Data often misses values because of errors or technical issues. Even if we are not missing values, we may have cause to suspect certain values. Once we doubt data values, derived values such as the arithmetic mean, which we learned to calculate in this chapter, become questionable too. It is common for these reasons to try to estimate how reliable the arithmetic mean, variance, and standard deviation are.

A simple but effective method is called Jackknife resampling (see http://en.wikipedia.org/wiki/Jackknife_resampling). The idea behind jackknife resampling is to systematically generate datasets from the original dataset by leaving one value out at a time. In effect, we are trying to establish what will happen if at least one of the values is wrong. For each new generated dataset, we recalculate the arithmetic mean, variance, and standard deviation. This gives us an idea of how much those values can vary.