If you can read this, either the style sheet didn't load or you have an older browser that doesn't support style sheets. Try clearing your browser cache and refreshing the page.

(Reuters)   That awkward moment when you, a lowly grad student, realize you spotted a basic math error in the most influential paper of two of the most prominent theorists in your field   (reuters.com) divider line 113
    More: Interesting, term papers, graduate students, Kenneth Rogoff, Carmen Reinhart, research papers, theorists, maths, mistakes  
•       •       •

10683 clicks; posted to Geek » on 18 Apr 2013 at 12:38 PM (52 weeks ago)   |  Favorite    |   share:  Share on Twitter share via Email Share on Facebook   more»



113 Comments   (+0 »)
   
View Voting Results: Smartest and Funniest

Archived thread

First | « | 1 | 2 | 3 | » | Last | Show all
 
2013-04-18 10:53:41 AM
#DIV/0!
 
2013-04-18 10:58:21 AM
As they poured over the spreadsheets Herndon had requested

Reuters, how the hell do you f*ck that up?
 
2013-04-18 10:58:54 AM

Langston: As they poured over the spreadsheets Herndon had requested

Reuters, how the hell do you f*ck that up?


They didn't get a grad student to double-check it.
 
2013-04-18 11:04:06 AM
My login name says it all. They thought they were above fact checking.
 
2013-04-18 11:06:21 AM
I can do arithmetic pretty quickly in my head, and have gotten in the habit of always double checking the math on spreadsheets/columns of numbers shown in presentations.

I have lost count of the number of times I have caught errors, occasionally very significant ones. And it is usually the bean-counters that make the worst ones.  I shutter to think of the boneheaded decisions made over the years because someone summed the wrong column.
 
2013-04-18 11:07:40 AM
why is this headline greenlit?

it wasn't a 'basic math error'.

never mind reading comprehension subby, just stick with trolling.
 
2013-04-18 11:08:36 AM
Someone's about to get their PhD dissertation rejected.
 
2013-04-18 11:12:56 AM
So now it goes lies, damned lies, statistics, and theses by former employees of Bear Stearns and the Federal Reserve.

That sounds about right.
 
2013-04-18 11:13:30 AM
That's how I passed differential equations.  My lab partner did the DiffyQ and I would check his addition and subtraction.
 
vpb [TotalFark]
2013-04-18 11:18:13 AM
Reinhart was a chief economist at investment bank Bear Stearns

Old habits die hard I guess.
 
2013-04-18 11:20:26 AM

vpb: Reinhart was a chief economist at investment bank Bear Stearns

Old habits die hard I guess.


What Reinhart might look like:
graphics8.nytimes.com
 
2013-04-18 11:27:46 AM
A quick CTRL~ would have done wonders for these guys.
 
2013-04-18 11:51:32 AM
Reinhart and Rogoff, however, say their conclusion that there is a correlation between high debt and slow growth still holds.

Nah, you guys predicted negative growth not slow growth.

I really love this story.
 
2013-04-18 11:57:39 AM
ChangB wrote:

Hmm 2.2 versus 0.1 not a significant difference. Someone doesn't want their funding taken away.

Apr 18, 2013 7:28am EDT  --Report as abuse


Jesus Dumbass Christ on so many levels. I'm going to report the comment as abuse to the human intellect.
 
2013-04-18 11:58:32 AM

calbert: why is this headline greenlit?

it wasn't a 'basic math error'.

never mind reading comprehension subby, just stick with trolling.


knowing you need to include all the data elements when calculating an average?  I learned that math in 4th grade how about you ?
 
2013-04-18 12:05:59 PM
FTFA:"It is sobering that such an error slipped into one of our papers despite our best efforts to be consistently careful," they said in a joint statement. "We do not, however, believe this regrettable slip affects in any significant way the central message of the paper or that in our subsequent work."

From my understanding this spreadsheet in question is the basis for the entire study. So yes, yes it does call into question your conclusions based on this study in a significant way and every subsequent work based on this study.
 
