Also: assuming there is not some bad link in the terrible telephone game between scientist and journalism, this seems to be a rather good article describing a study of the relationship between kids' learning mundane things like math skills, and their beliefs about cognition and learning.
The story seems a nice capsule of how science ought to work even: Scientist interested in some thing. Scientist grabs a bunch of existing data, to see if there is a correlation. Oh my, yes! There is. Kids who think that intelligence is more flexible and acquired tend to have better grades. Scientist thinks, could this be causal? Let's design an experiment to find out. The article even mentions what was done to the control group: they had a presentation on good study skills, while the experimental group was taught a little about neurobiology and told above all that your brain changes when you learn new things. Result: experimental group has better math grades.
Even if this is a totally valid result and I haven't failed to think of some confound (which I am suspicious I have, especially because I want this result to be true) there would be some work left fine-tuning what it is that we should be telling kids about their own brains for maximum effect. Would even delivering false information, being encouraging past the point of correctness, still be effective?