This article demonstrates that the effect of religion on public opinion is partially contingent on a previously overlooked variable: whether religious identifiers have accurate personal understanding of church teaching on political issues.This is an interesting and important argument, so set aside the author's love love of first person throughout the piece (I actually find that refreshing and leads to tighter writing, but some journal editors hate it).
So how does one measure religious-political sophistication? Largely by scoring respondents as agreeing with church orthodoxy, though it gets more complicated than that. As is often the case in political science research, you get bombarded by statistical models and it can appear the journal pays the author by the number of tables published. Simply put, correctly knowing the position of one's religious leaders influence's one opinion on the issue in the direction of the religion. I'm oversimplifying, but I'm rushed. Read it yourself for all the nuance and, especially, at the end the question of whether this is merely projection or a real effect on public opinion.