As profiled today in Inside Higher Ed, the Center for Community College Student Engagement (a research center at UT Austin) released a study which shows that students who take full loads at least some semesters (preferably early on) are more likely to graduate than students who are always/mostly part-time. There are at least three plausible reasons why this might be true:
1) The more units you take the closer you are to finishing.
2) Attending full-time produces benefits beyond the accumulated credits, e.g. more interaction with faculty and classmates. They provide data in support of that.
3) People who attend full-time have the advantage of some amount of financial security and stability in their personal lives, so they can focus on school.
It appears that they did indeed ask students if they received Pell grants, i.e. they did ask about personal financial situations, but the summary that they provide says nothing about the analysis of that data, and simply says that everyone should attend full-time as much as possible. When the summary and recommendations say nothing about the analysis of financial information, it's hard to know whether they controlled for the third possibility, so it's hard to know if full-time attendance for all is a good recommendation or not. But they don't dwell on that. They just tell everyone to go as much as possible.
Sadly, this is par for the course in much of educational research.