| 所在主题： | |
| 文件名: test.xlsx | |
| 资料下载链接地址: https://bbs.pinggu.org/a-1395774.html | |
| 附件大小: | |
|
本文目前做一组面板数据，在混合回归、固定效应、随机效应选择时，选择了固定效应，考虑到可能存在的时间效应，引入双向固定效应。 但是发现，直接采用时间虚拟变量的双向固定效应与引入时间趋势t的双向固定效应结果相差过大，后本人又使用lsdv方法做了一遍，还是相差过大，求解释。
我的理解中，这两个回归的结果应该可以直接对比才对吧？怎么会连符号都是反的？ 提供原始数据求赐教，不胜感激 xtreg growth lineq lineq2 sk lab gov urb free inf year2-year26,fe vce(clusterstate) Fixed-effects (within) regression Number of obs = 676 Group variable: state Number of groups = 26 R-sq:within= 0.6633 Obs per group: min = 26 between = 0.1240 avg = 26.0 overall = 0.5596 max = 26 F(25,25) = . corr(u_i, Xb)= -0.3070 Prob > F = . (Std. Err. adjusted for 26 clusters in state) ------------------------------------------------------------------------------ | Robust growth | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- lineq |-.8635252 2.093989 -0.41 0.684 -5.176176 3.449126 lineq2 | .2544478 .3179352 0.80 0.431 -.400352 .9092476 sk | .1288633 .0299593 4.30 0.000 .0671611 .1905656 lab | .0107866 .0796543 0.14 0.893 -.1532645 .1748377 gov | .1416641 .0755192 1.88 0.072 -.0138707 .2971989 urb |-.0743788 .0183281 -4.06 0.000 -.1121262 -.0366315 free |-.0260178 .0212988 -1.22 0.233 -.0698835 .017848 inf |-1.429839 1.706492 -0.84 0.410 -4.944424 2.084747 year2 |-.5238741 1.311369 -0.40 0.693 -3.224689 2.176941 year3 | .8228026 1.239182 0.66 0.513 -1.729341 3.374946 year4 |-2.114758 1.239279 -1.71 0.100 -4.6671 .437584 year5 | 13.8368 1.769161 7.82 0.000 10.19314 17.48045 year6 |-3.979693 1.311379 -3.03 0.006 -6.680528 -1.278858 year7 | 3.010125 1.397518 2.15 0.041 .1318832 5.888367 year8 | 3.805373 1.621175 2.35 0.027 .4665005 7.144246 year9 | 6.094947 1.922175 3.17 0.004 2.136154 10.05374 year10 | 17.1785 2.521036 6.81 0.000 11.98633 22.37067 year11 | 11.81559 2.065792 5.72 0.000 7.56101 16.07017 year12 | 2.527148 1.681131 1.50 0.145 -.9352063 5.989503 year13 |-3.081213 1.424111 -2.16 0.040 -6.014225 -.1482022 year14 |-7.828048 1.301543 -6.01 0.000 -10.50863 -5.147469 year15 |-6.041721 1.406253 -4.30 0.000 -8.937954 -3.145488 year16 |-6.479232 1.518945 -4.27 0.000 -9.607558 -3.350905 year17 |-3.812392 1.529501 -2.49 0.020 -6.962459 -.6623263 year18 |-3.405987 1.8565 -1.83 0.078 -7.22952 .4175454 year19 | .9172535 2.07527 0.44 0.662 -3.356846 5.191353 year20 | 3.00897 2.054839 1.46 0.156 -1.22305 7.240991 year21 | .1064575 1.859886 0.06 0.955 -3.72405 3.936965 year22 |-1.020915 2.19289 -0.47 0.646 -5.537257 3.495428 year23 | 1.401365 2.306282 0.61 0.549 -3.348512 6.151242 year24 | .1244292 2.393551 0.05 0.959 -4.805181 5.054039 year25 |-2.373303 2.40414 -0.99 0.333 -7.324722 2.578116 year26 | .1464293 2.432795 0.06 0.952 -4.864006 5.156865 _cons | 11.17588 5.293276 2.11 0.045 .2741766 22.07759 -------------+---------------------------------------------------------------- sigma_u |2.5780625 sigma_e |4.3832578 rho |.25702135 (fraction of variance due to u_i) ------------------------------------------------------------------------------ xtreg growth lineq lineq2 sk lab gov urb free inf t,fe vce(cluster state) Fixed-effects (within) regression Number of obs = 676 Group variable: state Number of groups = 26 R-sq:within= 0.2144 Obs per group: min = 26 between = 0.0112 avg = 26.0 overall = 0.0611 max = 26 F(9,25) = 14.62 corr(u_i, Xb)= -0.8464 Prob > F = 0.0000 (Std. Err. adjusted for 26 clusters in state) ------------------------------------------------------------------------------ | Robust growth | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- lineq | 21.50038 4.686173 4.59 0.000 11.84903 31.15174 lineq2 | -1.901 .7040963 -2.70 0.012 -3.351113 -.450886 sk | .2079216 .0454613 4.57 0.000 .1142924 .3015509 lab | .2471607 .1181887 2.09 0.047 .0037465 .4905749 gov | .0352942 .1315356 0.27 0.791 -.2356084 .3061969 urb |-.0795639 .0781119 -1.02 0.318 -.2404385 .0813106 free | .0120288 .0194182 0.62 0.541 -.0279637 .0520213 inf |-.4286012 1.910528 -0.22 0.824 -4.363407 3.506205 t |-.9213756 .1123752 -8.20 0.000 -1.152817 -.6899345 _cons |-34.73503 7.191513 -4.83 0.000 -49.54623 -19.92383 -------------+---------------------------------------------------------------- sigma_u |5.6862185 sigma_e | 6.568536 rho |.42837332 (fraction of variance due to u_i) ------------------------------------------------------------------------------ 下文2楼给出lsdv结果 |
|
熟悉论坛请点击新手指南
|
|
| 下载说明 | |
|
1、论坛支持迅雷和网际快车等p2p多线程软件下载，请在上面选择下载通道单击右健下载即可。 2、论坛会定期自动批量更新下载地址,所以请不要浪费时间盗链论坛资源,盗链地址会很快失效。 3、本站为非盈利性质的学术交流网站,鼓励和保护原创作品，拒绝未经版权人许可的上传行为。本站如接到版权人发出的合格侵权通知，将积极的采取必要措施；同时，本站也将在技术手段和能力范围内，履行版权保护的注意义务。 (如有侵权，欢迎举报) |
|
京ICP备16021002号-2 京B2-20170662号
京公网安备 11010802022788号
论坛法律顾问：王进律师
知识产权保护声明
免责及隐私声明
