The Pew Center used a phone survey of a nationally representative sample of 2,142 adults ages 18 and older. Pew also used an online survey done in association with the Chronicle of Higher Education, among the presidents of 1,055 two-year and four-year private, public and for-profit colleges and universities.
It makes sense that they used these methods because it allows for a large sample and quick responses. It was also important that they used these methods because those surveyed who did not attend college are more likely to respond honestly if their identity is somewhat concealed. Yet, it is important to note how this concealed identities may have also skewed the results. It is impossible to clarify that the telephone respondents were indeed 18 or older.