Methodology: adjusted versus unadjusted gap

Every year the study reports an unadjusted gap and an adjusted one, and every year members quote whichever suits them. Before the 2027 round I think the council should settle how we define and present the two.

My concern is the adjusted figure. Last year we let respondents report it using whatever controls they chose, and the number of controls ranged from three to nineteen. An organisation that controls for level, function, tenure, performance rating and location has explained away much of the unadjusted figure, including things that may themselves reflect bias.

I propose we ask for a standard minimum set and report anything further separately. Views by the 21st, please.