The UK media frequently publish articles reporting on research that rates places on various criteria, with indices that can be structured into league tables. Such indices are frequently based on statistical procedures that over-simplify the differences between places, assuming that averaging data on a number of different criteria presents a valid representation of a general pattern. After a critique of such methods this paper suggests an alternative procedure and applies it to the data used for the recent production of a UK Prosperity Index. It shows that the geographies of the forty-three separate variables deployed in producing that index are more complex than can reasonably be assumed.
|Journal||International Journal of Market Research|
|Early online date||3 Apr 2017|
|Publication status||E-pub ahead of print - 3 Apr 2017|