Analysis of insect survival after they are injected with 2000 bacteria cells.

This analysis was performed using a non-parametrical Cox proportional hazard model of survival. In this analysis, we controlled for insect weight and for the number of injected CFU which were found to have no significant effect on survival. A total of trials are analyzed here, with the differences between the lots of insects used in this experiment modeled as a random block factor (with an estimated variance of 0.008). The analysis demonstrates that insects injected with selected bacteria die earlier than those injected with the ancestral lineage.