"The survey was conducted online, with an e-mail sent to 19,843 possible participants. Of the 19,843 possible participants in the study, 1,175 completed at least a portion of the survey. The final sample of students who completed the entire survey was 910. More than two-thirds of the participants were undergraduates, and 32.1 were graduate students."

Ignoring the issue of over-representation of graduate students, this survey had a woeful response rate of 4.5 percent.  For this kind of online survey, which is very much like a mail-in survey for methodological purposes, one needs a very high (75 percent or so) response rate to allow any conclusions to be drawn about the statistical universe being sampled (i.e., the whole student body).  A much better methodology is actually to draw a random sample of students (about 500 would more than do) and then call them and ask them the survey questions over the phone.  Or, work very hard to get significant numbers of this 500-person sample to answer an online survey.  A major methodological problem with the type of online survey conducted here is a phenomena known as "self selection into the sample."  Simply put this means that the researcher does not select the sample but respondents choose whether or not to participate.  As a general rule what happens in this situation is that you get a survey of extreme positions.  In other words, those motivated to participate tend to be have strong opinions on the issue -- either for or against.  You tend to miss those in the middle ground who care less passionately.  

For academics to tout such a survey as a "research study" is laughable.  I know Rich Vedder, and respect him and consider him a friend, but let me suggest thet he stick to a subject he knows better -- economics.  If he wants to do a legitimate survey, I suggest he contact the Bush Research Insititute in the School of Journalism.  I believe Guido Stempel, emeritus professor, is still active with it.  He's a nationally recognized expert in this area.