e********e 发帖数: 21 | 1 首页 - 分类讨论区 - 学术学科 - 经济版 - 同主题阅读文章 |
f*****0 发帖数: 489 | 2 from my quick reading, the umich article only stated that AJR used incorrect
/ inconsistent / inadequate methods to estimate mortality rates and once
those mistakes were corrected their analysis lacked robustness.
he did not go as far to call intentional misuse / academic fraud.
the paragraph you quoted was BEFORE the DTA analysis, and the AER polilce.
the author of the 3rd paper did analysis what happened after the AER policy
was put in place. |
z***c 发帖数: 2959 | 3 Levitt最有名的两篇文章,都是错的
很多有名文章的结果都不能replicate,以前也不要公开code,你复制不出
是你coding能力不行
后来就要求公开,结果一公开麻烦就来了有coding error的文章很多,
过去大家都你好我好,现在挑错成了潮流了
journal’s
Anderson (1986),shocked the economics profession. The vast majority of the
research could
【在 e********e 的大作中提到】 : 首页 - 分类讨论区 - 学术学科 - 经济版 - 同主题阅读文章
|
s*******e 发帖数: 122 | 4 真有这样的事情啊……我最近也在跟教授作一个项目,follow他朋友的一篇文章,但是
我使用几乎一模一样的数据和计算方法就是得不出他的结果来,难道……
journal’s
Anderson (1986),shocked the economics profession. The vast majority of the
research could
【在 e********e 的大作中提到】 : 首页 - 分类讨论区 - 学术学科 - 经济版 - 同主题阅读文章
|
d**********i 发帖数: 418 | 5 上次听说那个long-run risk model也不能被replicate。sigh |
A******g 发帖数: 4 | 6 Levitt的文章不应该说是错的吧,
只是结论不会那么显著,并且问题不在他,数据源有问题
【在 z***c 的大作中提到】 : Levitt最有名的两篇文章,都是错的 : 很多有名文章的结果都不能replicate,以前也不要公开code,你复制不出 : 是你coding能力不行 : 后来就要求公开,结果一公开麻烦就来了有coding error的文章很多, : 过去大家都你好我好,现在挑错成了潮流了 : : journal’s : Anderson (1986),shocked the economics profession. The vast majority of the : research could
|
h*****0 发帖数: 145 | 7 reduced-Form的估计应该CODING很容易吧 |
z***c 发帖数: 2959 | 8 所以我以前讲过,现在的问题是有些文章的结论too good to be true
我以前一个做生物化学的朋友说他follow一个大牛的文章做了两三年
都出不来结果,后来发现那人就是错的。他同时也说科学界很多文章
都复制不出来,我当时还说经济金融不是如此,但最近老听金融的人
讲类似的事情,我估计经济也好不到哪里去。一个差别是金融的人
处理这种事情不如经济的严肃,所以最近讲这种现象的都是经济系的老师
在fight
【在 d**********i 的大作中提到】 : 上次听说那个long-run risk model也不能被replicate。sigh
|
l********s 发帖数: 430 | |
e******c 发帖数: 3 | 10 没听错吧
这个不是theoretical paper吗
calibrate一下应该没那么容易出错吧
【在 d**********i 的大作中提到】 : 上次听说那个long-run risk model也不能被replicate。sigh
|
|
|
S****Y 发帖数: 4634 | 11 就没几个paper能完全replicate的
【在 z***c 的大作中提到】 : Levitt最有名的两篇文章,都是错的 : 很多有名文章的结果都不能replicate,以前也不要公开code,你复制不出 : 是你coding能力不行 : 后来就要求公开,结果一公开麻烦就来了有coding error的文章很多, : 过去大家都你好我好,现在挑错成了潮流了 : : journal’s : Anderson (1986),shocked the economics profession. The vast majority of the : research could
|
e********e 发帖数: 21 | |
s*******e 发帖数: 122 | 13 别的地方不知道,我是见过不要求的,但是如果发表了人家来向你质疑总得拿出结果来。
~~
【在 e********e 的大作中提到】 : EDSG
|
z***c 发帖数: 2959 | 14 会计也一样
【在 S****Y 的大作中提到】 : 就没几个paper能完全replicate的
|
S****Y 发帖数: 4634 | 15 我说的就是accg/finance这些
【在 z***c 的大作中提到】 : 会计也一样
|
s******r 发帖数: 11 | 16 Claker, I would consider this if I were you:
I assume your are not any one of the editor and referees, and you will not
take any credit or debt if the paper finally get published. For if yes, you
should just do what you are supposed to do.
