Like many statistical conclusions, the devil is in the details.
The time frame measured is six years after ENTERING the college. Thus, a student who took six years to complete an undergraduate degree AND who has either entered or is looking to enter the work force is included. Conversely, students who are in graduate school (and presumably adding to their skill level which in some way translates into a higher salary) ARE NOT included. Eliminating from analysis the highest future wage earners, creates a sloppy, tilted playing field.
The first few years after graduating are generally spent rapidly building skills that are actually useful in the work place. As skills are built, promotions and raises follow. Comparing a HS graduate SIX years after graduating with a college grad TWO years after graduating seems designed gloss over that inconvenient fact. (In other words, the HS grad has accumulated six years work experience while the college grad has two year accumulated experience [assuming the student actually completed a four year degree in four years].)
(Going through a California state university in four years is quite a feat - due to "impacted majors" many Califorina students can't finish their degree in four years.)
The survey would have been more valuable if it compared the wages of each category six years after finishing schooling; then noting the additional investment of funds needed to get the degree, thereby leaving the reader with an understanding of how much (if anything) ROI the degree represents.
A working life is a marathon, not a sprint, so metrics which focus on the very beginning of the working life are, IMO, all but useless.