The null hypothesis was accepted in the test to compare cycle time means for the two
operators (Table 5). However, when comparing fuel consumptions with the null
hypothesis, 0 : A B H , the hypothesis was rejected. The research team then proceeded
to test the hypothesis that operator A was consuming more fuel/cycle than operator B
(null hypothesis and corresponding alternate hypothesis shown in parenthesis in Table 5).
Again, there was enough evidence to reject the null hypothesis. One would have to
conclude based on these t-tests at 95% confidence, that: (i) means of cycle times for the
two operators are equal; (ii) means of fuel/cycle for the two operators are not equal; and
(iii) the mean fuel/cycle for operator A is not greater than the mean fuel/cycle for
operator B. This leads to an inconclusive overall conclusion. On the one hand, cycle
times for the two operators are similar but there are indications that the fuel/cycle is not
the same. Yet, one cannot definitively say that the fuel consumption of operator A is
higher than that of operator B. More data over a longer period, and possibly involving
15
more operators, is needed to better characterize the impact of operators on fuel
consumption. Given the foregoing, the research team concluded that different trucks and
different operators made no significant difference and, hence, all data will be treated as
one population.