Statistical modelling
We can similarly analyse annual adult mortality figures. We find once more a highly significant effect of tsetse control on annual mortality of bulls. Tsetse control decreased average annual male mortality from 0.22 to 0.08 (SED 0.03). Note that we use the standard error of the lagged parameter estimate for the standard error of the difference between the two predicted least squares estimates. There was no significant interaction with period when the interaction was first included in the model and so we can conclude that similar reductions in mortality occurred when either of the two control methods was applied.
The other variables contained in CS10Data5 can be analysed in a similar way. The student can tackle these in the study questions.
|