I saw a TV news report a little while back on a lie detector being marketed that was based on voice analysis. It may have been this gadget, or another that works on similar principles. The news guy was interviewing the creator of the gadget, asking about its reliability, accuracy, etc. At the end of the spot, he asks with a sly smile “would you be willing to answer a question about whether your lie detector works while it is turned on?” They proceeded to do just that.
Umm, what? If the thing works, he’ll answer that it does and he’s telling the truth, which will trigger no alarm. If the thing does not work, he’ll answer that it does and he’s telling a lie, which - because it does not work - will still trigger no alarm.
Unsurprisingly the guy answered that it worked and no alarm went off, leaving us just as uninformed as before.
Before you try to be clever… think.
I know, I know… if the thing did not work, it might just randomly generate alarms. But it is the principle of the thing.