Sheffield ‘turned down £10 milllion’ for apprenticeships

Correlation is not causation. It's not clear from the article whether assessment in the universities in the study is anonymised. I'm pretty sure though that in all of my experience as a student and teacher in the social sciences, there has been a clear correlation between where students sit and their marks. This is because students sort themselves according to engagement - the keen ones sit at the front and the less keen at the back (including those arriving late). It's not because the ones at the front are the lecturer's 'favourites'. It would be interesting to see what would happen if seating was randomly assigned, or to experimentally test different seating arrangements, but I suspect it wouldn't have much effect.