I am running a linear regression with wage as dependent variable; education (dummy coded with 1=master; 0=bachelor) and gender (1=female; 0=male) are my categorical predictors. I am interested to see if the effect that gender has on wage is moderated by education education (i.e. comparing the gender wage gap for master and bachelor). Therefore, I have interacted education and gender and obtained the following results:
Code:
reg wage female##edulvl if cntry=="GB" Source | SS df MS Number of obs = 2,117 -------------+---------------------------------- F(3, 2113) = 76.15 Model | 1171.64829 3 390.549429 Prob > F = 0.0000 Residual | 10837.4098 2,113 5.12892088 R-squared = 0.0976 -------------+---------------------------------- Adj R-squared = 0.0963 Total | 12009.0581 2,116 5.67535827 Root MSE = 2.2647 ------------------------------------------------------------------------------- wage | Coef. Std. Err. t P>|t| [95% Conf. Interval] --------------+---------------------------------------------------------------- 1.female | -.426915 .1275451 -3.35 0.001 -.677042 -.176788 1.edulvl | 1.406627 .1492304 9.43 0.000 1.113973 1.699281 | female#edulvl | 1 1 | .1045125 .2019969 0.52 0.605 -.291621 .500646 | _cons | 4.507881 .0947753 47.56 0.000 4.322018 4.693743 -------------------------------------------------------------------------------
0 Response to Moderation/Interaction effect
Post a Comment