Newmark's Door

"Statistical control usually doesn't make an analysis causal and it can easily mislead."

(I especially appreciate the discussion of "overcontrolling" about halfway through the piece.)

Related: "It Can't Be A Coincidence!"

     

Related Stories

• "Wharton's Jeremy Siegel: There would've been a July rate cut had we known about the weak jobs data"
• "Free Markets, Economic Growth, and the Restoration of Civic Culture"
• "The Problem With Grocery Stores Isn’t Profits. It’s Reality."

 

See all current articles »