Re: Sum and Combine doubts

[Rayfo...@aol.com]

Hi Alejandra,

 

Perhaps this little model can help clarify:

 

SUM now just merges the underlying calibrated distributions without offering  a probability range, which would be less than meaningful.  Sum can be used in place of Phase.

 

R_Combine merges radiocarbon determinations BEFORE calibration.  Used when the determinations are known to be from the same C14 source. A Chi Sq test is performed and the T number should be LESS than the 5% number for degree of freedom df, other wise one or more of the determinations is suspect.

 

Warning! X-Test fails at 5% - Wide 1a X2-Test: df=2 T=17.103(5% 6.0)

 X2-Test: df=2 T=0.5(5% 6.0)  (a passed test)

Combine merges distributions AFTER calibration.  Can be used when determinations are not known to be from the same C14 source but are suspected to be co-eval. Although a Chi Sq test is performed and shown, the indicator to read is the Acomb test.  Here the Acomb percentage must be GREATER than the 'An' number to pass.

 

Warning! X-Test fails at 5% - Wide 1b X2-Test: df=2 T=12.723(5% 6.0)
Warning! Poor agreement - n=3 Acomb= 8.1%(An= 40.8%)

 Agreement n=3 Acomb=118.7%(An= 40.8%)  (a passed test)
 
The following model shows all four scenarios

 Plot()
 {
  R_Combine("Wide 1a")
  {
   R_Date("Rcomb1",4770,20);
   R_Date("Rcomb2",4680,20);
   R_Date("Rcomb3",4790,20);
  };
  Combine("Wide 1b")
  {
   R_Date("Ccomb1",4770,20);
   R_Date("Ccomb2",4680,20);
   R_Date("Ccomb3",4790,20);
  };
  Label("Closer dates");
  R_Combine("Close 1a")
  {
   R_Date("Rclose1",4770,20);
   R_Date("Rclose2",4780,20);
   R_Date("Rclose3",4790,20);
  };
  Combine("Close 1b")
  {
   R_Date("Cclose1",4770,20);
   R_Date("Cclose2",4780,20);
   R_Date("Cclose3",4790,20);
  };
 };

 

Hope this helps

 

regards

 

Ray

 

In a message dated 21/11/2014 10:43:31 GMT Standard Time, alejandra....@gmail.com writes:

Hi everyone,

I am writing to you, because I have a doubt about Sum, Combine and R_Combine functions.

The first one is what is the difference between Sum and Combine functions and how the last one is calculated by OxCal. 

Finally, my last question is about Acombine test. I don't know how to interpret its results. What is the percentage to pass the test? If it pass it, can we intrepret that the samples from different sources are coeval? 

Thank you very much for your attention,

Alejandra 

--
You received this message because you are subscribed to the Google Groups "OxCal" group.
To unsubscribe from this group and stop receiving emails from it, send an email to oxcal+un...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[alejandra....@gmail.com]

Thank you very much for your attention, Ray.

Your comments are very useful.

Best wishes,

Alejandra

El diumenge 30 de novembre de 2014 12:49:44 UTC+1, Ray va escriure: