NavList:
A Community Devoted to the Preservation and Practice of Celestial Navigation and Other Methods of Traditional Wayfinding
Re: Another round on the fate of Amelia Earhart in today's news
From: Don Seltzer
Date: 2018 Mar 1, 14:31 -0500
Jantz's own argument is based upon matching the remains found to his database of 2776 forensic samples. Apparently this database was begun in 1986 and much of it came from the Univ of Tennessee and other contributing research institutions. What is key is that this database seems to be overwhelmingly 20th century American, and not representative of either other populations or bones from previous centuries. Jantz admits ,"It should be mentioned that a sample of Micronesian or Polynesian bone
measurements was unavailable to test against the Nikumaroro bones. I
consider it highly unlikely that inclusion of such a sample would have
changed anything."
To it he has added an additional set of Amelia Earhart's 'actual' bone measurements as supplied by TIGHAR's amazing photo analyst Jeff Glickman who is able to make highly accurate measurements from grainy photos through flesh and clothing.
The ensuing probability analysis is beyond my understanding, but two things seem to stand out. First, that Jantz concludes that the Nikumaroro bones are a rough match to Glickman's estimate of Earhart's arm bones, but there were 18 other samples in his database that were even closer. But he does conclude that the odds are somewhere between 84 - 154 more likely to belong to Amelia Earhart than to any random individual who ended up on the island.
Even that is not very good, as he admits "Likelihood ratios of 84–154 would not qualify as a positive
identification by the criteria of modern forensic practice, where
likelihood ratios are often millions or more." But he doesn't give up with that, because TIGHAR has already proved it likely that Earhart was on the island. This allows him to assign a 10 to one probability that Earhart was there, elevating the likelihood ratio of the match to a much more respectable 840-1540!
I am nearly struck speechless by the force of his logic.
Don Seltzer
From: Don Seltzer
Date: 2018 Mar 1, 14:31 -0500
My main takeaways from the Jantz paper:
Jantz dismisses Hoodless's original determination that the pelvic bone was from a male because the bones were all worn and eroded. Yet he then goes on to use all of the other bone measurements as completely accurate when they support his thesis.To it he has added an additional set of Amelia Earhart's 'actual' bone measurements as supplied by TIGHAR's amazing photo analyst Jeff Glickman who is able to make highly accurate measurements from grainy photos through flesh and clothing.
Among the comments in the Wikipedia article on Jantz's database are these:
"According to the authors of the program, there are limitations that should be taken into account when using this program. Some of these limitations include the fact that ForDisc will classify any unknown into the ‘closest’ group, this means that even if an individuals ethnic group or race is not represented in the database, the program will classify it to the ‘closest’ group...
...The last limitation deals with archaeological populations. This limitation is due to the fact that most of the measurements in the data set that the classifications are based on in the program are from remains that are from the 20th century, and should not be used for classification of archaeological remains...
"According to the authors of the program, there are limitations that should be taken into account when using this program. Some of these limitations include the fact that ForDisc will classify any unknown into the ‘closest’ group, this means that even if an individuals ethnic group or race is not represented in the database, the program will classify it to the ‘closest’ group...
...The last limitation deals with archaeological populations. This limitation is due to the fact that most of the measurements in the data set that the classifications are based on in the program are from remains that are from the 20th century, and should not be used for classification of archaeological remains...
A 2009 study found that FORDISC 3.0 "is only likely to be useful when an unidentified specimen is more or less complete and belongs to one of the populations represented in its reference samples", and even in such "favorable circumstances it can be expected to classify no more than 1 per cent of specimens with confidence."[10]
In 2012 research was presented at the 81st Annual Meeting of the American Association of Physical Anthropologists, which concluded ForDisc ancestry determination was not always consistent, and the programs' recommended acceptance criteria did not separate correct and incorrect determinations. The authors concluded that the program does not perform to expectations and should be used with caution."
