When the Bonferroni correction was applied to account for the chance of statistical significance due to multiple comparisons (p = .0018), the number of types of events that refugees were more likely to report was reduced to eight (these were the same as those with p < .001 in Table 2).