Oded Shenkar

To cognitive scholars, the correspondence problem is the central scientific puzzle in imitation. This challenge of converting the imitation target into a copy that will preserve the favorable outcome observed in the original is as relevant for business as it is for the sciences, and no company can hope to become strategically agile without being able to decipher it. For instance, many firms copied General Electric’s three-rung, performance- evaluation system, only to find out that what worked for GE did not work for them. In other words, there was no correspondence that would make the model relevant and applicable in the imitator’s context. Without deciphering causality in the model, it is impossible to establish causality in the recipient system.

Like this content? Why not share it?
Share on FacebookTweet about this on TwitterShare on LinkedInBuffer this pagePin on PinterestShare on Redditshare on TumblrShare on StumbleUpon

Leave a Reply Cancel reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.