I have the following regression of panel data using fixed effects, I was wondering after a lot of search if can I add a time trend to the regression and if so how to test if is better that the present regression
Code:
Fixed-effects (within) regression Number of obs = 348 Group variable: j_SA Number of groups = 29 R-sq: Obs per group: within = 0.7610 min = 12 between = 0.0029 avg = 12.0 overall = 0.0003 max = 12 F(5,314) = 199.99 corr(u_i, Xb) = -0.7526 Prob > F = 0.0000 ------------------------------------------------------------------------------ l_arrival | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- l_incomepc | .5538903 .2660921 2.08 0.038 .0303413 1.077439 l_price | .1738661 .0723778 2.40 0.017 .0314594 .3162728 l_violence | .0009841 .0217045 0.05 0.964 -.0417205 .0436887 l_trade | .4550106 .1074208 4.24 0.000 .243655 .6663661 l_transport | .2258229 .0742288 3.04 0.003 .0797742 .3718716 _cons | -3.003777 2.029769 -1.48 0.140 -6.997444 .9898895 -------------+---------------------------------------------------------------- sigma_u | 1.9418083 sigma_e | .20554152 rho | .9889198 (fraction of variance due to u_i) ------------------------------------------------------------------------------ F test that all u_i=0: F(28, 314) = 200.43 Prob > F = 0.0000
0 Response to Time trends in panel data
Post a Comment