28 
BSH Report No. 56 
Var(Jf) = ^pi ■ (it - p) 2 . 
1=1 
m = ^2?* ■ x i 
t=i 
Formula 1: Top: variance, bottom: 
weighted average. 
• The chain length is the random value x, 
• Pj represents the normalised Intensities of 
n-alkanes. 
• p is the weighted average (wa) 
• a = yj(variance) the standard deviation 
(stdev) 
Similar samples show Identical bell shapes as n-alkane Intensities describe, 
ideally, the same curve at the same position In the chromatogram. Conse 
quently, those derived values are similar as well. In Fig. 5, standard deviations 
from all samples are plotted against their corresponding weighted averages. 
Fig. 17: Weighted average plotted against standard deviation for all samples. 
As can be seen at first glance, the majority of the a/b pairs overlap or are 
located in very close proximity to each other. This confirms the similarity of the 
duplicates. Samples 1 to 3 (first set) from all participants are located in the 
lower left corner of the plot (ranges from 2.5 - 3.8 and 26 - 27.5 for standard 
deviation and weighted averages, respectively). 
However, there are larger differences in the data among the second and third set. 
One reason for this discrepancy might be the nature of the paraffin wax samples 
of this set, as the content of higher boiling n-alkanes (hba) is, especially in set 2,