The American history quiz that students from American colleges, including the most prestigious schools, all failed (on average), can be found here. I got a 95%, which seems about right. I mostly missed some monetary policy questions at the end, even though I had narrowed the choices. On the other hand, I got a few right where I had narrowed the answers and then picked what looked likely.
The basic story is that not only did students average less than a 65% - in some cases much less - but at the "best" schools, there was a tendency for seniors to do worse than freshmen, indicating that they had lost ground while going to Ivies and Ivy-wannabees. Yale, Duke, Cornell, Princeton, and UPenn actually damaged their students. When one looks at the questions most frequently missed, it becomes apparent that the problem is not forgetfulness, but reversing correct answers in favor of more politically convenient myths. The questions most missed included the understanding of just war theory, separation of church and state, income distribution, and Plato's ideal of government. What are the odds. eh?
Two encouraging notes: several schools that my sons considered and their friends went to (their college was not in the survey) did better than most, and going there improved average scores: Grove City, Wheaton, and Calvin. A strong correlate with higher scores was growing up in intact families where intellectual issues were often discussed. Those students not only entered college with higher scores, but learned more while at school, even though they were no longer at home.
I was dying to know if conservative students did better than progressive ones, because the intact family and evangelical school correlates would certainly suggest that, but political preference was not included in the survey. It also didn't measure whether history majors did any better than the others.
That college is often a place to learn the template rather than, y'know, learn stuff might explain the first story here.