[coccoliths] How many fossils to count in a sample?

[Ali Haidar]

Dear INA subscribers,



For those who are still counting only 300 specimens (when trying to be sure at 95%) or only 500 specimens (at 99%), please note:



https://arxiv.org/abs/1804.11226v2



Any count should therefore be expanded in case you are trying to detect more than one taxon together in the same sample (same stratigraphic interval).



It includes easy-to-use tables to tell in each case how many specimens to count according to:

                - how many species are to be detected using the same count.

                - how sure you would like to be.

                - what is the expected original proportion of each species in the population.



Basically, the procedure is the same as that of Dennison and Hay (1967), but using multinomial (instead of binomial) distribution.



The tables can be easily used even if the mathematical theory provided is not understood.



Regards,



Ali Haidar.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://webmail.fcnym.unlp.edu.ar/pipermail/coccoliths/attachments/20180514/0a4f918a/attachment.html>