Skip to main content

View Diary: Researchers finally replicated Reinhart-Rogoff, and there are serious problems. (124 comments)

Comment Preferences

  •  don't blame Excel (6+ / 0-)

    the researchers got the result they wanted through good ol' GIGO and the magic of hand-waving.

    •  It's a combination. And rather than GIGO... (1+ / 0-)
      Recommended by:

      ...there was a case of selective self-serving data exclusion, and a weird, indefensible (and also self-serving) weighting scheme - or rather lack thereof.

      btw, my boss just forwarded me this snippet that suggests that the "London Whale" story is also related to Excel spreadsheet fun.

      Bottom line, whether or not you have nefarious intent, Excel-based spreadsheet analysis is not a serious scientific tool. And I say that as someone who actually likes Excel; it's by far the best of Microsoft Office.
      It's great for small-scale to medium-scale data manipulation, for making pretty tables, or for minor accounting tasks. But for real research or anything else that requires proper documentations - it's simply the wrong tool.

      •  Yes, the whole thing (0+ / 0-)

        makes me suspicious that it could have been intentional rather than GIGO.

        We'll just exclude a little of this, a little of that, drop a few countries here and there, and voila!  We have the results we want!

        The banks have a stranglehold on the political process. Mike Whitney

        by dfarrah on Tue Apr 16, 2013 at 03:48:22 PM PDT

        [ Parent ]

    •  The averaging scheme is the worst (3+ / 0-)
      Recommended by:
      eyesoars, happy camper, Sychotic1

      I call fraud.  The way data is averaged is almost certain to cause skewing of the data in the direction of the minority response.  Ye Gods of Statistics, I'd flunk an undergraduate who used a similar scheme on biological data.

      Why do I suspect conservative economists have no understanding (or no interest) in an honest analysis

Subscribe or Donate to support Daily Kos.

Click here for the mobile view of the site