On this page
Use identical fictional component inputs
For this numerical exercise, a fictional issuer has component scores E = 80, S = 40 and G = 60 on a stipulated common 0–100 scale. Method A assigns weights 50%, 25% and 25%. Method B assigns 20%, 50% and 30%. The component inputs are held fixed.
Calculate each weighted sum
Method A gives 0.50×80 + 0.25×40 + 0.25×60 = 65. Method B gives 0.20×80 + 0.50×40 + 0.30×60 = 54. The 11-point difference is caused solely by the changed weights in this exercise.
| Component | Input | Method A contribution | Method B contribution |
|---|---|---|---|
| E | 80 | 40 | 16 |
| S | 40 | 10 | 20 |
| G | 60 | 15 | 18 |
| Total | 65 | 54 |
Inspect which input each method emphasizes
If S rose from 40 to 50 while the other inputs stayed fixed, Method A’s score would rise by 2.5 points and Method B’s by 5 points. A larger change under Method B follows from its larger S weight, not from a different observed change in S.
Separate methodological disagreement from data quality
This calculation shows that different scores do not logically prove a data error. It does not prove that real disagreements are always caused by weights: providers can also use different scopes, inputs and treatment of missing information. Review those conditions before comparing scores or combining managers’ reports.