2013-04-18 12:09:22 PM

sammyk: FTFA:"It is sobering that such an error slipped into one of our papers despite our best efforts to be consistently careful," they said in a joint statement. "We do not, however, believe this regrettable slip affects in any significant way the central message of the paper or that in our subsequent work."

From my understanding this spreadsheet in question is the basis for the entire study. So yes, yes it does call into question your conclusions based on this study in a significant way and every subsequent work based on this study.


Aren't these kinds of things peer reviewed too?
 
2013-04-18 12:18:02 PM

bdub77: sammyk: FTFA:"It is sobering that such an error slipped into one of our papers despite our best efforts to be consistently careful," they said in a joint statement. "We do not, however, believe this regrettable slip affects in any significant way the central message of the paper or that in our subsequent work."

From my understanding this spreadsheet in question is the basis for the entire study. So yes, yes it does call into question your conclusions based on this study in a significant way and every subsequent work based on this study.

Aren't these kinds of things peer reviewed too?


It's like peer reviewing witchcraft. Economics is no science, no matter how much folks think it is.
 
2013-04-18 12:27:43 PM

PC LOAD LETTER: bdub77: sammyk: FTFA:"It is sobering that such an error slipped into one of our papers despite our best efforts to be consistently careful," they said in a joint statement. "We do not, however, believe this regrettable slip affects in any significant way the central message of the paper or that in our subsequent work."

From my understanding this spreadsheet in question is the basis for the entire study. So yes, yes it does call into question your conclusions based on this study in a significant way and every subsequent work based on this study.

Aren't these kinds of things peer reviewed too?

It's like peer reviewing witchcraft. Economics is no science, no matter how much folks think it is.


Kind of. It is a science, it's just a social science. Like anthropology. There's no fundamental economy to the human brain, so it's inexact. Hence why a-holes shoot children and bomb marathons. Economics generally relies on the idea of rationality but we are at our very core very irrational beings with different cultural norms, so there's no way to properly measure it.

On the other hand, the math involved at the macroeconomic level should at least add up.
 
2013-04-18 12:30:51 PM
Reinhart and Rogoff, however, say their conclusion that there is a correlation between high debt and slow growth still holds.

yup
we got it pretty much completely wrong, but we are going to stand by our hypothesis, because otherwise we will look stupid
 
2013-04-18 12:31:54 PM

sammyk: FTFA:"It is sobering that such an error slipped into one of our papers despite our best efforts to be consistently careful," they said in a joint statement. "We do not, however, believe this regrettable slip affects in any significant way the central message of the paper or that in our subsequent work."

From my understanding this spreadsheet in question is the basis for the entire study. So yes, yes it does call into question your conclusions based on this study in a significant way and every subsequent work based on this study.


Well onlyin that the data doesn;t in any way match their conclusions.   But if you think this was a sort of "let's study this and then formulate a theory based on the results" sort of paper, it never was.  it was much more of a "Paul ryan called and he wants some impressive-looking charts and graphs to back up the budget the Koch brothers wrote for him" kinda exercise.
 
2013-04-18 12:36:05 PM

namatad: yup
we got it pretty much completely wrong, but we are going to stand by our hypothesis, because otherwise we will look stupid


It's also quite possible that the accident was deliberate. They faked evidence to push their agenda.
 
2013-04-18 12:37:41 PM

bdub77: Kind of. It is a science, it's just a social science. Like anthropology. There's no fundamental economy to the human brain, so it's inexact. Hence why a-holes shoot children and bomb marathons. Economics generally relies on the idea of rationality but we are at our very core very irrational beings with different cultural norms, so there's no way to properly measure it.

On the other hand, the math involved at the macroeconomic level should at least add up.


worse
the current state of macro economics has been unable to predict things.
are current theories able to predict inflation 2 years out? 1 year out?
what about stock market or housing market?
or pretty much anything?

not picking on them and their models, but in the end, they have been completely unable to predict much of anything useful.
worse, when their predictions dont pan out, people are surprised
 
2013-04-18 12:41:55 PM

Magorn: calbert: why is this headline greenlit?

it wasn't a 'basic math error'.

never mind reading comprehension subby, just stick with trolling.

knowing you need to include all the data elements when calculating an average?  I learned that math in 4th grade how about you ?


cherry-picking your data to achieve a desired result is not a math issue, it's an ethical issue.

the fact that they knew what to include and what to omit to obtain their preferred results shows a complete understanding of the math involved.
 
