Using Statistics to Prove a Point

http://docs.google.com/Doc?id=dgnwtft2_12gffj246t

This page from Good Magazine was a large spread filled with various statistics regarding the "average" American's sugar consumption. The article states "between 1970 - 2003, annual U.S. per capita sugar consumption increased by 19%." Decision Modeling has a lot to say about this kind of blanketed statement.

For one thing, what is the standard deviation of this average American's consumption use? What is the variance? Is this a normal distribution of are there tails to consider?

Another question to consider when seeing these large general assessments is for each statistic is the "Average American" the same data point?

What would this distribution look like if it was not "average." I used the handy dandy BestFit tool to assign some probability by teaspoons and lo and behold this did not come up to look like a normal distribution. See http://docs.google.com/Doc?id=dgnwtft2_16fq2pxrgp

Lessoned learned from this article was too make sure and always match what you hold to be possibilities against what other people claim to be "normal." "Normal" is only what we assign to something as a term to describe a mean but the distribution around that mean could vary greatly.