn't evidence of fairness. Good intentions and undesirable outputs can coexist. Charlie explains what fairness analysis across protected characteristics looks like, and why the risk rises as the sift gets harder.
📕 The one question to ask every vendor: do you have a technical manual? Charlie's version of looking under the bonnet, and why it doubles as your defence. Can you explain to a rejected candidate why they were rejected? And would that explanation survive a tribunal?
Tune in to hear how the standards built over decades in occupational psychology apply to a market moving faster than the evidence behind it.
Don't forget to hit 'subscribe' to be sure not to miss more episodes of the TA Disruptors podcast!
------------------------------------------------------------
Thanks for listening, and here are some helpful links:
Connect with Robert on LinkedIn: https://hubs.li/Q0282KjJ0
Connect with Charlie on LinkedIn: https://www.linkedin.com/in/charlieeyre/
British Psychological Society - Test-related documents: https://www.bps.org.uk/about-psychological-testing-centre
European Federation of Psychological Associations Test Review Model – Version 2025: https://www.efpa.eu/resource/efpa-test-review-model-version-2025/
Sign up for the TA Disruptors newsletter — it's the easiest way to be updated with the industry's brightest ideas and the freshest need-to-know data — every week: https://hubs.li/Q0282H370
------------------------------------------------------------
Talent acquisition is changing faster than it ever has before. ChatGPT and other GenAI tools are literally turning the selection process on its head. This podcast is designed for disruptive Talent Acquisition leaders who want to get ahead. For the innovators, the movers and shakers, and the TA leaders who will survive and thrive in this new era by moving quickly to future-proof their recruitment processes. Even when you're feeling overworked and overwhelmed, we'll help you keep up with the latest ideas from the industry's brightest innovators and the freshest need-to-know data from new research. TA Disruptors has got you covered.
All brought to you from the team at Arctic Shores.
Hosted on Acast. See acast.com/privacy for more information.
