R******o
发帖数: 83 | 1 能举例子最好。谢谢各位了。
1. Difference between data cleaning and data validation?
2. What is ad hoc analysis? |
x***x
发帖数: 3401 | 2 1. data validation is like auditing, you need to compare data with other
sources to make sure the data is correct. data cleaning is just a general
term of making changes to data (to make the data correct)
2. ad hoc means something specially designed to fit a purposes. |
f****r
发帖数: 1140 | 3 I thought ad hoc means something diverse, or whatever coming up.
Maybe I am wrong:)
【在 x***x 的大作中提到】
: 1. data validation is like auditing, you need to compare data with other
: sources to make sure the data is correct. data cleaning is just a general
: term of making changes to data (to make the data correct)
: 2. ad hoc means something specially designed to fit a purposes.
|
R******o
发帖数: 83 | 4 Thanks a lot. Could you give more details? especially about ad hoc? |
g******n
发帖数: 339 | 5 In pharmaceutical industry, any analysis that is not Planned (in protocol)
before the data are collected are adhoc analysis.
【在 R******o 的大作中提到】
: Thanks a lot. Could you give more details? especially about ad hoc?
|
R******o
发帖数: 83 | 6 Thanks a lot! Does that mean 'Simulation'??
Using computer generated data to do some theoretical analysis instead of
using real data??? |
D******n
发帖数: 2836 | 7 1) they are obviously different if u check out the meaning of validation and
clean.
2)ad hoc, means u can't find any known procedure or algorithm for solving
the problem at hand. You kinda analyse the problem , and pound your head
against the wall and come up with a set of procedures or analyses that might
solve the problem or shed some light on it.
Ad hoc in Latin means "for this", so no one can tell you what ad hoc is, it
is up to you, and the problem u r working on.
just my 2 cents. |
R******o
发帖数: 83 | 8 Oh. That made some sense now...Thanks. |
R******o
发帖数: 83 | 9 Ad hoc, For example, you got some data, but nobody had given you the
protocol or guide to analyze them. So you started to look at the data and
came up with some strategies.
This does not necessarily mean the problem is complicated.
Is this interpretation correct? |
R******o
发帖数: 83 | 10 再确认一下，在clinical study里， 是不是没有Statistical Analysis Plan的就是ad
hoc analysis, 有SAP的就不是？？？？？ |
n*****g
发帖数: 112 | 11 a study has to have a SAP. It could either a standalone SAP or embedded in
the protocol. All the analyses included in the SAP are the pre-specified
analyses; any additional analyses done after the data is unblinded are ad
hoc.
ad
【在 R******o 的大作中提到】
: 再确认一下，在clinical study里， 是不是没有Statistical Analysis Plan的就是ad
: hoc analysis, 有SAP的就不是？？？？？
|
R******o
发帖数: 83 | |
