Statistic errors in Business / Data Analysis

Taken from:

http://www.information-management.com/newsletters/statistics-analytics-data-quality-mistakes-10021691-1.html?ET=informationmgmt:e2823:2262124a:&st=email&utm_source=editorial&utm_medium=email&utm_campaign=IM_Daily_122711

1. Sophistication in statistics compensates for lack of data and/or business understanding. - hey - I can do the second derivative and rotate around the mean, and do advanced analysis - but ... what? You only wanted the average - or what does this mean? [Many years ago, teaching mathematics, students would divide 1000 by 50 and get something like 2000 - a little common sense would have helped!!]

2. Extracting meaning out of randomness. - there are times where there really isn't a pattern in the data (or ... you are not using the correct data - or not using the correct data in a meaningful fashion) - this might be another version of #1 - what is the business understanding - what does this really mean and imply?

3. Correlation versus causation – modeling will help uncover causal relationships. - so which came first - the chicken or the egg? Did A CAUSE B - or did B cause A - or are they just related? Did Steve fall in love with Sara; or did Sara fall in love with Steve or did they both fall in love with each other? (the article has this illustration:)