4 ms·
I love the recent one on piketty. Step 1: Decide on ideological argument and proposed policy remedies; Step 2: Manipulate data as needed to justify; Step 3: Pu
by kenrikm 12y ago
I love the recent one on piketty.
Step 1: Decide on ideological argument and proposed policy remedies; Step 2: Manipulate data as needed to justify; Step 3: Publish book...
Step 4: Ideological allies receive book rapturously, say it cannot be challenged due to quantitative rigor of underlying data analysis...
Step 5: Data analysis carved up by outside analysts; Step 6: Response from allies = "Data & analysis don't matter -- we KNOW the truth."
- caseydurfee 12y agoDoes anyone have a link for Step 5? Or step 6, for that matter? I don't know what he's referring to here.
- declan 12y ago#5 is likely the FT independent analysis here, which found math errors, cherry-picking of data (to meet manufactured-in-advance conclusions), etc.: http://www.ft.com/cms/s/2/e1f343ca-e281-11e3-89fd-00144feabdc0.html#axzz32Tbqv13O http://www.ft.com/cms/s/2/e1f343ca-e281-11e3-89fd-00144feabd... Though over at MR, they noted there "were already several major problems" with Piketty even before the FT debunking, including that if you adjust for the housing bubble a few years ago, his results disappear: http://marginalrevolution.com/marginalrevolution/2014/05/what-do-the-piketty-data-problems-really-mean.html http://marginalrevolution.com/marginalrevolution/2014/05/wha...
- gohrt 12y agoIf that comment were originally posted on HN, it would deservedly be downvoted as a mischaracterization.