Anyone who has hired for a few years knows the feeling. Two candidates interview very well, the references are glowing, but the decision still comes down to a hunch. Hunches aren’t useless, but they’re a shaky foundation for an appointment you’re going to be living with for years. Cognitive ability tests won’t solve the guessing game entirely, but they do provide one piece of hard evidence about how a person reacts to unfamiliar information.
Here's more or less how they work. The candidate receives a set of questions with a strict time limit and must work quickly. Some are written in words, some in numbers, some in shapes and patterns with no language at all. Nobody expects perfection. Many are designed so that not many people do everything. This is the way they divide the people who learn fast from those who need more time.
The section you choose is more important than people think. A bookkeeper’s day is full of numbers, so statistical reasoning has something useful to tell you. The verbal section is where a communications adviser proves his or her worth. Having every applicant go through the entire battery “just in case” often results in annoyed candidates and a mountain of data that nobody reads.
Who the score is compared against is another question. A result by itself is nothing; it only makes sense in a group of other people. Tests are often pitched at a certain level, for example general staff, graduates or senior managers, so using the wrong one skews everything. Ask the provider which level the vacancy falls under. Ask if the comparison group includes New Zealanders.
Candidates must be warned. Tell them the test is coming, how long it will take approximately, and that it is timed. A few practice questions are helpful for anyone who hasn’t done this since school, and people with conditions affecting timed work, such as dyslexia, should be able to request adjustments. Anxiety can bring down a score, and that’s not what you’re trying to measure.
"Don't over-interpret the numbers. Someone in the 68th percentile falls in the same range as the 72nd percentile. What you can say sensibly is whether someone is comfortably above the level of need of the job, or roughly at the level of the job, or well below it. That is a rough judgment, and it's the honest one.
The mistake I’d avoid is seeing cognitive ability tests as a pass/fail gate. Someone who trips over abstract puzzles can be a great customer-facing hire. The score is only one voice in the conversation, along with the interview, the references, and everything else you have learned. If it doesn't agree with your thinking, that is a sign to dig deeper, not dismiss the rest.
On the practical side, most of this is done online now. The candidate gets a link, completes the test in one sitting, and the report is delivered soon after. Always insist on seeing a sample report before you commit. If you can’t work it out in five minutes, your hiring panel won’t either. A provider who’ll walk you through the results afterwards is worth a lot more than one who just sends you a PDF.
Conclusion
Cognitive tests can’t pick your next hire. What they can do is replace a vague impression of someone’s reasoning with something you can actually look at. Select the sections that match the job, explain the process to candidates, and keep the result in proportion.
Comments