I managed to find some comparative [Scotland/England] data to add to my recent report:
‘Glass half full or half empty? The Markit July figures are looking healthy.’
The author, Chris Peace, is clearly better than I am at finding such information from the Markit report. The comparative figures, where 50 would mean no change are:
Permanent Jobs Temporary Jobs
Scotland 66.2 70
England 62.5 62.7
I have to admit that I don’t know just how significant those differences are but they’re in the right direction for the Scottish economy and presumably further evidence of good government. They’re clearly not percentages. Here’s what Wikipedia has to say:
‘Formula, calculation, and reading
PMI data are presented in the form of a diffusion index, which is calculated as follows:
P M I = ( P 1 ∗ 1 ) + ( P 2 ∗ 0.5 ) + ( P 3 ∗ 0 )
- P1 = Percentage number of answers that reported an improvement.
- P2 = Percentage number of answers that reported no change.
- P3 = Percentage number of answers that reported a deterioration.
Thus, if 100% of the panel reported an improvement, the index would be 100.0. If 100% reported a deterioration, the index would be zero. If 100% of the panel saw no change, the index would be 50.0 (P2 * 0.5).
Therefore, an index reading of 50.0 means that the variable is unchanged, a number over 50.0 indicates an improvement, while anything below 50.0 suggests a decline. An index of 50.0 would arise if either all respondents reported no change or the number of respondents reporting an improvement was matched by the number of respondents reporting a deterioration. The further away from 50.0 the index is, the stronger the change over the month, e.g. a reading of 55.0 points to a more frequently reported increase in a variable than a reading of 52.5. The degree of confidence experienced by respondents reporting an improvement and the degree of concern experienced by respondents reporting a deterioration are not factored into the index.’
There you go. That’s that cleared up, I think. Does that mean that if the range is only from 50 to 100 that the gaps between the Scottish and English figures are actually twice as big as they would be if they had been percentages based on a 0 to 100 scale?
I could be all at sea here. Readers good at maths; are you out there?