Now it’s time to play “What’s Wrong With This Test?”. From ScienceNews:
Researchers in Hungary and Mexico used brain-scanning technology on 20 pet dogs to measure responses to faces. The dogs were trained to lie still in a sphynx position inside an MRI tube, resting their head on a chin rest while watching a screen. The scientists played four types of two-second video clips for the dogs to view: the front or back of a human head, and the front or back of a dog head. Thirty human volunteers in MRI machines saw the same short videos.
As many earlier studies have found, faces were captivating for people. When shown a face — either human or dog — a large swath of these people’s visual systems became active. These brain regions were quieter when the people saw the backs of heads.
The vision-processing parts of the dogs’ brains, however, didn’t seem to care about faces, the researchers report October 5 in the Journal of Neuroscience. No brain areas had greater activity when viewing a face compared with the back of a head. Instead, areas of the dogs’ visual systems were more tuned to whether the video featured a dog or a human.
They provide a hint in the next paragraph:
Still, the study measured brain responses — not behavior. The results don’t mean that dogs themselves don’t see, or don’t care, about faces. Other studies have shown that canines can recognize people’s facial cues.
So, what’s wrong with the test that they conducted? Your answer should reconcile the difference between the test that was conducted and the many other studies that have found that dogs respond to human facial expressions.
I wonder what the flicker rate of the videos was. There’s anecdotal evidence that dogs are becoming more attentive to screens in the age of TVs with faster refresh rate.
Other factors too….being put in a Sphinx position in a tube so they’re not in a natural relaxed state, and cognitively they don’t know that 2D images correlate to 3D so they’re not likely to tune in.
Dogs responded to whether the image was dog or human, dogs were trained to sit in the Sphinx
Position:
The test didn’t draw dog responses, it drew trained dog responses. From dogs trained by the testers.
All of the defects mentioned are worth considering as is the small scale of the study. Something else to consider: dogs’ perceptual space is drastically different from that of humans. Dogs rely on scent, then hearing, then sight. Humans rely on sight, then hearing (in a different frequency range) then scent. The images to which the dogs were exposed didn’t smell like people and most dogs do not think abstractly.