Hi,
I ran a multiple regression with three variables: yvar (continuous, dependent variable), xvar (continuous, independent variable), Hetero (dummy variable) and the interaction of xvar and Hetero (Heteroxvar). I created two graphs with the scatterplot of yvar and xvar given Hetero == 0 and Hetero == 1 and added the expected value lines in both cases. More specifically, I used the following code:
Does anyone know how can I add a confidence interval around the expected value lines? In addition, how can I write the value of the standard error of the coefficient on xvar on the graph? That is, I want to extract this standard error with e.g.
and then put a textbox on the graph or write on the borders of the graph something along the lines of "The standard error is equal to `SEvariable' ".
I ran a multiple regression with three variables: yvar (continuous, dependent variable), xvar (continuous, independent variable), Hetero (dummy variable) and the interaction of xvar and Hetero (Heteroxvar). I created two graphs with the scatterplot of yvar and xvar given Hetero == 0 and Hetero == 1 and added the expected value lines in both cases. More specifically, I used the following code:
Code:
reg yvar xvar Hetero Heteroxvar scalar coeff1 = _b[xvar] scalar coeff2 = _b[Heteroxvar] scalar coeff3 = _b[Hetero] scalar constant = _b[_cons] scalar slope = coeff1 + coeff2 scalar adder = constant + coeff3 graph twoway (scatter yvar xvar if Hetero == 1) (function y = adder + slope*x, range(xvar)), name(interactiontermregression) graph twoway (scatter yvar xvar if Hetero == 0) (function y = constant + coeff1*x, range(xvar)), name(noninteractiontermregression)
Code:
local SEvariable = _se[xvar]
