225Analyses of Single Samples
due to chance. The K-S procedure tests for the largest difference between the
cumulative frequency of the theoretical distribution and the cumulative fre-
quency of the observed sample data. This means that the largest difference is
based on the average (cumulative frequency) of the distributions. The value
of this largest difference is called the maximum deviation (D) as specied in
Figure8.28.
Critical values for the sampling distribution of maximum deviations under
the null are known and have been calculated based on sample size. If D is
derived through hand calculation, critical values of D are displayed in a table
by alpha and N for fewer than 35 samples. For larger samples, the critical
value at alpha = 0.05 ...