If there were separate scattergrams for UCLA L&S and UCLA engineering (or each engineering major), that may show how much of the overlap is due to selectivity differences by division or major, versus by evaluation of subjectively-graded aspects of the applications (e.g. essays).