Bootstrapping repeated measures data (effect of trypanosomiasis on plasma luteinizing hormone level in female goats)

Abstract

This is a study on the effect of trypanosomiasis on plasma Luteinizing (LH) hormone level in female goat Mixed model methodology and the bootstrap are used to analyse the repeated measures data and to increase the power of the test, respectively. Here the analysis of repeated measures had consideration for the correlation structure. The minimum value of Akaike information Criteria from the model with mean structure "Treatment.' as the only fixed effect, and compound symmetry as the covariance structure indicated the suitable model. In order to increase the power of the test, cast' resampling bootstrap technique was employed. This technique demonstrated an increase in the power such that a significant difference in plasma LH level was revealed between the infected and non-infected female goats. Hence bootstrapping can be used as a remedy for studies with small sample sizes. If it were possible, the study could he redone with a bigger sample size and also adjusting for covariates like age and parity. Such results could be generalized to a broader population of animals.