The article says that lower prices imply lower ratings, but the data presented only shows that lower prices correlate with lower ratings. One could equally well draw the conclusion that lower-priced software is of lower quality, hence the bad ratings.