Suppose it is published one day and you still do not agree with the authors
and can not stay away. Then you can just write a short article about your
points and arguments and send it to the editor of the journal. If the editor
likes your idea, it will be quick |
c****r 发帖数: 75 | 17 Thanks for the advice--really appreciated it. My concern is that they will use fake data now, and then I will have no way to write a critique, since I don't have the raw data. I only have their original "forthcoming" paper, which they will change
soon. Or, they may just delete the relevant graphs that show their argument
is completely wrong; and since it's already an accepted paper, it'll be
published and no one can tell what's wrong any longer.
you
authors
editor
【在 s******r 的大作中提到】 : Claker, I would consider this if I were you: : I assume your are not any one of the editor and referees, and you will not : take any credit or debt if the paper finally get published. For if yes, you : should just do what you are supposed to do. : Suppose it is published one day and you still do not agree with the authors : and can not stay away. Then you can just write a short article about your : points and arguments and send it to the editor of the journal. If the editor : likes your idea, it will be quick
|
h*****n 发帖数: 924 | 18 I cannot find your original post, and guess you want to uncover some problem
s of a lousy paper.
I assume that you are a junior person. If you are not, ignore my reply.
my 2 cent is never to touch such things. First and foremost, it is bad for y
our career. They will hate you and seek revenge in the future. second, it is
somewhat unlikely to affect them. Third, it is equally unlikely that you ge
t any credit. Even you are proved to be right, many would think that you are
some junior person who w
【在 c****r 的大作中提到】 : Thanks for the advice--really appreciated it. My concern is that they will use fake data now, and then I will have no way to write a critique, since I don't have the raw data. I only have their original "forthcoming" paper, which they will change : soon. Or, they may just delete the relevant graphs that show their argument : is completely wrong; and since it's already an accepted paper, it'll be : published and no one can tell what's wrong any longer. : : you : authors : editor
|
c****r 发帖数: 75 | 19 Thanks for your advice. What happened was that I found a number of critical
errors in a forthcoming article--basically the author's own data shows their
argument is completely wrong, and now they are probably going to fake
something in order to make the article look better. I know I should not try
to establish any enemy, but on the other hand it's really difficult to
swallow it when they do this in front of your eyes. I was just wondering if I should warn him/her in private (politely, of course)
【在 h*****n 的大作中提到】 : I cannot find your original post, and guess you want to uncover some problem : s of a lousy paper. : I assume that you are a junior person. If you are not, ignore my reply. : my 2 cent is never to touch such things. First and foremost, it is bad for y : our career. They will hate you and seek revenge in the future. second, it is : somewhat unlikely to affect them. Third, it is equally unlikely that you ge : t any credit. Even you are proved to be right, many would think that you are : some junior person who w
|
d**********i 发帖数: 418 | 20 具体看John Campbell最近的一篇paper,他说不能replicate
【在 e******c 的大作中提到】 : 没听错吧 : 这个不是theoretical paper吗 : calibrate一下应该没那么容易出错吧
|
d****n 发帖数: 23 | 21 wise guy!
problem
y
is
ge
are
Unrobus
【在 h*****n 的大作中提到】 : I cannot find your original post, and guess you want to uncover some problem : s of a lousy paper. : I assume that you are a junior person. If you are not, ignore my reply. : my 2 cent is never to touch such things. First and foremost, it is bad for y : our career. They will hate you and seek revenge in the future. second, it is : somewhat unlikely to affect them. Third, it is equally unlikely that you ge : t any credit. Even you are proved to be right, many would think that you are : some junior person who w
|