Wrist-worn device for monitoring people's emotions during market research


Originally published at: http://boingboing.net/2017/07/17/wrist-worn-wearable-for-monito.html


So, this creeps me out a bit.


Ugh. Beware of market research companies’ claims of what they can tell you that you don’t already know. Just like the fact that lie detectors are simply not effective at what they are supposed to do, I would be skeptical that this kind of biometric data can be used to tease out complex human emotions… so many confounding factors in play. Maybe it can work, but I’d love to see some scientific research rather than claims by a for-profit company looking to score some new contracts.


I am not paid enough to have emotional needs or wants!


Aw, it’s just skin conductance, and not an accelerometer? I was hoping it could track me making the “wank” gesture.


I did research work using GSR around a decade ago. The group I worked with found we could detect changes in focus or arousal (not just the sexual kind) accurately, but every other emotional response we tried to infer never got significant enough results.

Then we built a sweet chair and trolled people with an interactive art exhibit


And this is different from the pseudo-scientifically designed lie detector how…? What emotion will be detected during anal clenching, I wonder?


Didn’t they do stuff like that already in the 1960ies/1970ies?
Oh, right, the relevant buzzwords have changed since then…
Just remember to throw in “AI” and “deep learning” a couple of times, and Bob’s your uncle.


how do I make something like this so a computer can tell me what feelings I’m having? because i’m really bad at noticing otherwise and this would help with my anxiety/depression/undiagnosed emotional regulation issues


You call that an emotion reader? There’s not even a dodgy, underpowered fMRI study to go with it! Amateurs.


E meter bollocks. I’m calling ‘woo’


“It’s your birthday. Someone gives you a calfskin wallet. How do you react?”


A decade ago, I was a grad assistant in the Com dept and assisted with eye-tracking experiments.
Turns out what people said they were looking at was different from what they actually looked at. Also, car chases are exciting no matter what screen size you watch them on.
The point being, any kind of biometric tracking gives researchers more information than surveys and focus groups. Just one more tool in the arsenal to identify people’s behaviors, and change or modify them.


The problem is that this sort of data typically doesn’t mean what people think it does. As others have pointed out, lie detectors aren’t reliable, because while the data is accurate, the assumption that it’s a latent variable that can be used to detect lies doesn’t hold up. Bad statistical practices can also make a lot of noise look like something it’s not.

So, while you’re right that people are getting more data, there needs to be careful thought as to what kind of data you’re getting, what statements you can actually make about it, and how you can analyze it. I imagine eye tracker data is better because the assumption that the fovea is pointed at the area of attention seems pretty solid.



This topic was automatically closed after 5 days. New replies are no longer allowed.