Local Standards or Informative Priors in Applied Forensic Anthropology? - PowerPoint PPT Presentation

1 / 50
About This Presentation
Title:

Local Standards or Informative Priors in Applied Forensic Anthropology?

Description:

Local Standards or Informative Priors in Applied Forensic Anthropology? ... L.A. Forensic. 232. Korean War (KIA) N. Sample. Korean War Dead Transitions ... – PowerPoint PPT presentation

Number of Views:25
Avg rating:3.0/5.0
Slides: 51
Provided by: konig5
Category:

less

Transcript and Presenter's Notes

Title: Local Standards or Informative Priors in Applied Forensic Anthropology?


1
Local Standards or Informative Priors in Applied
Forensic Anthropology?
  • Lyle W. Konigsberg, Ph.D.
  • University of Tennessee
  • Knoxville, Tennessee

2
95 Ranges (Katz Suchey)
3
Problems With Using Percentiles of Age
  • ) Percentiles are estimates that may require
    standard errors.
  • ) Percentiles give an incomplete description.
  • ) Percentiles depend both on the biology of
    aging and on the demographic composition of the
    reference sample.

4
Phase I
5
Phase I
6
Phase I
7
Phase I
8
Phases V VI
9
Phases V VI
10
Reference Samples
Sample N
Korean War (KIA) 232
L.A. Forensic 737
Korean War (POW) 126
Terry Anatomical 422
Kosovo (M F) 226

11
Korean War Dead Transitions
12
Cumulative Probabilities
13
Universal Weights (Likelihood)
14
Phase V VI
15
Phase V VI
16
Phase V VI
17
Transition I/II
18
Transition II/III
19
Transition III/IV
20
Transition IV/V
21
Phase I Posterior
22
Phase II Posterior
23
Phase III Posterior
24
Phase IV Posterior
25
Phase V VI Posterior
26
Survivorships
27
Survivorships
28
Survivorships
29
Survivorships
30
Prior Age-at-Death Distributions
31
Phase I Posterior
32
Phase I Posterior
33
Phase I Posterior
34
Phase II Posterior
35
Phase II Posterior
36
Phase II Posterior
37
Phase III Posterior
38
Phase III Posterior
39
Phase III Posterior
40
Phase IV Posterior
41
Phase IV Posterior
42
Phase IV Posterior
43
Phase V VI Posterior
44
Phase V VI Posterior
45
Phase V VI Posterior
46
Conclusions
  • ) It is possible for Lyle to give a paper
    showing absolutely no equations!
  • ) For large representative reference samples,
    use survivorship within each phase.

47
Conclusions, cont.
  • ) For Bayesian analyses use an appropriate
    prior.

48
Age-at-Death Distribution
49
Conclusions, cont.
  • ) If appropriate reference material is not
    available, it may be possible to use information
    from large generic samples.

50
Acknowledgements
  • Dr. Nicholas Herrmann
  • Dr. David Hunt
  • Erin Kimmerle
  • Dr. Daniel Wescott
  • Dr. Judy Suchey
  • National Science Foundation (SBR-9727386)
Write a Comment
User Comments (0)
About PowerShow.com