Fisher FERITSCOPE FMP30, Руководство оператора

Страница: 218 / 241

Operator‘s Manual FERITSCOPE
®
FMP30 Page 213
Glossary
backscatter and X-ray fluorescence methods). During normalization, the
calibration curve is adjusted to the individual measurement application
that the open Application is calibrated for. The coefficients of the adjust-
ed calibration curve are stored in the open Application. The coefficients
of the master calibration curve, which are stored in the EEPROM of the
probe plug, are not affected.
Normalized countrate Countrate
Normal distribution (Gaussian normal distribution, Gaussian distribution)
Probability distribution discovered by C. F. Gauß in 1794. If a quantity
X can be classified as having normal distribution, 68.3 % of the ob-
served values of X are within the

interval around the mean value

of
the variable X, i.e., the following applies to 68.3 % of the observed val-
ues:

-
 
X
 
+

.
This interval is indicated in Figure 16.7 by the gray area below the
curve.
Figure 16.7: Probability distribution P(X) of a quantity X, which can be classified as having
normal distribution
The probability distribution P(X) is symmetrical around the mean value

of the quantity X, which can be classified as having normal distribu-
tion.
Skewness and Curvature are zero for the normal distribution.
The populations, which are tested for technical purposes, often can be
classified as having approximately normal distribution. However, the
following fact is of great importance: if several random samples with
equal size are drawn of a population, and the mean values of these ran-
dom samples are determined, these mean values can be classified as
having normal distribution (Central Limits Theorem). The mean value of
these sample mean values is an estimated value (
) for the mean
value

of the population. The uncertainty of measurement u can be de-
X

-
  
+

Probability P(X)

ˆ
Fe. =