day6lm

# Interval agehd

This preview shows page 1. Sign up to view the full content.

This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: for one group and nega:ve for the other, when you combine the 1 and y 2 on likely Marginal e↵ects of xgroups, xou are Ey : to get a zero eﬀect. •  To deal with this, we need to interact (mul:ply) @ Ey @ Ey the appropriate independent variables. = 2 + 4 x2 and = 3 + 4 x1 @ x1 a + b1 X1 + b2 X2 @ x2b3 X1 X2 + e y= + The interpretation of these e↵ects and of the individual parameters is very Interac:on term speciﬁc to theoretical model behind the relationship. 10 2/17/12 Interpre:ng interac:on coeﬃcients Predicted income Extra effect of education if female y = a + bed Xed + b fem X fem + bint Xed X fem + e Effect of education for men Effect of being female Mean income when all Xs are zero Another way to think about it (assuming all Xs are dichotomous) Predicted income if Male (Xfem=0) Female (Xfem=1) Low Ed (Xed=0) a a+bfem High Ed (Xed=1) a+bed a+bed+bfem+bint . . . . ! Intercepts by gender +Slopes by gender Regression w/ interac:ons drop if marriedhd==1! global controls "agehd black hisp femalehd educhd”! gen femed=femalehd*educhd! reg lnfaminc \$controls femed! Source | SS df MS Number of obs = 4272! -------------+-----------------------------F( 6, 4265) = 115.36! Model | 570.735236 6 95.1225393 Prob > F = 0.0000! Residual | 3516.81104 4265 .824574686 R-squared = 0.1396! -------------+-----------------------------Adj R-squared = 0.1384! Total | 4087.54627 4271 .957046657 Root MSE = .90806! ! ------------------------------------------------------------------------------! lnfaminc | Coef. Std. Err. t P>|t| [95% Conf. Interval]! -------------+----------------------------------------------------------------! agehd | .0027621 .000813 3.40 0.001 .0011682 .0043561! black | -.2664906 .0294689 -9.04 0.000 -.324265 -.2087163! hisp | .0763496 .0571102 1.34 0.181 -.0356161 .1883153! femalehd | -.6155345 .151239 -4.07 0.000 -.9120416 -.3190275! educhd | .1029402 .0092153 11.17 0.000 .0848734 .1210071! femed | .0312132 .0116579 2.68 0.007 .0083577 .0540688! _cons | 9.026822 .1284679 70.27 0.000 8.774958 9.278686! ------------------------------------------------------------------------------! ! •  I restricted the sample to unmarried so we are comparing single men to single women. •  The interac:on is posi:ve. But the eﬀect of...
View Full Document

## This note was uploaded on 03/28/2012 for the course PADP 8130 taught by Professor Fertig during the Spring '12 term at LSU.

Ask a homework question - tutors are online