This repository contains the source code and anonymized data used to perform all analyses of the article "Technical Note: Reliability of Pöch’s Facial Shape Classification System: A Morphometric Approach."
This study evaluates the reliability of Pöch’s facial classification system. Although previous publications have discouraged the use of such classification schemes, we recognize their relevance in facial identification where standardized methods for describing, collecting, and comparing facial morphological variation are essential. Our analyses assessed whether Pöch’s facial shape categories are representative of the facial variation in a Mexican sample and evaluated the intra- and inter-observer agreement when using these categories.
The results show that Pöch’s categories are redundant and do not fully capture the observed morphological diversity. However, the use of more morphometrically distinctive categories led to higher inter-observer agreement, aligning with patterns observed in psychological research.
We suggest different strategies to minimize inconsistencies in facial descriptions:
- Prioritize classification systems based on distinctive and representative face shapes.
- Oval, round, square and rectangular face shapes could be particularly useful due to their greater morphological separation (i.e. distinctiveness).
- Complement face shape categories with additional features such as hairline shape (round or straight) and overal facial proportions (narrow and long vs short and wide).
