In short, it means that an assessment accurately measures what it says it will. There’s quite a few different types of validity, though, which makes it a bit trickier. But don’t worry – let’s briefly explore the four most common types together:
This establishes whether the test ‘looks’ how it should. It’s not exactly what you’d call ‘rigorous’, and so it probably shouldn’t be the only measure of validity you consider if you’re looking at psychometric tests.
This considers what a test says it measures, and then looks at whether it contains the right elements to measure those traits. It’s a slightly more detailed version of face validity.
Construct validity is often what people mean when they simply refer to “validity” in psychometric testing. It basically establishes whether an assessment measures what it says it will.
Testers will often get to the bottom of this with some complex statistical modelling, but it’s also possible to establish construct validity by comparing tests to previously-validated examples.
Confusingly, there actually two types of criterion validity, predictive and concurrent:
Concurrent – considers whether the scores in an assessment relate to something else that we’re able to measure now.
Why is validity important?
In short, it pays to know what you’re measuring.
Imagine if you wanted to hire resilient people, for example. And all through your assessment stage, you thought you were measuring your candidates’ resilience. Then you make a hire based on the results of your assessment, only to realise that you weren’t ever really measuring resilience at all!
In this way, it’s easy to end up making expensive mis-hires with an invalid assessment.
So validity’s pretty important, and you should never feel awkward asking any test provider about it. If you’re on the hunt for an assessment that measures what it says it does, then we’d love to walk you through our behaviour-based assessment. Just click here to sort a quick, free demo.