本人用实际GDP的对数来消除异方差,进而进行单位根检验,在进行ADF检验时,选取LEVEL,有截距和趋势进行,滞后项为3时,结果为平稳,而用无截距和趋势,有截距检验时,则是不平稳的。不知道这是怎么一回事,从GDP线型图看,它有明显的时间趋势(注意:我想要LNGDP原序列不平稳,从而进行一阶平稳成立)
我把数据发上来,请高手们看看,是我数据有错,还是怎么一回事。
年份 实际GDP LNGDP
1978 3645.20 8.201167
1979 3922.20 8.274408
1980 4228.20 8.349532
1981 4452.30 8.401176
1982 4853.00 8.487352
1983 5381.90 8.590797
1984 6200.00 8.732305
1985 7037.00 8.858937
1986 7663.30 8.944198
1987 8552.20 9.053944
1988 9518.60 9.161003
1989 9908.90 9.201189
1990 10285.40 9.238481
1991 11231.70 9.326495
1992 12826.60 9.459276
1993 14622.30 9.590303
1994 16537.80 9.713404
1995 18340.50 9.816867
1996 20174.50 9.912175
1997 22050.70 10.00110
1998 23770.70 10.07621
1999 25577.30 10.14946
2000 27725.80 10.23012
2001 30027.00 10.30985
2002 32759.40 10.39695
2003 36035.40 10.49226
2004 39675.00 10.58848
2005 43721.80 10.68560
2006 48793.50 10.79535
2007 54600.00 10.90779
2008 59514.00 10.99397
2009 64989.30 11.08198
2010 71683.20 11.18001
Null Hypothesis: LNGDP has a unit root
Exogenous: Constant, Linear Trend
Lag Length: 3 (Automatic based on AIC, MAXLAG=3)
t-Statistic Prob.*
Augmented Dickey-Fuller test statistic -4.427963 0.0076
Test critical values: 1% level -4.309824
5% level -3.574244
10% level -3.221728