First the link to this week’s complete list as HTML and as PDF.
***
Whatever the risks and benefits of vaccination may be, what the study by Horne et al. really is about is the efficacy of large scale brain washing programs. So every responsible citizen and critical journalist should take these results to heart and be very alert about the methods recommended here being employed in entirely different contexts.
***
And again Miller et al. use the standard error, a measure describing not a population but an idealised, hypothetical standard member thereof. Even if their conclusions were valid, the scatter of data is so large, that the predictive value of their result is essentially nil. On the positive side, they do give us real data and not just regression results as so many others do.
***
As Epstein & Robertson is about voting outcomes, i.e. neither the group nor the individual is of interest, only their aggregate, this is one exception where the standard error is used appropriately. Nevertheless it would have been nice to be shown results and their scatter and not just tabulated regression coefficients.