Description

Langley et al reported a classification and regression tree (CART) analysis for determining the sex of human remains from features of the skull. The authors are from the Mayo Clinic and Lincoln-DeBusk College of Osteopathic Medicine.


Parameters:

(1) glabella

(2) zygomastic extension (posterior)

(3) mastoid

 

Each parameter is scored by comparison with images.

 

Parameter

Points

none

1

mild

2

moderate

3

moderate to marked

4

marked

5

 

A CART analysis then uses these scores to determine the probable sex.

 

Glabella

Zygomatic

Mastoid

Sex

>= 4

>=3

NA

male (96%)

>= 4

< 3

NA

female (77%)

< 4

>=3

>=4

male (69%)

< 4

>=3

<4

female (73%)

< 4

<3

NA

female (98%)

 

Performance:

Langley et al reported an accuracy of 93.5%.

Garvin and Klales found an accuracy of 72% with a sex bias (females 94% correct, males 49%).

Three of end-points have a prediction from 69-77% for a sex so would be wrong with some frequency. The other 2 end-points show a much better accuracy.


To read more or access our algorithms and calculators, please log in or register.