2.3. Statistical Analysis

Because the normality of residuals was significantly different on the Shapiro–Wilk test, we used the Kruskal–Wallis test as a non-parametric test to examine whether different methoprene concentrations affected the number of feeding holes and eggs in Experiment 1 and whether the absence/presence of 20E and/or methoprene altered the number of feeding holes and eggs in Experiment 3. The two-parameter Hill equation (where a is the inflection point and n is the Hill coefficient, which controls the curvature) was fit to the data using nonlinear regression.

In addition, the paired Wilcoxon signed-rank test was used to test whether the number of both feeding holes and eggs was different after the preoviposition period among individual females used in experiments 1 vs. 2 (e.g., methoprene retreatment from Day 8 to 14) and experiments 3 vs. 4 (e.g., 20E plus methoprene retreatment from Day 8 to 24). The unpaired T-test was conducted to test whether feeding holes and eggs were different among treatments in Experiment 2. The Chi-square test was used to test whether proportions of ovipositing females, number of eggs, hatching rates, and adult mortality were different among three treatments.