Introduced on day 3
Standard error
The standard deviation of a statistic's sampling distribution; for a mean of n independent values it is sigma / sqrt(n).
In detail
It measures how much an estimate would vary across repeated samples, not how much individual observations vary. It shrinks with the square root of the sample size, so quadrupling n halves it.