Wilcoxon signed-rank test was used to compare the peak heights obtained in the sensitivity study and inhibitor tolerance study, as well as the number of alleles detected for the real casework study. For multiple comparisons due to different concentrations of DNA and inhibitors, Holm-Bonferroni correction was employed. Repeated measures ANOVA was used to compare the heterozygous balance and stutter ratios of the two kits with the following stratifications: kit within concentration and kit within locus, respectively. ANOVA assumptions were checked with normal quantile-quantile plots (sample quantiles against theoretical quantiles) and homoscedasticity plots. A p-value of < 0.05 was considered statistically significant.
3. Result