Ted Gup in the Chronicle of Higher Education repeats some of the usual fake statistics aimed at promoting conservatives as victims on campus: “The American Council of Trustees and Alumni found that almost half of students believed in-class discussions of politics were too one-sided. Almost one-third said they felt they had to agree with the professor to get a good grade.”
I can find nothing in ACTA's 2004 survey about one-sided classroom discussions; 49% did agree that “some panel discussions and presentations on political issues seem totally one-sided,” but even I would agree with that. Anyone with a strong point of view who gives a speech on campus is going to be one-sided.
It's true that 29% did agree, “On my campus, there are courses in which students feel they have to agree with the professor's political or social views in order to get a good grade." However, ACTA surveys never ask students for their own experiences; they always ask students to speculate about whether there are some courses where this happens, which makes them entirely worthless.
In an ACTA survey in Missouri, 51 percent agreed that “On my campus, there are courses in which students feel they have to agree with the professor’s political or social views in order to get a good grade.” Omitted in the report (unless you examine the full list of questions in the appendix) was this question: “On my campus, some professors penalize students because of the student's political or social views.” Only 12 percent of students agreed with this secondhand rumor that “some” professors do this, even though it seems almost identical to the question that got 51 percent support.
A 2008 Georgia study that ACTA did not conduct found much more ambiguous political results. Only 23% of students reported having to agree with a professor to get a good grade. 25.2% of Republicans (vs. 19.3% of Democrats) reported this. For two-thirds of the students, this only happened once in college (and it was less likely to happen frequently to Republicans: 1.8% of Republicans, but 6.2% of Democrats, reported it happened 5 or more times). Part of the party difference may reflect a desire by Republicans not to have their views challenged. Only 40.5% of Republicans (vs. 59% of Democrats) considered it important for instructors to challenge their beliefs. More Democrats (22.8%) than Republicans (16.8%) reported "low" respect for the political views of all students.
Of course, even if we did find more conservatives than liberals report feeling bias, it would not be proof of it. Self-reported bias is highly unreliable, and the fact that creationists might feel uncomfortable in a biology class discussion about evolution is the fault of their own ignorant and unscientific beliefs, not political bias by professors. A student who devoutly believes in astrology is not necessarily the victim of political bias in an astronomy class, even if they feel like they must betray their own ideas and agree with scientific reality in order to get a good grade.