"Talent assessment tools" can mean two things: the providers that deliver assessments, and the assessment methods themselves. I'll mainly talk about the methods, but also comment on what to look for in the provider behind them.
They're one of the most widely used assessment tools for good reason. The biggest studies into hiring methods, looking at close to 100 years of data, show general cognitive ability to be the strongest single predictor of job performance across almost every role and level. The value these have added to hiring processes is huge.
There are three main types, and most providers offer all of them (sometimes combined into one test):
Each one only measures part of overall cognitive ability, so to get that top predictive power from 'general cognitive ability' you want to be combining them. If you want more detail on which test suits which role, our guide to the best pre-employment tests goes through each one.
Unfortunately, it's not as simple as going to any provider and using their ability tests. Two numerical reasoning tests can look very similar and be completely different in quality. To make sure you're getting a good deal and not sub-par quality, look out for these things:
Ability tests were a no-brainer 5 years ago, but things have changed a little. With static question formats and long time limits, the traditional ability tests have become too easy to cheat with AI. You may already be seeing it in your own data: average scores rising, many candidates scoring in the 90+ percentiles.
Most providers know this, and many now market their tests as "AI-resistant" because they've added anti-cheat features: tab-switching detection, copy-paste blocking, webcam proctoring, behavioural tracking and so on. These will help with some of the methods, but the truth is there are still ways around these which will let candidates score over 90% without answering a single question themselves.
So be careful before believing the marketing. Ask for a trial and take the tests yourself. If it's a static question with a generous time limit, it's not AI-resistant, no matter how many behavioural tracking measures sit on top of it. Genuine resistance comes from the design of the test itself, not from monitoring the candidate while they take it.
Why they're one of the top assessment tools: Ability tests let you screen a huge number of candidates remotely, and take your applicant pile from hundreds or thousands down to a shortlist of 10-20, without anyone having to read a single CV. When the tool doing that sifting is also the strongest predictor of job performance we have, it should be the best friend of any TA team. The catch is just making sure the results are actually the candidate's, not ChatGPT's.
3 minutesBen Schwencke, our Chief Psychologist, goes into the individual assessments in more detail: verbal, numerical and inductive reasoning tests, personality questionnaires, SJTs and interviews, and what each one predicts.
Now don't be put off by the name. Done well, these assessments give the same validity as ability tests, but solve their two biggest issues: the time they take to complete, and how easily you can cheat with AI.
Gamified (or game-based) assessments swap traditional question formats for interactive tasks. Historically, and what's also given them a bad name, these tests focused on personality traits and presented themselves in a whimsical fashion, like inferring risk-taking from how much you'll blow up a balloon before it pops (the makings of any great employee). These were supposed to make for an engaging candidate experience, but honestly candidates constantly complain online about how unserious and unrelated to the job they are.
But there are now good, professional, game-based assessments which measure cognitive ability, which are certainly worthy of consideration.
They measure verbal, numerical and inductive reasoning directly, with validated scoring, so you keep the predictive power of an ability test without the static format that makes it easy to cheat. If you're replacing or adding to your ability tests, this is the type you want.
Our MindmetriQ™ assessments were built to do exactly this. Each test takes 4-6 minutes, so a candidate can complete verbal, numerical and inductive reasoning in 12-15 minutes, which is quicker than a single traditional test, and gives you that measure of general cognitive ability we're always seeking. Each game also measures sub-facets like working memory and visual processing alongside the core ability, and the dynamic format means AI can't help.
Why they make the list: Gamified cognitive assessments do the same job as ability tests (early, remote, high-volume screening) but in a fraction of the time, and without the AI problem. So if you're choosing a tool to cut down your applicant numbers today, this is where I'd start. Just make sure it's measuring cognitive ability, and not how fast someone blows up a balloon.
What's the difference between these and ability tests? Well, ability tests predict a candidate's potential ability to do the job, and personality questionnaires and SJTs tell you how they're likely to go about it. I've combined personality and SJTs as they do a similar job from different angles, so you'd normally pick one or the other rather than use both.
A personality questionnaire asks candidates to rate how well statements describe them, and builds a profile of traits like conscientiousness, emotional stability, openness and how they work with others. Used well, they're good at predicting how someone will behave at work, how they'll fit a team, and (often importantly) whether they're likely to stick around. They work best alongside other data rather than as a pass/fail filter on their own.
