Note that the article does not do a breakdown by subject, an interesting ommission. It is possible that the data might show girls out performing boys because t

Note that the article does not do a breakdown by subject, an interesting ommission. It is possible that the data might show girls out performing boys because they are taking easier courses, so in aggregate it looks like they're ahead, since a high grade in $hardsubject is taken as equivalent to a high grade in $easysubject. Hard to know without the data to analyse for this.