Unexpected omitted variables

[Minchul Park]

Hi, I have a question about unexpected omitted variables.
When I run a regression with dummy variables, stata sometimes omits "two or more" variables.
Of course, stata may be trying to prevent the multicollinearity, but stata remove variables I do not want to omit occasionally (i.e. stata sometimes eliminates variables those have no problem I think)

So, my question is that is it fine to use the results omitting the unexpected variables? or I have to find other solution?
If using the result could be problematic, is there any solution for the problem?

Thank you for your time spending to read this question.

[Clyde Schechter]

For specific advice, you need to post example data and also show the actual code and output of your regression. Please use -dataex- to show the data example, and be sure to post the Results between code delimiters (see Forum FAQ #12 if you are not familiar with these) for readability. Also provide a full explanation of the context of your problem and a clear statement of your research question.

If Stata thinks your data are colinear, then they are. And if you think they aren't supposed to be then either the data are incorrect, or your understanding of your data is incorrect. So you need to consider which variables Stata has omitted and figure out whether or not colinearity among them is supposed to happen or not. If it's not supposed to happen, then you need to fix the data errors.

Another possibility is this: remember that in any regression, any observation that has a missing value for any variable mentioned in the regression command is omitted from the regression's estimation sample. It is possible that variables are not colinear in the full data set, but are colinear within the subset that remains in the estimation sample. In that case, you need to consider resolving the missing data, or accepting that within the limits of your sample you cannot include some variables you would otherwise have wanted to.

[Minchul Park]

Clyde Schechter Thank you for your good advice! your comment helped with the problem and is really informative for me. Also, I will keep in mind that I have to follow this community's convention such as how to upload my question. Thank you again.

[Carlo Lazzaro]

Minchul.
as an aside to Clyde's excellent advice, please note that collineraity is, non technically speaking, a matter of partnership between variables: that's why you find "unexpected" variables omitted du to perfect coliinearity.

[Minchul Park]

Carlo Lazzaro Thank you for your good reminder! Sometimes I forget about a basic (but very important) point. This kind of comment is really helpful! Thank you again.