Phase, Sum, Sequence?

[Michael Stites]

I am definitely a newbie to the intricacies of radiocarbon analyses AND OxCal, but I have been playing around with the Phase(), Sum(), and Sequence() commands. I need some advice on the best methods for analyzing regional and site-specific sets of radiocarbon dates on temporally successive cultural complexes. I have tried the Sequence, Begin, Phase, End, Begin, Phase, End, etc. method, but I'm really not sure I completely understand the output. Geez, it feels like I'm a passenger trying to land a plane here. Thanks in advance for any advise!

[MILLARD A.R.]

> From: Michael Stites
> Sent: 16 April 2016 02:20

Have you tried Charles McNutt's "Multi-Plots and Bayesian Analysis of Radiocarbon Dates with OxCal 4.2 for PC Dummies"?
https://www.academia.edu/5758385/Bayesian_analysis


Best wishes

Andrew
--
 Dr. Andrew Millard 
e: A.R.M...@durham.ac.uk | t: +44 191 334 1147
 w: http://www.dur.ac.uk/archaeology/staff/?id=160
 Senior Lecturer in Archaeology, & Associate Director
of the Institute of Medieval & Early Modern Studies,
Durham University, UK

[Rayfo...@aol.com]

Hi Michael,

 

The manual referred to by Andrew has many useful features for a learner.  However, may I insert a note of caution.

 

The first example of three R_Dates:

 

STRATUM I 1050 +/- 50
STRATUM II 1080 +/- 60
STRATUM III 1000 +/- 60

 

does not produce the calibrated dates shown in the plot.  I believe the correct R_Date for Stratum III would need to be 1060+/- 60.  I think this is just a typo, 1000 instead of 1060, but confusing if you are following the story.

 

 

 

Furthermore the text says :

 

"There is obviously a great deal of overlap; these measurements are statistically the same at the 95% level."

 

What is omitted here, is that the broken lines under the plots are showing the 68% and 95% ranges of the calibrated dates.  In no way does it show the measurements are the same 'at the 95% level', far from it.

 

The paper then goes on to say that the 1 sigma range and 2 sigma range is equivalent to the 68% and 95% ranges.  e.g. when calibrating, it says:

 

The result will be displayed at the right (+ 1 and 2 s.d.) in the white panel.

 

In fact the white panel is showing the 68.2% and 95.4% probability ranges.  These are the preferred way of showing Bayesian probability ranges and are not equivalent to 1 and 2 sigma.

 

With these caveats, there is a lot of useful information in the paper, but it is as well to start with a clean sheet.

 

Best wishes

 

Ray

 

In a message dated 21/04/2016 09:41:53 GMT Daylight Time, a.r.m...@durham.ac.uk writes:

https://www.academia.edu/5758385/Bayesian_analysis