This is anecdotal, but I have also noticed the related inverse problem as well. Students who think they are "one of the dumb kids" do much worse than similar students who do not think that. These students will randomly guess the answer to a multiple choice[1] question if they don't instantly know the answer, even if the answer to the question was basically given in another question on the test. Part of helping this k…
It's amazing how few students are actually as bad as they appear to be on first pass. I basically go into every student/teacher interaction assuming the kid is above average, and I'm very rarely disappointed. That's because "average" is often pulled way down by issues surrounding self-image and study behavior, both of which have no bearing on raw ability.
Basically, "confidence + hard work" is usually enough to take a kid from below average to slightly above average.