69 of 74 people found the following review helpful
selecting on the dependent variable,
By A Customer
This review is from: In Search of Excellence: Lessons from Americas Best Run Companies (Paperback)
I don't have much to add to the other reviews here on the content, but as a couple of reviewers here have pointed out, there's a problem with the way they reached their conclusions. They chose a series of metrics as indicators of "excellence": they ranked companies on these metrics to identify a sample of "excellent companies": they then profiled these companies to find common features. Statisticians call this "selecting on the dependent variable": all excellent companies might have a certain feature, but you can only say that the feature has something to do with their excellence if non-excellent companies don't have it. The features that Peters picks out might be important, but the research they do doesn't in any way prove that.
There was a follow-up piece of research done some years later (not by the authors) in a paper called "excellence revisited", which argued that excellence was basically a temporary phenomenon, and that even these companies reverted to the mean. This looked at the "excellent companies" subsequent performance and found that on average, they had deteriorated significantly in all measures of performance. They then picked a sample of "non-excellent companies" using the same ranking criteria as the original book did at the time that the original research was done. Sure enough, these on average improved significantly in performance.
Sort: Oldest first | Newest first
Showing 1-1 of 1 posts in this discussion
Initial post: Feb 2, 2015, 8:09:27 AM PST
Last edited by the author on Feb 2, 2015, 8:09:56 AM PST
At the highest level, it all looks like investing, in which in the long run, everybody is average. The notion that excellence is a temporary phonomenon seems to be many natural examples. In "Discipline of Market Leaders", I think half of the companies cited as "leaders" were no longer in such a position 5-7 years later. In this decade, it is astonishing to see Nokia and Motorola fall so quickly. However, the "dependent variables" are the lessons here, not the companies that may have possessed them, temporarily.
‹ Previous 1 Next ›