I'm interested in seeing this study... I'd also like to see the breakdown in tribal percentages and such (was one tribe represented more than another, etc). Also, was the poll regarding the term "redskins" or just NA mascots, in general?
I have to go back into my university's database to find my earlier research sources, but I remember numbers that wholeheartedly contradict the 9-13 percent found in both this Annenberg poll and an SI poll (conducted with half the sample size a year earlier). The percentage of college mascots and images that have been transformed or removed altogether dwarfs their professional sports counterparts. The NCAA has started to crack down because the activism is coming directly from college campuses.
Post article on polls
This article even shows how several polls reveal totally opposite results within a NA sample (not exclusively activists, either). One polls contradicts the other with entirely flip-flopped percentages. I'm guessing methodology (sample selection, survey language, etc) varied immensely between the Annenberg/SI polls and the Indian Country poll.
You can't discount cultural factors, as well. Rob Schmidt (who represents activism through his comics... much like Aaron McGruder does with "Boondocks") states, "Native people are known for not trusting Anglos and not speaking openly to them in initial encounters. What steps did Harris Field Center take to overcome this propensity? Can Harris assure us the Native people questioned said what they really think? That they didn't say mascots are okay to avoid 'making waves'?"
I believe polls surveying opinion/emotion have much higher probability of flawed methodology and possibly bias (the way a question is loaded has everything to with results), as opposed to polls based on hard numbers like scores, salaries, etc. Here's a link to a number of educational research based on scientific methodology regarding this very issue (skip the APA statement and go to the citations for the actual studies):
APA Resolution