Authors get paid when people like you upvote their post.
If you enjoyed what you read here, create your account today and start earning FREE BLURT!
Sort Order:  
  ·  2 years ago  ·  

True, knowing that, the design of experiments (DOE) can randomize the results during data collection. This way the results are only extracted at the last step.

I Love it when the answer is Not what I expected, that's where the treasure lives!

👍💗😁🤠💙🎉

  ·  2 years ago  ·  

it would also go a long way towards advancing "confidence"

if there was (a hell of a lot) more focus on sigma


Posted from https://blurtlatam.intinte.org

  ·  2 years ago  ·   (edited)

Scientific prostitutes offend me at a base level! I have told clients Not to tell me what they expect I'll find. Then I randomize the data collection process. I have fixed a number of production processes, by identifying a true root cause; and ignoring what they have tried to use in the past!

👍💗🤠💙💩🎯

  ·  2 years ago  ·  

i love it


Posted from https://blurtlatam.intinte.org

  ·  2 years ago  ·  

It is fun to turn their processes over and rebuild them using their own data! They are usually surprised and happy....

👍💗🤠💙🎯💲

  ·  2 years ago  ·  

i remember figuring out what "sample-bias" is when a friend of mine explained how the market-research company they work for designs the surveys to "make the customer happy"


Posted from https://blurtlatam.intinte.org

  ·  2 years ago  ·  

A fine example of prostituted results! The risky part is when the customer thinks the results are honest, and invest heavily on the product involved. Some companies get lucky, some just survive, but the rest crash and burn!

There is always a cost from bad science, it is mostly hidden; by it always lethal!

👍😭🤢🤮💲🤢