From the actual article (and not the news article): "The results show that students’ change in programming patterns is only weakly predictive of course performance"

Not as "juicy" as the website - but then again, it's the actual results :)