This site uses cookies. See our Terms of Service and Privacy.

Outliers

Posted by Peter Donovan 14 years, 8 months ago

In general, statistical accuracy increases with the square root of sample size. Doubling your sensitivity and accuracy quadruples your cost. It's a power law, not a normal distribution, and it pushes us toward extremes.

In measuring soil carbon using traditional sampling, what this means is that the high achievers are easiest and cheapest to measure (circled red in the diagram below). A sampling scheme that is adequate for measuring a large change in soil carbon between an initial baseline and resampling, may not yield a significant result if the change turns out to be small.

Measurement will thus tend to highlight the more successful strategies, and the more motivated and successful managers.

Seth Godin, author of the book TRIBES, writes, "there is no tribe of normal. People don't coalesce into active and committed tribes around the status quo."

realities of sampling