That's a example of a quantitative parameter
-30 = 10 points
31,00 = 8 points
32,00 = 6 points
33,00 = 4 points
34.00 - 2 points
So if the weighting factor for tax rate as .10 (out of a total of 1.00),
1) City-1 with a 33% rate would have a raw score of 4 and weighting would add 0.4 to the city's composite score
2) City-2 with a 31% rate would have a raw score of 8 and weighting would add 0.8 to the city's composite score
>>>>>> This has the built-in assumption that low tax rates are better than high tax rates <<<<<<
Population, city size (square KM), public parks (square KM), %employment etc. could be treated similarly
City Size, weighting .02
0 - 10,000 = 10 point
10,001 - 25,000 = 9 point
...
Over 10M = 1 point
1) City-1 with 20,000 population would have a raw score of 9 and weighting would add 0.18 to the city's composite score
2) City-2 with 50,000 population would have a raw score of 5 and weighting would add 0.1 to the city's composite score
>>>>>> This has the built-in assumption that small towns/villages are better than large cities <<<<<<
Subjective parameters (quality of life, education, etc.) are a little trickier
I recommend that the data and the scoring algorithm is first, and the reporting is after you're happy with the technique
I think it's easier and more understandable if all scores be 0 - 9, but 10 brackets are not necessary (i.e. 0, 2, 4, 6, and 8 as responses)
Can you provide a small but realistic sample of your data in the format that you collect it in?
