« Normal Distribution III | Main | Lognormal Distribution »
September 27, 2005
Normal Distribution IV: Skewness
Plotting the frequency distribution of our random data set is quite straightforward and allows us to visually evaluate how symmetric about the mean our data is.We will also look at a function that measures how symmetric a data set is.
We have already met the FREQUENCY function in a previous tip
In the example below, we set up our intervals in B22:B30 and the FREQUENCY array function is added to C22:C30
The final step is to create a chart from the newly created frequency table.
A function that measures how symmetrically distributed a set of data is is SKEW. For perfectly symmetric data, the SKEW function returns 0.
In the plot below, the data is tailing off to the right - giving a positive skew. This is confirmed by the value of the SKEW function.
In the next tip, we will show how to generate a skewed set of random data - known as the Lognormal distribution
Posted by Dave at September 27, 2005 08:55 PM