Normal population
The observations with each data set should be from normally distributed
populations. This assumption is usually relaxed when there are equal observations in
each data set, which is not the case in this analysis. Therefore, a one-sample
Kolmogorov-Smirnov test was conducted on each sample set to check the validity of this
assumption. The following null hypothesis was tested in this analysis.
H0: There is no difference between this sample and the population in term
of normality.
The test resulted in significant P-values (Table 8, Appendix) for Kuwait and the
USA. Adopting a 5% significance level, both P-values were greater than 0.05, therefore
the null hypothesis was accepted and the normal population assumption was fulfilled.