1

I have several variables that describe the quality of a trading relationship between a client and financial institutions (FIs) the client deals with. The variables are populated with observed data over a period of time. Lets call the variables A,B,C,D. The client is able to use this data to filter on FIs based on either A,B,C or D right now.

What I am trying to do is create a single rank/score which aggregates a chosen number of variables and a defined weighting to each variable based on what the client cares about. So for example, client ACME might create a rank based on xA + yB where x and y are weights applied to the two variables it cares about with other variables are ignored.

I need this rank to be simple and easy to explain to clients who are not stats experts. I welcome advice and suggestions on how to approach this problem.

kjetil b halvorsen
  • 63,378
  • 26
  • 142
  • 467
JKL
  • 21
  • 1
  • 3
  • 3
    Closely related: http://stats.stackexchange.com/questions/9358. The classic text in the theory of valuing multiple attributes is [Keeney & Raiffa](http://www.amazon.com/Decisions-Multiple-Objectives-Preferences-Tradeoffs/dp/0521438837). – whuber Apr 15 '13 at 17:10

0 Answers0