1 of 1 people found this helpful
There are two function it takes - SKEW() and SKEW.P() and the difference is that incase of SKEW only, it considers the distribution of the data as a sample and incase of SKEW.P, it considers the data as the entire population. Mathematically, in place of n if you take n-1, the formula will take the data as a population distribution. So your -0.64 comes when you apply sample skew function, and -0.54 is observed when you calculate skewness of a population.
SKEW() on a sample
SKEW.P() on a population
Hope this help.