Using country dummy variables? PreviousNext
Mplus Discussion > Categorical Data Modeling >
Message/Author
 Job Nauta posted on Saturday, January 04, 2014 - 7:23 am
Hi,
Short introduction I`m currently researching whether institutional factors have an effect on Earnings management. Not sure earnings management is a common term on this forum, but it doesn`t matter, I just want to discuss some basic statistics.

Here it comes;
My database consists of 9 countries and 6 years, I have 4 institutional factors en several control variables and using linear regression. My first thought was to control for country as well as years by using country/year-dummies. But now the catch, when I tried using country dummy variables I got very high VIF values and multicollinearity issues (on a side note, the year dummies don`t give me any problems).

My question is whether I should even be using country dummy variables, since my whole research is actually finding out whether there is a country specific effect? And it is quite logical it correlates with the institutional (which are country specific) factors?

ps. I have considered the dummy trap by using n-1 dummys (so 8 country dummies en 5 year dummies)

ps1. sorry to bust in like this and the forum(being my first post a question),but will try to contribute in the future ;) :-)
 Linda K. Muthen posted on Saturday, January 04, 2014 - 1:11 pm
Please send the outputs with and without the dummy variables along with your license number to support@statmodel.com.
Back to top
Add Your Message Here
Post:
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Password:
Options: Enable HTML code in message
Automatically activate URLs in message
Action: