我分别研究全要素生产率和劳动生产率对我国OFDI的影响。16个行业作为截面数据，9年，分别以OFDI为因变量+ATFP为自变量，OFDI为因变量+LTFP为自变量，做面板数据。在选择模型时，我先假设是变截距模型，做了固定效应的冗余变量似然比检验，显示混合回归模型无效，那么就是应该变截距固定效应模型咯。但是我后来用各个模型的残差平方和计算出的F2就小于5%的临界值，那就说明要用混合回归模型啦！这前后不是矛盾吗？！请问各位大神，这种情况我该怎么处理？！谢谢！下图是固定效应的冗余变量似然比检验的结果：
Redundant Fixed Effects Tests | [size=9.0000pt] | [size=9.0000pt] |
Pool: POOL2 | [size=9.0000pt] | [size=9.0000pt] | [size=9.0000pt] |
Test cross-section fixed effects | [size=9.0000pt] |
| | | | |
| | | | |
Effects Test | Statistic | d.f. | Prob. |
| | | | |
| | | | |
Cross-section F | 8.821333 | (15,112) | 0.0000 |
Cross-section Chi-square | 112.317109 | 15 | 0.0000 |
| | | | |
| | | | |
| | | | |
Cross-section fixed effects test equation: | [size=9.0000pt] |
Dependent Variable: OFDI? | [size=9.0000pt] | [size=9.0000pt] |
Method: Panel Least Squares | [size=9.0000pt] | [size=9.0000pt] |
Date: 05/19/15 Time: 17:11 | [size=9.0000pt] | [size=9.0000pt] |
Sample: 2004 2012 | [size=9.0000pt] | [size=9.0000pt] |
Included observations: 9 | [size=9.0000pt] | [size=9.0000pt] |
Cross-sections included: 16 | [size=9.0000pt] | [size=9.0000pt] |
Total pool (balanced) observations: 144 | [size=9.0000pt] |
| | | | |
| | | | |
Variable | Coefficient | Std. Error | t-Statistic | Prob. |
| | | | |
| | | | |
C | -1120319. | 215719.7 | -5.193400 | 0.0000 |
_TRANS--LTFP_TRANS | 435182.9 | 73706.88 | 5.904238 | 0.0000 |
_INF--LTFP_INF | 314452.6 | 62168.33 | 5.058083 | 0.0000 |
_SELL--LTFP_SELL | 446808.9 | 61493.40 | 7.265964 | 0.0000 |
_HOTEL--LTFP_HOTEL | 337543.6 | 68330.14 | 4.939894 | 0.0000 |
_REALTY--LTFP_REALTY | 278030.9 | 52787.05 | 5.267030 | 0.0000 |
_RENT--LTFP_RENT | 864099.5 | 78090.89 | 11.06531 | 0.0000 |
_SCI--LTFP_SCI | 449318.5 | 87588.09 | 5.129904 | 0.0000 |
_IRRI--LTFP_IRRI | 632203.4 | 128130.1 | 4.934075 | 0.0000 |
_RES--LTFP_RES | 266984.4 | 53257.23 | 5.013111 | 0.0000 |
_HYG--LTFP_HYG | 564612.6 | 114894.4 | 4.914189 | 0.0000 |
_CUL--LTFP_CUL | 417198.5 | 84524.56 | 4.935826 | 0.0000 |
_MIN--LTFP_MIN | 551025.8 | 69678.46 | 7.908122 | 0.0000 |
_MAN--LTFP_MAN | 441834.5 | 71048.44 | 6.218778 | 0.0000 |
_POWER--LTFP_POWER | 371423.9 | 71420.88 | 5.200494 | 0.0000 |
_CON--LTFP_CON | 449559.7 | 85547.45 | 5.255092 | 0.0000 |
_AGR--LTFP_AGR | 271956.6 | 53387.88 | 5.093976 | 0.0000 |
| | | | |
| | | | |
R-squared | 0.769268 | Mean dependent var | 210779.9 |
Adjusted R-squared | 0.740199 | S.D. dependent var | 433227.0 |
S.E. of regression | 220818.6 | Akaike info criterion | 27.55856 |
Sum squared resid | 6.19E+12 | Schwarz criterion | 27.90916 |
Log likelihood | -1967.216 | Hannan-Quinn criter. | 27.70102 |
F-statistic | 26.46385 | Durbin-Watson stat | 0.983839 |
Prob(F-statistic) | 0.000000 | | | |
| | | | |
| | | | |
扫码加好友，拉您进群
全部版块
我的主页
收藏
