This preview shows page 1. Sign up to view the full content.
Unformatted text preview: Necessary conditions for an experiment to be informative o Non-satiation Youre always trying to get more o Saliency Something significant that shows how you doing the experiment will benefit you o Dominance Payoffs of experiment must outweigh any transaction costs o Privacy Avoid any external interactions when payoffs are known to other subjects o Parallelism Can you make inferences from your experimental results to a naturally occurring market or system...
View Full Document
- Spring '11