3 ms·
While I could point out several questionable aspects of their methodology, I'm even more troubled by the complete absence of the word 'ethics' from this paper.
by anigbrowl 7y ago
While I could point out several questionable aspects of their methodology, I'm even more troubled by the complete absence of the word 'ethics' from this paper.
Do the people working on these projects never reflect on the ridiculously broad scope of the capabilities they're claiming, the multitude of possibilities for false or inappropriate positives (eg people who are trying to conceal their emotional states, but for wholly valid reasons), the abuses that will be committed by over reliance on such technology and rationalized with spurious claims of objectivity, and the overall rush to make public spaces and the world at large into a giant prison camp where techno-utilitarians pass judgement on the rest of the human race?
Almost every paragraph I look at in this paper makes me roll my eyes. A small sample:
We recruited 88 participants (51 female, 30 male, 7 preferred not to say, age = 20.35) from a university campus.
Wow, fantastically diverse sample.
In the deceptive condition, the participant is supposed to conceal their activities from these “onlookers.” The experimenter stresses that [an experimental prop] is a secret, and because of the amount of money involved, it is really important that the participant keeps anyone from noticing as he/she does these things [with the prop]. In the natural condition, the confederate is wearing a hoodie with the hood pulled down. However, in the deceptive condition, the confederate is wearing the same hoodie with the hood pulled up while wearing sunglasses. Because of this setting, the participant makes a deliberate attempt to conceal his/her activities, resulting in deceptive behavior
We use these findings and consider the following set of gestures:{Hands In Pockets, Looking Around, Touching Face, Touching Shirt/Jackets, Touching Hair, Hands Folded, Looking at Phone}. We chose this set because it includes all the hand gestures observed in the walking videos of participants, and these gestures have been reported to be related to deception
Figure 7. Gesture Features: We present the percentage of videos in which the gestures were observed for both deceptive and natural walks. We observe that deceivers put their hands in their pockets and look around more than participants in the natural condition.
So you got a recruited a bunch of college students and then told some of them to be super-covert while dealing with someone dressed up to look like a stereotype of a shady person, but somehow you imagine this will give an 'honestly deceptive' result that accurately reflects the behavior of deceptive people, rather than the awkward-embarrassed parody of deceptive behavior from someone who feels they're playing a silly game.
Besides all the ethical pitfalls, their research is bad. It's like trying to predict crime based on the observed behavior of people in a school drama class putting on a play about cops and robbers. This is why we have so much of a problem with junk forensic science.