282 • Alan H. Goodman 



Table 1. Comparison of mean ages at death for individu£ds by cultural horizon and number 

 of hypoplasias-stress periods at 3.5-7.0 years developmental age (see text) 



Sample Mean S.D. 

 size 



l-way A priori contrasts (T-values) 



ANOVA 



(F-mtio) A vs B A vs C A vs B + C 



Late Woodland 



Hypoplasias (A) 



1 Hypoplasia (B) 

 2-3 Hypoplasias (C) 



M. A. L. W. 



Hypoplasias (A) 



1 Hypoplasia (B) 

 2-3 Hypoplasias (C) 



Middle Mississippism 



Hypoplasias (A) 



1 Hypoplasias (B) 

 2-3 Hypoplasias (C) 



TOTAL SAMPLE 



Hypoplasias (A) 



1 Hypoplasia (B) 

 2-3 Hypoplasias (C) 



.35 



1.44 



6.52 



4.99 



.55 



1.22 



1.53 



2.25" 3.50" 



1.84* 3.04" 



.55 



1.69 



3.52 



3.08 



Notes: + = 2-tailed p< .10; * = 2-tailed p< .05; ** = 2-tailed p< .01; '** = 2-tailed p< .001 



years (Goodman, Armelagos, and Rose 1984). However, all 

 individuals yielded a complete record of stress-hypoplasias 

 for the seven half-year periods from 3.5 to 7.0 years. 



The mean age at death of individuals in the three cultural 

 horizons and in the total sample is compared for those indi- 

 viduals with no defects-stress period (total n = 50, 45.0%). 

 one defect-stress period (total n = 45, 40.5%), and two or 

 more defects-stress periods (two defects: total n = 14, 

 12.6%; three defects: total n = 2. 1 .8%) (Table 1 , Figure 2). 

 The lowest overall frequency of defects is found in the LW 

 group. All individuals in this sample have either one or no 

 hypoplasias-stress periods between 3.5 and 7.0 years. Indi- 

 viduals with one hypoplasia-stress period have a slightly 

 greater mean age at death (34.7 years) than individuals with 

 no hypoplasias-stress periods (31.6 years). This difference, 

 however, is not statistically significant (F-ratio = .35). 



This association between hypoplasias-stress periods and 

 longevity is reversed during the MALW periods. The mean 

 age at death of individuals without hypoplasias-stress periods 

 is 36.6 years, or 5.5 years greater than those with one 

 hypoplasia-stress period (31.1 years) and 8.0 years greater 

 than those with two or more stress periods (Figure 2). 



This inverse association between stress periods and mean 

 age at death is most pronounced during the Middle Mississip- 

 pian. The mean age at death of individuals without hypo- 

 plasias-stress periods is 37.5 years, or 7.3 years longer than 

 those with one hypoplasia-stress period and 15.7 years long- 

 er than those with two or more hypoplasias-stress periods 

 (Figure 2). A one-way ANOVA, testing for the statistical 

 significance of differences in ages at death among hypo- 

 plasia-stress period groups in the MM (Nie et al. 1975) 

 yielded an F-ratio of 6.52 (Table I; p < .01). 



Furthermore, I tested for thz specificity of differences be- 

 tween groups in the MM with a series of a priori contrasts 

 (Table 1). These provide a comparison of the mean age at 

 death in group A (no stress periods) with: (1) group B (one 

 stress period), (2) group C (two or three stress periods), and 

 (3) group B -(- C combined (one or more stress periods). For 

 the MM group, all a priori contrasts yielded statistically sig- 

 nificant results at the .05 level of confidence. The most sig- 

 nificant differences are found in comparing individuals with- 

 out any stress periods with those with one or more or two or 

 more stress periods (t = 3.50 and 3.52, p < .001 and < 

 .001). 



Zagreb Faleopalhology Symp. J9S8 



