Paterson (1982) used water dissolved in organic solvents and water dissolved in glass to generate a calibration for water in quartz. His method has been widely adopted to more general studies of 'water' in minerals.
Here is a comparison of our water contents in olivive determined with NRA analysis with the same data analyzed with Paterson's method using, 1) unpolarized light, and 2) with three polarized directions.
In general, we feel that a single conversion factor between the Paterson method of estimating the water content and the analytical results determined by NRA does not exist. Yet, in the case of olivine, the constant factor is remarkably good.