Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | David Hoaglin <dchoaglin@gmail.com> |
To | statalist@hsphsun2.harvard.edu |
Subject | Re: st: Regression question |
Date | Sat, 26 May 2012 20:43:09 -0400 |
Hi, Mike. The situation often turns out to be more complicated than the initial question reveals. As a quick answer, I can see a couple of possibilities. One would combine the media types into categories for which all (or, at least, most) companies have data. You might have to omit companies with only 1 type, but you might get an indication of whether media type matters. If you have a large number of companies, you could separate them into subsets in which the companies use the same types of media (or perhaps the same categories). Within a subset, a regression model could get at the contributions of the various types. A preliminary analysis could use as X2 the total AD spending for all types that the company used. If X2 did not make an important contribution (perhaps unlikely), you would not need to pursue the individual types of media. David Hoaglin On Sat, May 26, 2012 at 5:43 PM, Mike Kim <kalisperos@gmail.com> wrote: > Hi David, > > Thank you for your opinion. The data structure is more complicated in fact. > Say, there are 50 different media types and each company (i) has different > number of media spending (from 1 to 50). So, setting all these as > independent variables is not possible. > > Anyway, the regression form I specified below does not seem correct. > Probably the only way is to aggregate information about (j) and make all > variables specific to only (i). > > Thank you, > Mike. * * 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/