This paper describes our attempts to make radiation gauges and analyze
rs operate in more of a 'black box' manner-being more independent of o
r at least more insensitive to changes in such things as sample compos
ition and position. Our overall approach relies heavily on Monte Carlo
simulation and that is discussed first in relation to radiation gauge
and analyzer design and usage. Then the principles and applications o
f radiation gauges and analyzers including the Measurement Chi-Square
and Monte Carlo Library Least-Squares, respectively, are treated. Fina
lly, future work in this area is discussed. (C) 1997 Elsevier Science
Ltd. All rights reserved.