Don't sweat it. It's only a study to establish correlation. It isn't at the stage of a hypothesis. The study also mentions other competing factors such as air quality and so forth. That said they aren't saying that autism is entirely a creation of television, and that no other instances of autism occurred before the advent of television, but they are showing a strong correlation between spikes in onset and times when television was being viewed in greater quantites by children. Take a look at a book called "The Plug In Drug." It's written like a soccer mom scare book, but the research didn't seem too shaky, and it draws similar conclusions.