Red Flags
The Red Flags section offers guidance on areas of public opinion research where findings may be misleading, unstable, or easily misinterpreted.

Public Agenda uses several indicators to judge when survey results should be reported and used cautiously:

Results change when survey questions are reworded slightly

Results change when implications or trade-offs of a policy are pointed out.

Results may be misleading if reported in isolation or out of context

Other research suggests that people have incomplete or inaccurate knowledge in this area.

Affirmative Action: What a Difference a Word Makes. Survey responses on affirmative action and increasing diversity on campus vary dramatically depending on how questions are worded and whether they emphasize the goal of diversity or focus explicitly on the issue of preferences. Surveys also suggest that many Americans may not completely understand much of the legal debate and press coverage of the issue. For example, surveys show that Americans have different definitions of what the words "affirmative action" mean and fairly large numbers say they aren't sure. Majorities of Americans say they support programs that offer "assistance" for minorities in college admissions or jobs, but support drops dramatically if the question is reworded to ask about "preferences." …
http://www.publicagenda.org/issues/red_flags.cfm?issue_type=higher_education