May 2014


May 12, 2014

As a corollary to a prior post I titled Site Your Source, a pet peeve of mine is that person who tries to force a point with a completely unacceptable sample size.  While siting your source is important, it is more important to draw your conclusion from a reasonable sample of the population segment you are attempting to represent.

Some statisticians think n=8 is the minimum required to draw an inference across a relatively homogeneous population. All I know is the more the merrier as people are not so similar, and the stronger you will be able to stand your position. I just want to laugh and/or cry when I ask people “how many people did you talk to?” and they reply “two.”

Your credibility will not hold up with big sweeping generalizations about “our customers, our products, or our employees” without sound data from a good sample size!