各位，请教了！

fy	Y	X2		prais y x2 zy,twostep
1983	11.87764	6.231122974
1984	12.04982	7.519523192		Iteration 0: rho = 0.0000
1985	12.15839	8.414325566		Iteration 1: rho = 0.4594
1986	12.27428	8.030865039
1987	12.59321	8.445765931		Prais-Winsten AR(1) regression -- twostep estimates
1988	12.80043	9.852932091
1989	12.91313	10.25292001		Source SS df MS	Number of obs	24
1990	12.95381	10.17341858			F( 2, 21)	1574.56
1991	13.03224	10.73479539		Model 13.9943638 2 6.99718191	Prob > F	0
1992	13.17632	11.34905951		Residual .093321662 21 .004443889	R-squared	0.9934
1993	13.43864	11.78540989			Adj R-squared	0.9927
1994	13.65296	12.05581204		Total 14.0876855 23 .612508065	Root MSE	0.06666
1995	13.74767	11.35268405
1996	13.86559	11.36750193
1997	13.91665	11.53476836		y Coef. Std. Err. t P>t	[95% Conf.	Interval]
1998	13.93874	11.55374932
1999	13.99078	11.74970885		x2 .0049811 .0306621 0.16 0.873	-0.0587842	0.0687464
2000	14.05503	11.9168582		zy .9709511 .0578415 16.79 0.000	0.8506631	1.091239
2001	14.15252	11.95823596		_cons .4683023 .5068272 0.92 0.366	-0.5857025	1.522307
2002	14.23349	12.10552831
2003	14.33085	12.35885426		rho .4593725
2004	14.44974	12.62398537
2005	14.56763	12.92092696		Durbin-Watson statistic (original) 1.066868
2006	14.71255	12.12415178		Durbin-Watson statistic (transformed) 1.544177
2007	14.93836	12.10489736

上面的分析中，为什么t值不合，能否有其他办法？求救了！