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: RE: perform regression on minimum number of observations stata


From   Nick Cox <[email protected]>
To   "[email protected]" <[email protected]>
Subject   Re: st: RE: perform regression on minimum number of observations stata
Date   Mon, 29 Jul 2013 16:20:15 +0100

Gregorio's loop can be modified something like this

forval y=1989/2012 {
      forval  i= 1/57 {
        di "year = `y' and industry = `i'"

      reg DV IV_1 IN_2 IN_3 if  Industry== `i' & year==`y' & nobs>15
}
Nick
[email protected]


On 29 July 2013 16:15, Nahla Betelmal <[email protected]> wrote:
> Thanks Gregorio, but how can I exclude the regression with less than
> 15 observation due to missing data rather than no observation? any
> idea!
>
>  many thanks again for the great help. from what I have read in the
> field the magic number is 15, one or two papers uses 20. Maybe due to
> the nature of the data availability from the population dataset.
>
> Thanks again for the kind help
>
> Nahla
>
> On 29 July 2013 15:57, Impavido, Gregorio <[email protected]> wrote:
>> You could modify your loop as follows:
>>
>> bysort year industry: gen nobs = _N
>> forval y=1989/2012 {
>>       forval  i= 1/57 {
>>         di "year = `y' and industry = `i'"
>>       reg DV IV_1 IN_2 IN_3 if  Industry== `i' & year==`y' & nobs>15
>> }
>>
>> However, I thought that the "magic number" for the central limit theorem to apply was 30.
>>
>> -----Original Message-----
>> From: [email protected] [mailto:[email protected]] On Behalf Of Nahla Betelmal
>> Sent: Monday, July 29, 2013 10:32 AM
>> To: [email protected]
>> Subject: st: perform regression on minimum number of observations stata
>>
>> Hi Statalist,
>>
>> my data consists of several years and several industries (each industry contain several firms). I run  cross-sectional regressions as follow :
>>
>> forval y=1989/2012 {
>>       forval  i= 1/57 {
>>            reg DV IV_1 IN_2 IN_3 if  Industry== `i' & year==`y'
>> }
>>
>> I have two questions please:
>>
>> 1- According to literature, any industry-year regression with less than 15 observations (i.e. 15 firms) should be disregarded. the tricky thing is that almost every industry-year group has more than 15 firms but some are with missing values to perform the regression. Therefore, the regression could be done with less than 15 observation. How can I account for this in the above loop, so regression with less than 15 complete observations produce missing values instead of actual values!
>>
>> 2- the output does not provide information about the matched year and industry. Is there a way to do it?  I only need that for reporting purpose.
>>
>> Thank you
>>
>> Nahla Betelmal
>> *
>> *   For searches and help try:
>> *   http://www.stata.com/help.cgi?search
>> *   http://www.stata.com/support/faqs/resources/statalist-faq/
>> *   http://www.ats.ucla.edu/stat/stata/
>>
>> *
>> *   For searches and help try:
>> *   http://www.stata.com/help.cgi?search
>> *   http://www.stata.com/support/faqs/resources/statalist-faq/
>> *   http://www.ats.ucla.edu/stat/stata/
>
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/faqs/resources/statalist-faq/
> *   http://www.ats.ucla.edu/stat/stata/

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/


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