My Google-fu is failing me…
But I recall a study from some years ago, a legit medical study with the results captured on video.
The subjects were put in a safety harness so they wouldn’t actually fall.
They were then invited to walk toward a clearly visible stumbling block and asked to trip over it.
That’s where the tricky part came in.
The researchers wanted actual stumbles, not deliberate near-falls, so they had a remotely controlled block that popped up, well in front of, the decoy stumbling block.
It worked quite well, as memory of watching the video serves.
I just wish I could find it to link here.