Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: rectangulizing data


From   Dmitriy Krichevskiy <[email protected]>
To   [email protected]
Subject   Re: st: rectangulizing data
Date   Thu, 26 May 2011 15:20:37 -0400

I do think we mean different things by "nationally representative". In
my view there is no such thing as representative sample; agreeing to
participate in these survey(s) is not random.There are however nearly
representative samples. I am trying to figure out if this makes sense.
It is volatility we see months to month or wave to wave that is
enormous. For example, in my sample there is a 23 percent chance of
moving down from middle quintile every month. I cannot get my head
around that...

Thanks again for you comments and your paper
Dmitriy

On 5/26/11, Dmitriy Krichevskiy <[email protected]> wrote:
> I do think we mean different things by "nationally representative". In my
> view there is no such thing as representative sample; agreeing to
> participate in these survey(s) is not random.There are however nearly
> representative samples. I am trying to figure out if this makes sense.
> It is volatility we see months to month or wave to wave that is enormous.
> For example, in my sample there is a 23 percent chance of moving down from
> middle quintile *every month*. I cannot get my head around that...
>
> Thanks for you comments and your paper
> Dmitriy
>
>
> On Thu, May 26, 2011 at 2:30 PM, Austin Nichols
> <[email protected]>wrote:
>
>> Dmitriy Krichevskiy <[email protected]> :
>> The SIPP sample is nationally representative (at least in wave 1). Yours
>> is
>> not.
>> Look at job start and end dates; there is much of the story on
>> intrayear volatility.
>> And yes, it is much higher than you think it is based on your
>> convenience sample.
>> I will not take a position on how you should proceed with your research.
>>
>> On Thu, May 26, 2011 at 1:44 PM, Dmitriy Krichevskiy
>> <[email protected]> wrote:
>> > Thanks for the links Austin,
>> > I am lucky to find an expert on both the matter and the data. I am
>> > basically finding similar results as your paper though my interest is
>> > in self-employment vs. wage work. I am still puzzled by the enormous
>> > volatility an average individual faces. Out of the small sample of
>> > people I personally know no one is subjected to such fluctuations,
>> > most certainly no one working for a wage. This makes me uncomfortable
>> > because either I  and the people I know are not a good representative
>> > of an average individual or (more worrisome scenario) those
>> > participating in SIPP are some strange individuals self-selecting to
>> > participate.
>> >
>> > Back to the issue at hand: would you abandon attempts to calculate
>> > annual income, impute or drop people missing several months?
>> > By abandoning I presume switching to 4 month cumulative (or average)
>> intervals.
>> >
>>
>> *
>> *   For searches and help try:
>> *   http://www.stata.com/help.cgi?search
>> *   http://www.stata.com/support/statalist/faq
>> *   http://www.ats.ucla.edu/stat/stata/
>>
>


-- 
Dmitriy Krichevskiy Ph.D. Candidate
Economics Department
Florida International University
www.fiu.edu/~dkrichev

Research Associate, College of Education
Lumina Foundation Project
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/


© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index