Hello,
I am a user of EViews 12 and I want to analyze a panel dataset by running an OLS with an interaction term.
y = c(1) + c(2)*x + c(3)*z + c(4)*xz + c(4)*control variables + e
For this equation, I would like to display a marginal effects plot with a 95 percent confidence interval. This plot should show the influence of x on y, given the second independent variable in the interaction term z:
dy/dx = c(2) + c(4)*z.
Is there an easy way to do this in EViews? I could not find a solution in manual and therefore would be very grateful for help.
Best,
Sebastian
Marginal effect plot
Moderators: EViews Gareth, EViews Moderator
-
- Non-normality and collinearity are NOT problems!
- Posts: 3775
- Joined: Wed Sep 17, 2008 2:25 pm
Re: Marginal effect plot
Code: Select all
series dydz = c(3) + c(4)*z
scat z dydz
Re: Marginal effect plot
Thank you! That works very well and helps a lot. What I do not understand, however, is how to get the cofidence intervals (95 % in my case) into the plot. How does that work?
-
- Non-normality and collinearity are NOT problems!
- Posts: 3775
- Joined: Wed Sep 17, 2008 2:25 pm
Re: Marginal effect plot
You'll need to retrieve the variance-covariance matrix of the coefficients from the regression. Then get the variance of the point estimate using the standard formula for linear combinations of random variables. Take the square root, and then take the point estimates plus and minus twice that square root.
Who is online
Users browsing this forum: No registered users and 72 guests