L******r 发帖数: 141 | 1 PP是否增加RFE?很多人说会,有又很多人说是谣传。
我从www.trackitt.com/统计出来。
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
IMPORTANT!!! 觉得有用的同学,可怜我花时间统计,给个包子吧
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
+++++++++++++++++++++++++++++++++++
PP RFE EB1A EB1B
Yes Yes 31.00 28.00
Yes No 31.00 56.00
RFE Rate: 0.50 0.33
No Yes 168.00 137.00
No No 248.00 314.00
RFE Rate: 0.40 0.30
++++++++++++++++++++++++++++++++++++
结论:
在EB1A类别,PP没有明显增加RFE可能性
(Fisher's Exact Test: p-value = 0.0982)。
在EB1B类别,PP没有明显增加RFE可能性
(Fisher's Exact Test: p-value = 0.3375)
#######################################################
RE:
I did it indeed. But the data look biased.
For EB1A, 67 denied vs 353 approved
For EB1B, 18 denied vs 533 approved.
Looks like people tend to report when approved, but not to report when denied. |
c******g 发帖数: 9273 | |
c******g 发帖数: 9273 | 3 www.trackitt.com
是干啥的?
从来没去过
【在 L******r 的大作中提到】 : PP是否增加RFE?很多人说会,有又很多人说是谣传。 : 我从www.trackitt.com/统计出来。 : ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ : IMPORTANT!!! 觉得有用的同学,可怜我花时间统计,给个包子吧 : ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ : +++++++++++++++++++++++++++++++++++ : PP RFE EB1A EB1B : Yes Yes 31.00 28.00 : Yes No 31.00 56.00 : RFE Rate: 0.50 0.33
|
s*******t 发帖数: 7746 | |
s*******t 发帖数: 7746 | 5 People report comprehensive 140 and 485 timeline. you should check it out.
【在 c******g 的大作中提到】 : www.trackitt.com : 是干啥的? : 从来没去过
|
e******r 发帖数: 9977 | 6 大赞!lz再接再厉,把本版的数据也统计一下吧,哈哈
【在 L******r 的大作中提到】 : PP是否增加RFE?很多人说会,有又很多人说是谣传。 : 我从www.trackitt.com/统计出来。 : ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ : IMPORTANT!!! 觉得有用的同学,可怜我花时间统计,给个包子吧 : ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ : +++++++++++++++++++++++++++++++++++ : PP RFE EB1A EB1B : Yes Yes 31.00 28.00 : Yes No 31.00 56.00 : RFE Rate: 0.50 0.33
|
c******g 发帖数: 9273 | 7 many thanks
【在 s*******t 的大作中提到】 : People report comprehensive 140 and 485 timeline. you should check it out.
|
l******2 发帖数: 2502 | |
s***a 发帖数: 142 | 9 但是用pp和nonpp的人的潜在条件强弱导致的样本
差异无法估计
【在 L******r 的大作中提到】 : PP是否增加RFE?很多人说会,有又很多人说是谣传。 : 我从www.trackitt.com/统计出来。 : ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ : IMPORTANT!!! 觉得有用的同学,可怜我花时间统计,给个包子吧 : ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ : +++++++++++++++++++++++++++++++++++ : PP RFE EB1A EB1B : Yes Yes 31.00 28.00 : Yes No 31.00 56.00 : RFE Rate: 0.50 0.33
|
L******r 发帖数: 141 | 10 elsevier, thanks for your Bao Zi!
【在 e******r 的大作中提到】 : 大赞!lz再接再厉,把本版的数据也统计一下吧,哈哈
|
|
|
w*********m 发帖数: 4740 | 11 我早看过了说过了,只不过没做test
EB1A PP 似乎增加了一点RFE的Prob
EB1B PP似乎增加RFE的Prob更少,和其95%左右的通过率比简直可以忽略不计。
你要不再统计一下有RFE和没RFE的通过率吧
【在 L******r 的大作中提到】 : PP是否增加RFE?很多人说会,有又很多人说是谣传。 : 我从www.trackitt.com/统计出来。 : ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ : IMPORTANT!!! 觉得有用的同学,可怜我花时间统计,给个包子吧 : ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ : +++++++++++++++++++++++++++++++++++ : PP RFE EB1A EB1B : Yes Yes 31.00 28.00 : Yes No 31.00 56.00 : RFE Rate: 0.50 0.33
|
w*********m 发帖数: 4740 | 12 这个bias是没法避免的
【在 s***a 的大作中提到】 : 但是用pp和nonpp的人的潜在条件强弱导致的样本 : 差异无法估计
|
h********y 发帖数: 2092 | |
L******r 发帖数: 141 | 14 Got it. Thanks very much!
