Are You Getting the Right Reliability?

The idea would be to have multiple measures of the same phenomenon, or set of phenomenadand to see where they are dovetail. So if you have multiple teams of economists and they all converge on something that approximates an agreement, then you can be reasonably that you've converged on something that's real. The critical thing for you guys to get right is the rank ordering, and not the absolute magnet. As long so as long as your method is stable across all the different domains, then the rank oarning should be relatively stable.