But they only compared it against job listings on their own association website. This seems like incredibly lazy statistical research. They didn't look hard f

But they only compared it against job listings on their own association website. This seems like incredibly lazy statistical research. They didn't look hard for a number, but instead accepted the first, easiest number they could generate as a "rough figure."