Maybe the boys with a low digit ratio were more proud so their head was tilted at a higher angle, because there's a huge difference in the angle of the head in those images. In photos taken at a lower angle, the forehead appears shorter and more tapered, the mandible appears wider and more round, the lower part of the face appears bigger relative to the upper part, canthal tilt appears lower, the nose appears shorter and more upturned, the distance between the eyebrows and the eyes appears higher, the eyebrows appear thicker, and the position of the ears relative to the eyes appears lower.
The methods section of the paper says that "the children's heads were adjusted according to the Frankfort Horizontal Plane", but they clearly failed to adequately normalize the position of the head:
Frontal photographs were taken with a digital reflex camera (Canon EOS 300D) and a 116 mm lens. To standardize the photographs, all children were advised to look straight into the camera with neutral facial expression, to remove their glasses or any facial adornment and to tie their hair back. The camera was positioned at eye height 3.5 m away from the face. Studio lights helped standardize the light conditions. The children's heads were adjusted according to the Frankfort Horizontal Plane [26].
The sample size of the study was 17, but a bigger sample size could've helped to neutralize differences in the angle of the head between the morphs.
You can use Google's Cloud Vision API to estimate the angle of the head in the photos:
https://cloud.google.com/vision/docs/drag-and-drop. Click "Show JSON" and search for "tiltAngle". The "tiltAngle" parameter was returned as about -11.8 degrees for digit ratio 1.094 and as about 3.7 degrees for digit ratio 0.874.
The reason why the eyes of the left morph look messed up is that all images consist of the same morph, but they just applied different meshes to the morph. It would've been useful if they also included morphs of the individual subjects in each digit ratio class.
Bookmarks