This article is flawed!

For example, the first one regarding the representativeness of CR readers. Why do we need the sample to be representative in this case? If general public is stupid, then we want a sample that is equally stupid to guide us in selecting reliable cars? CR readers are wealthier and more educated, so they are smarter shoppers and their opinions are what we need for selecting reliable cars. Does that make sense?