Re: Consulting Time
- From: Art Kendall <Arthur.Kendall@xxxxxxxxxxx>
- Date: Wed, 17 May 2006 01:38:54 GMT
<insert tongue in cheek>
"never" may be too strong a word.
In consulting on 200 or so doctoral dissertations, and a thousand or so Congressional Investigations since 1972, I have had 2 instances where the data was as clean as the client told me.
<remove tongue from cheek>
My best guess is that the data cleaning/prep/exploration time can be as low as 80% of analyst time when the data comes from from Census or NCHS.
I have seen it exceed 95% on several occasions.
90% would probably be a good rule of thumb.
Art
Social Research Consultants
John_Kane wrote:
.
My,not too extensive, experience suggests that you should, at the very
least, triple the time you estimate to clean-up or reorganize the data.
Data is never as clean or nicely layed out as the client tells you it
is.
- References:
- Consulting Time
- From: Brett Magill
- Re: Consulting Time
- From: John_Kane
- Consulting Time
- Prev by Date: Re: Consulting Time
- Next by Date: Re: friedman's test
- Previous by thread: Re: Consulting Time
- Index(es):
Relevant Pages
|