Thanks for the interesting article. Sometimes I worry that data is skewed and whatever point is trying to be made will be supported by data that may only be "half truth". If one studied the boston plumbing reviews in order to have valid data all the reviews have to be looked at not just a particular subgroup.