问题如下：
When is the CBB needed?
解释：
The CBB is generally needed when the data being bootstrapped have time series dependence. A block bootstrap is required when the statistic being bootstrapped is sensitive to the serial correlation in the data. For example, when bootstrapping the mean, the CBB is required if the data are serially correlated. If bootstrapping the variance, the CBB is required if the squares are serially correlated—that is, if the data has conditional heteroskedasticity.