2013-04-18 12:45:30 PM

Magorn: sammyk: FTFA:"It is sobering that such an error slipped into one of our papers despite our best efforts to be consistently careful," they said in a joint statement. "We do not, however, believe this regrettable slip affects in any significant way the central message of the paper or that in our subsequent work."

From my understanding this spreadsheet in question is the basis for the entire study. So yes, yes it does call into question your conclusions based on this study in a significant way and every subsequent work based on this study.

Well onlyin that the data doesn;t in any way match their conclusions.   But if you think this was a sort of "let's study this and then formulate a theory based on the results" sort of paper, it never was.  it was much more of a "Paul ryan called and he wants some impressive-looking charts and graphs to back up the budget the Koch brothers wrote for him" kinda exercise.


that right there is what struck me about economics when i took some classes, even compared to sociology or anthropology. about finding your conclusion, picking your evidence, then formulating your hypothesis. it was like science in reverse
 
2013-04-18 12:48:09 PM
Most Americans figured Paul Ryan was full of sh*t anyway.
 
2013-04-18 12:53:20 PM
From what I read in the article it looks like the professors where working the data towards a conclusion instead of working the data to find a conclusion. By which I mean they wanted to find the results they got. I may be mistaken but I would also like to know who was funding their research and what political connections are in their past just in case.
 
2013-04-18 12:56:07 PM
"It is sobering that such an error slipped into one of our papers despite our best efforts to be consistently careful," they said in a joint statement. "We do not, however, believe this regrettable slip affects in any significant way the central message of the paper or that in our subsequent work."

Somehow I don't think everyone else is going to see it that way, you losers.
 
2013-04-18 12:56:53 PM

Langston: As they poured over the spreadsheets Herndon had requested

Reuters, how the hell do you f*ck that up?


Poured.

What is wrong with it. Past tense?
 
2013-04-18 12:58:14 PM
Just like the debunked vaccination-autism "study," the fact that this was proven wrong will have no impact on its usefulness to be wielded by rabid ideologues to further their agenda.
 
2013-04-18 12:58:59 PM

bdub77: sammyk: FTFA:"It is sobering that such an error slipped into one of our papers despite our best efforts to be consistently careful," they said in a joint statement. "We do not, however, believe this regrettable slip affects in any significant way the central message of the paper or that in our subsequent work."

From my understanding this spreadsheet in question is the basis for the entire study. So yes, yes it does call into question your conclusions based on this study in a significant way and every subsequent work based on this study.

Aren't these kinds of things peer reviewed too?


It might be different in economics, but peer review in hard sciences probably wouldn't have access to the actual data sets. We write the paper, submit it (along with any supplemental material), and the reviewers get that. Their job is to assess whether our methods seem appropriate, whether we've interpreted the existing literature correctly, and whether our data support the conclusions we've come to. For a newer, or highly specialized, technique, we'd submit any info that would help clarify our approach. But we're not sending all of our raw data from every step of the way. So it wouldn't surprise me if the reviewed paper simply sent their results along, but not necessarily the underlying calculations (since in *theory* they should've checked that themselves..)
 
2013-04-18 12:59:55 PM

Shirley Ujest: Langston: As they poured over the spreadsheets Herndon had requested

Reuters, how the hell do you f*ck that up?

Poured.

What is wrong with it. Past tense?


"Pored."
 
2013-04-18 01:03:00 PM

Nina_Hartley's_Ass: Shirley Ujest: Langston: As they poured over the spreadsheets Herndon had requested

Reuters, how the hell do you f*ck that up?

Poured.

What is wrong with it. Past tense?

"Pored."


Clogged pores?  Try pouring some ProActiv over those breakouts!
 
