It’s a familiar scene on daytime TV the world over, from Jerry Springer to Jeremy Kyle. Accusations of infidelity (and sometimes chairs) are tossed around (“What about that time at my sister’s wedding?”) and half-heartedly denied (“We never did anything, I swear”). But this is merely a prelude to the “lie-detector test”. “We asked Jason, since you’ve been with Kate, have you passionately kissed anybody else? You said no. But you were lying.”
Poor Jason. The only lie here is that a so-called “lie detector test” can tell for sure whether or not somebody is lying. In fact, measures of skin conductance – also known as galvanic skin response (GSR) – reflect merely intensity of arousal, not which particular emotion is being felt. So, if you already know that a person in a particular situation is likely to feel happy and excited (e.g, when winning the lottery), GSR can tell you how happy and excited that person is. Similarly, if you already know that a person in a particular situation is likely to feel stressed and anxious (e.g., because he knows an electric shock is coming), GSR can tell you how stressed and anxious that person is. This is why GSR is not much use as a lie detector. Most people get quite stressed and anxious if grilled about their sex lives, even if they have nothing to hide.
But if you do know what emotion your test participant is likely to be feeling at a particular moment, things start to get interesting. And potentially very useful. For example, the findings of a recent review1 of studies of gambling behaviour suggest that skin conductance levels may start to rise (reflecting increased excitement, and hence increased sweating) just before participants make a smart decision. That is, the GSR machine knows what decision the participant is going to make before she does! So, in terms of conversions, if you’re presenting your test participants with things that you’re pretty sure they’re going to like, and you set up your test very carefully, the GSR machine may be able to tell you just how much they like them. Indeed, if you suspect that your participants are just giving you the answers they think you want to hear, GSR could potentially be more reliable as a measure of how much they really like something than simply asking them.
Conversely, if you’re giving your test participants something that you know they’ll dislike, GSR can potentially be a great tool for finding out how much they dislike it. For example, in my last blog, we looked at the pain of paying. We learned that when deciding whether or not to make a purchase, people factor in not just how much they want whatever is on offer, but the pain they will feel as a result of spending the money. But again, asking people directly isn’t necessarily going to get at the truth here. If the total cost is the same either way, participants might feel a bit foolish admitting that buying on a credit card or a buy-now-pay-later scheme is more painful than shelling out immediately. If so, GSR holds the promise of telling you how much different methods of paying really hurt.
Just don’t expect the machine you be able to tell you whether or not any of your employees have been cheating on their partners!
Want to get started with GSR? Check out this practical guide.