Using data from Teng et al, (1992) (Figure 3.18), it can be illustrated how the sample
standard deviation is related to the population standard deviation. Figure 3.18 shows the
estimation of the undrained shear strength (Su) of clay at a particular site using three
different methods: (1) cone penetration tests (CPT) (2) vane shear test (VST), and (3)
laboratory consolidation tests based on the preconsolidation pressure (s0
p). It is seen from
Figure 3.18 that in each case the estimation can be improved by increasing the sample size
up to an optimum size of about 7. The corresponding standard deviation estimate can
possibly be interpreted as the population standard deviation. However, the best estimate
of the standard deviation that one can make varies with the specific technique used in the
estimation. Moreover, Figure 3.18 also shows that, based on the laboratory prediction
method, VST provides a much more accurate estimate of the ‘‘true’’ standard deviation of