More than 550. Considering lots of universities have 10,000s of students its an incredibly small sample size for one university let alone all of the US. The study says so itself that its a very small size. I would say a size of at least 1/5 of all students would make a study more credible