2013-04-18 01:04:38 PM

Nina_Hartley's_Ass: Shirley Ujest: Langston: As they poured over the spreadsheets Herndon had requested

Reuters, how the hell do you f*ck that up?

Poured.

What is wrong with it. Past tense?

"Pored."


More about pore:
http://www.dailywritingtips.com/poring-over-pore-and-pour/
 
2013-04-18 01:07:44 PM

PC LOAD LETTER: Economics is no science, no matter how much folks think it is.

bdub77: Kind of. It is a science, it's just a social science. Like anthropology.


I'd say -- like a lot of social sciences -- that at it's best, it's a science; and at it's worst, it's a cargo cult. I suspect it largely depends on the personalities of the people involved. Doing science looks like:
""So we pushed him and pushed him and pushed him, and after about a month of pushing him I said, 'Goddamn it, he's right.'"

A cargo cult looks like:
"We do not, however, believe this regrettable slip affects in any significant way the central message of the paper or that in our subsequent work."

I wonder how long the politicians will take to catch on that the data doesn't back the hypothesis.
 
2013-04-18 01:11:28 PM

YoungLochinvar: bdub77: sammyk: FTFA:"It is sobering that such an error slipped into one of our papers despite our best efforts to be consistently careful," they said in a joint statement. "We do not, however, believe this regrettable slip affects in any significant way the central message of the paper or that in our subsequent work."

From my understanding this spreadsheet in question is the basis for the entire study. So yes, yes it does call into question your conclusions based on this study in a significant way and every subsequent work based on this study.

Aren't these kinds of things peer reviewed too?

It might be different in economics, but peer review in hard sciences probably wouldn't have access to the actual data sets. We write the paper, submit it (along with any supplemental material), and the reviewers get that. Their job is to assess whether our methods seem appropriate, whether we've interpreted the existing literature correctly, and whether our data support the conclusions we've come to. For a newer, or highly specialized, technique, we'd submit any info that would help clarify our approach. But we're not sending all of our raw data from every step of the way. So it wouldn't surprise me if the reviewed paper simply sent their results along, but not necessarily the underlying calculations (since in *theory* they should've checked that themselves..)


As someone who works for an analytics company, let me tell you that the data sets are often wrong, or incorrectly calculated, or a number of other factors. There's nothing wrong with examining the underlying methods, BUT when it comes down to it, you've got to check the data too. Otherwise your cure for cancer is just bogus math because you missed a decimal.
 
2013-04-18 01:13:29 PM

PC LOAD LETTER: bdub77: sammyk: FTFA:"It is sobering that such an error slipped into one of our papers despite our best efforts to be consistently careful," they said in a joint statement. "We do not, however, believe this regrettable slip affects in any significant way the central message of the paper or that in our subsequent work."

From my understanding this spreadsheet in question is the basis for the entire study. So yes, yes it does call into question your conclusions based on this study in a significant way and every subsequent work based on this study.

Aren't these kinds of things peer reviewed too?

It's like peer reviewing witchcraft. Economics is no science, no matter how much folks think it is.


BURN HIM

Shirley Ujest: Langston: As they poured over the spreadsheets Herndon had requested

Reuters, how the hell do you f*ck that up?

Poured.  Pored.

What is wrong with it. Past tense?

 
2013-04-18 01:13:31 PM

namatad: Reinhart and Rogoff, however, say their conclusion that there is a correlation between high debt and slow growth still holds.

yup
we got it pretty much completely wrong, but we are going to stand by our hypothesis, because otherwise we will look stupid


Too late.
 
2013-04-18 01:14:48 PM
"Herndon said only 15 of the 20 countries in the report had been used in the average. He also said Reinhart and Rogoff used only one year of data for New Zealand, 1951, when growth was minus 7.6 percent, significantly skewing the results."

oh those silly, non-deliberate, spreadsheet errors.
 
2013-04-18 01:16:01 PM

namatad: bdub77: Kind of. It is a science, it's just a social science. Like anthropology. There's no fundamental economy to the human brain, so it's inexact. Hence why a-holes shoot children and bomb marathons. Economics generally relies on the idea of rationality but we are at our very core very irrational beings with different cultural norms, so there's no way to properly measure it.

