Paleodemography: Science, Art or Artifice?
Can the demographic features of a population be recovered from the age distribution of skeletons?
PPT Slide
Take the best dataset: Libben (Ohio), 800-1100 AD. skeletons by estimated age
Add The Whopper assumption:
My Approach: graphics over statistics
Meta-analysis of 40 paleodemographic datasets: all periods, places and populations
Three recent advances in paleodemography:
What’s new (in this presentation but not in the paper)? Three extras for the diehards:
Demographers know: fertility has a big impact on population age structure (and on the age distribution of deaths). Next figure shows fertility effects:
Fig. 1. Fertility has big effects on age structure of deaths
Great variations in mortality scarcely affect age distribution of skeletons. The next figure shows mortality effects (stable populations):
Fig. 2. Mortality offers a small target
Use faux hazard rates (central death rates) for data, models and graphs
Let’s accept a modified Whopper assumption
Example: The best dataset in North America Libben (Ohio) 800-1100 AD
Draw empirical hazard “rates” and confidence intervals over stable population “rates”
Not convinced? Consider conventional graphics (percent of deaths aged 5-14, 15-24, …45+) drawn over stable populations.
Following graphs use faux hazard rates instead of percentages
Fig 7. Age bias in skeletal datasets , but not historical ones.
Fig. 8. European prehistoric populations also show age bias
Another example: Skeletal data for the Oneota
Fig. 9 Oneota. Note poor fit between data and models
Next set of figures: faux hazard rates by ancestry--all show age bias
Regardless of ancestry, there is a consistent pattern of unprecedentedly high hazard rates from age 25-35 and beyond
Lovejoy adjustments are insufficient to overcome bias (remaining years of life from age 15)
Conclusions, I: 5 reasons why graphs reveal poor fit and statistics do not
Conclusions, II: paleodemography is methdologically challenged.
Science, Art or Artifice?
Two paleodemographic age ratios: Bocquet-Appel (D5-14)/D20+ Buikstra (D30+/D5+)
Email: rmccaa@maroon.tc.umn.edu
Home Page: http://www.hist.umn.edu/~rmccaa/
Download Microsoft PowerPoint Animation Player for ActiveX
Download presentation source