【在 h********y 的大作中提到】 : 给你发了两个包子。
|
n***p 发帖数: 7668 | 15 这样一来这种统计是没有意义的。 特别是当样本比较小的时候。
【在 w*********m 的大作中提到】 : 这个bias是没法避免的
|
b*********r 发帖数: 7139 | 16 re,弱CASE,强CASE PP结果天壤之别啊
【在 n***p 的大作中提到】 : 这样一来这种统计是没有意义的。 特别是当样本比较小的时候。
|
v*********b 发帖数: 785 | |
L******r 发帖数: 141 | 18 If we have information about applicants, such as number of papers, number of citations, etc. We could use linear regression model to take consideration of these factors. Unfortunately, how can we get such data for hundreds of people? So we have to assume those applicants have equal chance. |
P*********y 发帖数: 986 | 19 赞!
【在 L******r 的大作中提到】 : PP是否增加RFE?很多人说会,有又很多人说是谣传。 : 我从www.trackitt.com/统计出来。 : ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ : IMPORTANT!!! 觉得有用的同学,可怜我花时间统计,给个包子吧 : ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ : +++++++++++++++++++++++++++++++++++ : PP RFE EB1A EB1B : Yes Yes 31.00 28.00 : Yes No 31.00 56.00 : RFE Rate: 0.50 0.33
|
w****9 发帖数: 6338 | |
|
|
n***p 发帖数: 7668 | 21 既然你的数据不全,你的假设又不合理,那么你的统计就没有意义,而且会误导大家。
of citations, etc. We could use linear regression model to take
consideration of these factors. Unfortunately, how can we get such data for
hundreds of people? So we have to assume those applicants have equal chance.
【在 L******r 的大作中提到】 : If we have information about applicants, such as number of papers, number of citations, etc. We could use linear regression model to take consideration of these factors. Unfortunately, how can we get such data for hundreds of people? So we have to assume those applicants have equal chance.
|
L******r 发帖数: 141 | 22 This is the most complete data so far.
for
chance.
【在 n***p 的大作中提到】 : 既然你的数据不全,你的假设又不合理,那么你的统计就没有意义,而且会误导大家。 : : of citations, etc. We could use linear regression model to take : consideration of these factors. Unfortunately, how can we get such data for : hundreds of people? So we have to assume those applicants have equal chance.
|
T*******y 发帖数: 6523 | 23 Zan your effort!
Well, as it will be a pass or loss case for each individual, everyone wants to be sure about his/her decision, and this data may not be as convincing as one wants to see.
The ones who go for PP are the ones who are somewhat more confident in their cases, do not care about the money, or do not have enough time to wait. Due to the first reason, this population may have a bias towards being more likely to be approved.
We did see people with very strong evidences who got RFEs on or near the 15th day, and their RFEs passed without problems. So there is a tradeoff between saving time and spending some efforts for RFE. It may not be a bad trade-off.
【在 L******r 的大作中提到】 : This is the most complete data so far. : : for : chance.
|
n***p 发帖数: 7668 | 24 既然大家都(曾经)是做学问的人,那么当你的Data不足以得到(你想要的)结论的时
候,
我们应该怎么办呢?
假定你的帖子是向一个(比较严谨的)杂志投稿,十有八九会被拒掉的吧?
【在 L******r 的大作中提到】 : This is the most complete data so far. : : for : chance.
|
h****d 发帖数: 1305 | 25 good point!
wants to be sure about his/her decision, and this data may not be as
convincing as one wants to see.
their cases, do not care about the money, or do not have enough time to wait
. Due to the first reason, this population may have a bias towards being
more likely to be approved.
15th day, and their RFEs passed without problems. So there is a tradeoff
between saving time and spending some efforts for RFE. It may not be a bad
trade-off.
【在 T*******y 的大作中提到】 : Zan your effort! : Well, as it will be a pass or loss case for each individual, everyone wants to be sure about his/her decision, and this data may not be as convincing as one wants to see. : The ones who go for PP are the ones who are somewhat more confident in their cases, do not care about the money, or do not have enough time to wait. Due to the first reason, this population may have a bias towards being more likely to be approved. : We did see people with very strong evidences who got RFEs on or near the 15th day, and their RFEs passed without problems. So there is a tradeoff between saving time and spending some efforts for RFE. It may not be a bad trade-off.
|