Relating to the five heads of qualitative output criteria listed in Table I,
contingency tables were built up for each input variable, showing the
percentage of improvements in the subclass. The chi-square statistic was
computed to ascertain whether the input variable has a significant impact on
the output criteria. This non-parametric approach was adopted, since the
output criteria were qualitative in nature. The input variables were varied at
two levels in each case. The results obtained in terms of the chi-square statistic
are presented along with the level of significance in brackets in Table III. The
values underlined, correspond to those cases where the level of significance is
below 5 per cent.