Eye centring in selfies posted on Instagram.

Area of Science:

• Psychology
• Artistic Composition
• Computer Vision

Background:

• Previous research indicated artists exhibit an eye-centring bias in portraits.
• This bias, potentially rooted in psychological mechanisms, has faced theoretical and empirical challenges.
• Detecting this bias in non-professional self-portraits (selfies) remains an open question.

Purpose of the Study:

• To investigate the presence and characteristics of an eye-centring bias in a large, representative dataset of selfies.
• To compare selfie eye-centring patterns with previous findings in traditional art and theoretical predictions.
• To explore potential hemispheric asymmetries (pseudoneglect) and directional preferences in selfie composition.

Main Methods:

• Analysis of approximately 4000 spontaneously posted selfies from Instagram across six major world cities.
• Quantitative assessment of the horizontal position of the most-centred eye within each selfie image.
• Statistical comparison of observed eye-centring distributions against realistic Monte Carlo simulations.

Main Results:

• A significant eye-centring bias was observed, with the most-centred eye position peaking at the horizontal image center.
• The observed distribution significantly deviated from random chance or realistic rendering predictions.
• A statistically reliable, albeit small, pseudoneglect effect was detected, favoring left-eye centring.

Conclusions:

• The study provides strong evidence for an eye-centring tendency in non-artist self-portraits (selfies).
• This bias appears consistent across a large, diverse, and spontaneously generated sample of digital self-images.
• Findings suggest underlying psychological or compositional principles influence how individuals frame themselves in self-portraits.