On the other hand, the math involved at the macroeconomic level should at least add up.

worse
the current state of macro economics has been unable to predict things.
are current theories able to predict inflation 2 years out? 1 year out?
what about stock market or housing market?
or pretty much anything?

not picking on them and their models, but in the end, they have been completely unable to predict much of anything useful.
worse, when their predictions dont pan out, people are surprised


dealbreaker.com

"That's what I've been trying to tell you."

Seriously, if you give someone a narrative built on white noise they'll take it as a predictor of the future. Your best bet is to take risks with high rewards and limited losses (options and startups) or safe bets with reliable returns (municipal bonds), everything else is for rubes and governments.
 
2013-04-18 01:22:28 PM
Oh, economics.

Not like it's a real economic discipline or anything.
 
2013-04-18 01:27:35 PM

Tell Me How My Blog Tastes: PC LOAD LETTER: bdub77: sammyk: FTFA:"It is sobering that such an error slipped into one of our papers despite our best efforts to be consistently careful," they said in a joint statement. "We do not, however, believe this regrettable slip affects in any significant way the central message of the paper or that in our subsequent work."

From my understanding this spreadsheet in question is the basis for the entire study. So yes, yes it does call into question your conclusions based on this study in a significant way and every subsequent work based on this study.

Aren't these kinds of things peer reviewed too?

It's like peer reviewing witchcraft. Economics is no science, no matter how much folks think it is.

BURN HIM

Shirley Ujest: Langston: As they poured over the spreadsheets Herndon had requested

Reuters, how the hell do you f*ck that up?

Poured.  Pored.

What is wrong with it. Past tense?


Maybe they were just crying that hard with laughter.
 
2013-04-18 01:30:30 PM

bdub77: YoungLochinvar: bdub77: sammyk: FTFA:"It is sobering that such an error slipped into one of our papers despite our best efforts to be consistently careful," they said in a joint statement. "We do not, however, believe this regrettable slip affects in any significant way the central message of the paper or that in our subsequent work."

From my understanding this spreadsheet in question is the basis for the entire study. So yes, yes it does call into question your conclusions based on this study in a significant way and every subsequent work based on this study.

Aren't these kinds of things peer reviewed too?

It might be different in economics, but peer review in hard sciences probably wouldn't have access to the actual data sets. We write the paper, submit it (along with any supplemental material), and the reviewers get that. Their job is to assess whether our methods seem appropriate, whether we've interpreted the existing literature correctly, and whether our data support the conclusions we've come to. For a newer, or highly specialized, technique, we'd submit any info that would help clarify our approach. But we're not sending all of our raw data from every step of the way. So it wouldn't surprise me if the reviewed paper simply sent their results along, but not necessarily the underlying calculations (since in *theory* they should've checked that themselves..)

As someone who works for an analytics company, let me tell you that the data sets are often wrong, or incorrectly calculated, or a number of other factors. There's nothing wrong with examining the underlying methods, BUT when it comes down to it, you've got to check the data too. Otherwise your cure for cancer is just bogus math because you missed a decimal.


I don't disagree with you, I'm just telling you how it is. As a reviewer, any time you spend reviewing somebody else's work is time that isn't spent doing your own work. So yeah, you can pore over every little detail and spend months making sure they did everything right, or you can make assumptions on the types of things the authors *should've* done correctly (eg, triple-checking their own data for mathematical errors). Of course, another issue is that primary publications SHOULD NOT be treated as gospel. It's not until the results have been tested and verified by other, independent parties that they should be assumed to be solidly true.

Don't get me wrong, I get your point, but it really comes down to time management and I'm not sure the resources exist to do the type of check you're looking for. The real problem here, as I see it, is that people treated the original publication as gospel well before it had been fully tested and verified. It's almost the equivalent of reading about a promising anti-cancer drug in a mouse model, and then insisting that your doctor administer the "cure" for cancer to you immediately.
 
2013-04-18 01:31:49 PM

gameshowhost: ChangB wrote:

Hmm 2.2 versus 0.1 not a significant difference. Someone doesn't want their funding taken away.

