RESUMO
Bone heat-induced changes complicate osteometric stature estimation of human remains from forensic settings. The validity of current methods has not been tested to a great extent. Our aim was to determine how precise are stature estimations obtained from three different approaches, namely by using (i) Rösing's method (Rösing 1977), (ii) a 10% shrinkage correction factor (Strzalko et al. 1972) and (iii) chemosteometry (Gonçalves et al. 2020). For this purpose, pre- and post-burned head measurements from the humerus, radius and femur were used. The sample comprised 46 human skeletons of known sex and age-at-death. These were experimentally burnt to maximum temperatures ranging from 700 to 1100°C (attained after 90-188 min) for other research purposes. Stature estimations were performed through measurements in both pre-burned and burned bones using the three approaches and based on the method of Olivier and Tissier (1975). Mean absolute differences and the relative technical errors of measurements (TEM%) between the pre-burned and the estimations were calculated. Absolute mean differences indicated that the chemosteometric approach allowed for potentially more precise stature estimations than the other two procedures. However, the TEM% was very low for all approaches (smaller or close to 1%), and stature estimations were thus well within the error margin reported by Olivier and Tissier (1975). Results suggest that the chemosteometric approach enables more accurate predictions of the actual heat-induced shrinkage of each bone thus rendering more precise stature estimations. Nonetheless, the other procedures also provided quite reliable estimations although they require confirmation that the bone is calcined.
