Masters Theses

Date of Award

12-1996

Degree Type

Thesis

Degree Name

Master of Arts

Major

Anthropology

Major Professor

Lyle W. Konigsberg

Committee Members

William M. Bass, Richard L. Jantz, David Gerard, Esteban Walker

Abstract

This study attempted to augment the Stout and Paine (1992) and Stout and colleagues' (1996) methods of histological aging for the clavicle. In the 1992 study, the predictive equation was generated from a sample of only 40 individuals taken from an autopsy population (mean ages 28.6 years) with no prior sampling strategy. The 1996 expansion (Stout et al., 1996) tested the 1992 equation, added 41 males and 42 females from a Swiss cemetery sample to the original autopsy sample to generate a new predictive equation based on all 123 individuals.

In this study, an independent autopsy population was used to construct a sample of 95 individuals with a random stratified profile and an approximately equal number of individuals from each decade between the ages of 22 and 88 years. Roughly equal numbers of males (n = 50) and females (n = 45) were included in the sample. Construction of an independent sample assured comparability of all measures and excluded possibility of intraobserver error. This sample was used to achieve four research goals: 1) Test for sexual dimorphism in the OPD variable. 2) Generate predictive equations using inverse calibration, the method most commonly used in histological aging methods, achieved by regressing age upon the histological variable OPD. 3) Test the 1992, 1996, and current equations using both the Stout and Paine (1992) and the current samples. 4) Generate predictive equations for the Stout and Paine (1992) and current sample using classical calibration, which is regression of the OPD variable upon age.

The generation of predictive equations using both the inverse and classical calibration allowed comparison of the characteristic distributions associated with each, and tests for accuracy of their estimates.

While females exhibited a greater correlation between age and the OPD variable (r2 = .5763) than males (r2 = .2489), this difference did not result in a significant difference in the OPD variable between males and females: Based on the full sample, the relationship between age and the OPD variable was found to be highly significant (p < .0001).

A linear response and plateau function (lrp) was performed on the data to locate the most likely asymptote, a point where further remodeling activity cannot be detected. The lrp placed the most likely start of the plateau at the age of 73 years. Based upon this result, a subsample of those individuals between the ages of 22 and 73 years were used to generate all predictive equations (n = 74).

The inverse predictive equation reported by Stout and colleagues (1996) provided the most accurate age estimates for the current sample, followed by the equation produced from the subsample in the current study. The classical calibration equations generated for the current subsample and Stout and Paine (1992) samples both resulted in greater mean squared error values than the their inverse counterparts when applied to the full current sample.

Comparison of the mean squared error values for the classical and inverse predictive equations for the current subsample (n = 74) showed that while the inverse calibration equation shows evidence of slight bias, the error for the inverse calibration estimates is smaller than those for the classical equation across all age ranges (22 - 73 years). This suggest the inverse form of the predictive equation is the calibration method of choice for histological age estimation for the clavicle.

Download
Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Included in

Anthropology Commons

Share

COinS