...preferably one who understands statistics.
This morning I was greeted with the bizarre heading (p.2, col.5) telling me that there is 'Little Support for Welsh Devolution', followed by the headline findings of a survey by Aberystwyth's Institute of Welsh Politics that 39% of people surveyed wanted a 'parliament', 31% liked the Assembly the way it is, 15% wanted to go back to colonial rule and 10% favoured independence.
Aside from the maths (the WM reveals that 5% are don't know/refused to respond), I see that as 4/5ths of the Welsh population being in favour of devolution of power on some scale (on a 2,500 survey which is quite large for these things - a little under 1 in 1000 people in Wales).
With its PA news and lack of editorial comment or investigative journalism, Metro can often be simplistic, but 80% in favour seems a pretty hefty majority to me. Whether that means support for a further devolution of powers is a different thing altogether.
Turning to the survey itself, one problem that I often have with the methodology is with the discrete element of the questioning - would you say that you are in favour of independence or a parliament? Are these mutually incompatible? Can you not be in favour of both, especially if you perceive this as a process, not an event (c Ron Davies).
Either way, that's 49% in favour of more powers, 31% in favour of the status quo (but will that transmit itself as a yes or no vote in a referendum?) and 15% against the whole thing (I'll take that as a 'no' then!)