September 2004 www.pqsystems.com Vol. 6, No. 9
Quality Quiz
from Professor Cleary

Congratulations!
You're right!

For a more comprehensive explanation of skewness, click here for a video discussion.

It is clear that Marge Orrine has no clue about the meaning of skewness, which implies "screwed up" to her. Help was right at her fingertips, with a skewness explained in her CHARTrunner help file, but she failed to use this resource.

Skewness relates to the shape of data, the measure of the asymmetry of a histogram (frequency distribution). A histogram that follows a normal distribution is symmetrical; that is, the same amount of data falls on both sides of the mean. A normal distribution has a skewness of 0. The chart that Marge was looking at had a skewness of -.4 which is not that different from zero. Visually it looks similar to a normal distribution.

If on the other hand your data has a long tail that goes out to the right then you would have a large positive value for your skewness, as seen in the chart below. The histogram tail goes to the right and the skewness is 1.9.

On the other hand if the tail of your distribution goes to the left then you will have a negative skewness as shown in this chart.

Of course, the other failure that Marge demonstrates lies in her intention to change the data to make it "come out right." The data itself does the talking, and Marge failed to comprehend this critical concept.

Register for the free Quality Gamebox drawing
Please submit the form only once. Multiple entries in one month will be discarded.