THE MARKER SYSTEM - HOW IT APPEARS TO WORK

The Dar data is/was processed using the Symphony spreadsheet and the 'Marker' system, for which absolutely no documentation has been found. Verbal description and a brief look at the Symphony spreadsheets lead one to assume the following:

- approx 2 - 3 times a year a large sample is processed, and

the results of this processing are incorporated as

bogus records in with the actual data collected for each

market-day.

- this enables GSC's to be reported on which are not present

in the sample for the market-day

- they also however skew the data for all other grades and sexes from the pure calculated values

I am currently (February 1995) awaiting a written description of the Marker System from Jim Airey by request through Percy Robb of TLMP.

It appears that the Arusha data does not use any such system for quoting on non-sampled GSC's or for otherwise altering the pure calculated values, but this possibility was not investigated to any great extent.