Drawing the right conclusions from the numbers (again)

Posted a comment  again (scroll to bottom)  on the LSE's British Politics and Policy blog. I sincerely hope it isn't taken personally. It is not intended to be anything other than a comment on the facts of the matter and the consequences of essentially arbitrary decisions about the way to measure things (like the gap between different groups' participation rates in HE and trends therein). Our old friend the 1958 Birth Cohort features as does the degree of belief we should attach to the results for one data point and the desirability of taking into account all the relevant data, not just those bits of it that are consistent with your prior.