Just like ability tests, quality varies a lot here. I should point out, personality tests for hiring are not the same as the MBTI tests you'll find online, which are built for personal development and team-building. They sort people into categories, they aren't designed to predict job performance, and they shouldn't be used to make hiring decisions. For selection, you want a questionnaire built on an established trait model, such as the Big Five, with norm groups and validity evidence behind it.
Better still is a bespoke personality assessment built around your own competency framework, so candidates are measured against the behaviours that actually matter in your organisation rather than a generic profile. We build these for clients, and they tend to be faaaar more useful to hiring managers than an off-the-shelf report.
SJTs give candidates workplace scenarios and ask them to rate or rank how they'd respond. Rather than asking people to describe themselves and inferring their behaviours from that, they ask directly how someone handles the kind of situations they'll actually face in the role.
The catch is that SJTs only work when they're role-specific. A generic customer service scenario won't tell you much about a trainee solicitor, so you will need bespoke SJTs to make them worth it.
SJTs are fairly susceptible to "cheating". Each scenario has options keyed as the most and least desired actions, so an AI tool, when supplied with information about the company and role, will be able to tell a candidate which options are likely the desired choices. Personality tests are a little less cheatable, as there's no real 'right answer'. You often have to answer 150+ statements and so it's much slower to use AI.
Where they fit: These are really useful, but mainly as supplementary data. Screening a large pool on personality or SJT results alone isn't something I'd recommend. Where they come into their own is alongside your ability tests. If candidates are already sitting an assessment, adding a personality questionnaire is an extra 10-15 minutes for them and a lot of extra info for you, which you can use to shape your interview questions and final decisions. So their usefulness is pretty tied to whether you're running ability tests. On their own they're a nice-to-have, but paired with them they're well worth it.
Interviews are the cornerstone of hiring decisions. They've always been used, will always be used, and should always be a big part of your hiring.
You might not know this but the validity of interviews does differ quite a bit depending on the style. Unstructured interviews (informal chats without a strict framework) have lower validity as they'll be influenced by how likeable or confident someone is, and less reliable as some candidates will likely be given a harder time with questions.
A structured interview sits closer to ability tests, whereby every candidate gets the same questions, and answers are scored against the same criteria. Done this way, interviews are one of the stronger predictors of job performance, and they let you probe and follow up in a way no online assessment can. Our guide to structured vs unstructured interviews covers the differences in more detail.
It's kind of surreal (and frustrating) how much AI has inserted itself into things. Video interviews, especially pre-recorded ones, are a much weaker choice now, as it's easy for candidates to script answers or use tools that listen in and feed them answers in real time. In-person interviews are less convenient and only suitabe for late stage but remain AI-resistant, which is why many organisations, like Google, are focusing on in-person interviews instead of video ones.
Are they a top assessment tool? Interviews are essential, but they work best at the end of the process, not the start. They take too much time to use for screening, so let your online assessments do the heavy lifting first. Then save your (structured, in-person) interviews for the handful of candidates who are left, where you can actually dig into the detail.
Well done, you've made it to the final part of this article, and like many graduate schemes, the final part is an assessment centre.
Assessment centres have strong validity due to the range of tasks being asked (group exercises, role-plays, interviews etc.), and as it's in-person they are AI-proof. Overall, they're a great option.
Wait, they cost how much?!
Yes, they take a lot of time to design, run and staff, so they only really make sense as a final stage for a small shortlist, typically on graduate schemes or senior appointments - and for those of you that have the resources to run them.
Where do they rank on the list? Assessment centres are a bit of a luxury. They're effective, but a lot of what they measure overlaps with tools you're probably already using: there's usually an ability test and an interview in there somewhere. If you have the budget and the headcount to run them, great. But they're not essential by any means, and a good mix of ability tests and structured interviews will get you most of the way there for a fraction of the cost.
So what have we learned? There's no single tool that gives you the full picture, and the most effective processes use two or three of them together. But if you're wondering where to start, the most useful tool is normally the one that can sift through and cut down your application pile the most. As the bulk of screening happens early, and ideally remotely, that stage is where AI does the most damage, so using an AI-resistant tool like gamified cognitive assessments makes the biggest difference to the rest of your process.
When comparing providers, always check the key points I listed, and try the tests yourself before believing the AI-resistance claims.
Test Partnership offers ability tests, MindmetriQ gamified assessments, personality questionnaires and situational judgement tests, including bespoke versions built around your competencies. They all run on the same platform, and our team can help you work out the right combination for your roles.
Head of Digital Marketing at Test Partnership. Over seven years' experience as a writer, content strategist, SEO and digital marketer.
