Organizations who make their living off of data from public opinion polls do good work - e.g., Gallup, Roper etc. Campaign surveys, interest group surveys, media organization surveys which come from these survey professionals who operate under the standards and social science norms established and supported by AAPOR are trustworthy. However, as Dude notes, not all organizations and political operatives do this. The good news is that it never lasts long because they get "outed" for doing it. The robo call or bad survey ends up in the press and they are embarrassed. It can still influence people though. Tends to happen less in lobbying settings because if you as an interest group give bad data to a politician or bureaucrat who then trumpets it and gets called out, the interest group will have damaged the relationship.
All that being said, most small institutions, like schools, think they can write surveys because they are unaware of how easy it is to write bad questions. There are so many ways, often inadvertent, which bias responses.
For the ipad original question - there are stock survey questions that can be found online which would allow more appropriate answers - plus there is also the fill in the blank question. Textbooks often highlight these types of bad questions with the classic - how often do you beat your dog - it does not allow the respondent to say you do not do it or have ever done it - particularly if a 0 answer is not provided
DeHe