Apr 18, 2013 7:28am EDT  --Report as abuse

Jesus Dumbass Christ on so many levels. I'm going to report the comment as abuse to the human intellect.


To me, that reads as sarcasm aimed at the two study authors, not a dismissal of the grad student's findings.
 
2013-04-18 01:34:31 PM

bdub77: sammyk: FTFA:"It is sobering that such an error slipped into one of our papers despite our best efforts to be consistently careful," they said in a joint statement. "We do not, however, believe this regrettable slip affects in any significant way the central message of the paper or that in our subsequent work."

From my understanding this spreadsheet in question is the basis for the entire study. So yes, yes it does call into question your conclusions based on this study in a significant way and every subsequent work based on this study.

Aren't these kinds of things peer reviewed too?


Their work was published in the AER's Papers and Proceedings; which is not really peer-reviewed.  It's more of a collection of 'interesting' papers, conference presentations, and other similar work.  There is very little double-checking of any work in them.
 
2013-04-18 01:35:39 PM

Bondith: Oh, economics.

Not like it's a real economic discipline or anything.


FTFM.  Kids, don't Fark while distracted.
 
2013-04-18 01:35:40 PM

YoungLochinvar: It's almost the equivalent of reading about a promising anti-cancer drug in a mouse model, and then insisting that your doctor administer the "cure" for cancer to you immediately.


This is what happens with politicians mistake faith for fact.
 
2013-04-18 01:37:47 PM

KhanAidan: bdub77: sammyk: FTFA:"It is sobering that such an error slipped into one of our papers despite our best efforts to be consistently careful," they said in a joint statement. "We do not, however, believe this regrettable slip affects in any significant way the central message of the paper or that in our subsequent work."

From my understanding this spreadsheet in question is the basis for the entire study. So yes, yes it does call into question your conclusions based on this study in a significant way and every subsequent work based on this study.

Aren't these kinds of things peer reviewed too?

Their work was published in the AER's Papers and Proceedings; which is not really peer-reviewed.  It's more of a collection of 'interesting' papers, conference presentations, and other similar work.  There is very little double-checking of any work in them.


Ooh - you know more about econ - would a full dataset typically be sent for review as well in a fully peer-reviewed work? I'm just curious more than anything...
 
2013-04-18 01:42:34 PM

bdub77: vpb: Reinhart was a chief economist at investment bank Bear Stearns

Old habits die hard I guess.

What Reinhart might look like:
[graphics8.nytimes.com image 307x315]


More like:

www.mediabistro.com
 
2013-04-18 01:47:09 PM

abb3w: PC LOAD LETTER: Economics is no science, no matter how much folks think it is.
bdub77: Kind of. It is a science, it's just a social science. Like anthropology.

I'd say -- like a lot of social sciences -- that at it's best, it's a science; and at it's worst, it's a cargo cult. I suspect it largely depends on the personalities of the people involved. Doing science looks like:
""So we pushed him and pushed him and pushed him, and after about a month of pushing him I said, 'Goddamn it, he's right.'"
A cargo cult looks like:
"We do not, however, believe this regrettable slip affects in any significant way the central message of the paper or that in our subsequent work."
I wonder how long the politicians will take to catch on that the data doesn't back the hypothesis.


This. Social sciences are messier and more tentative than the hard sciences, so you have to be that much more rigorous and skeptical about your methods and data analysis, because, yes, any old random thing really could conceivably be acting as a compounding variable when you are studying complex social systems. Remember that what gets called microeconomics or microsociology (just for examples) are analyzing events which would be considered a complex macro-scale phenomenon by most of the more reductionist natural science disciplines. That is, barring some of the geosciences like climatology and maybe environmental biology. They also deal with complex macroscale stuff, and can probably relate to the kind of headaches the social scientists have trying to rule out confounds.
 
Displayed 50 of 113 comments

First | « | 1 | 2 | 3 | » | Last | Show all

View Voting Results: Smartest and Funniest


This thread is closed to new comments.

Continue Farking
Submit a Link »






Report