求教eviews5.0中面板数据协整检验

以下是引用minixi在2007-5-17 18:34:00的发言：

EViews 6.0 beta用engel granger思想，通过残差检验面板序列之间是否协整？

平稳=>被检验的验面板序列之间协整。

这样理解是否正确？怎样看结果？

Pedroni Residual Cointegration Test    
Series: 已经删去      
Date: 05/17/07   Time: 17:09    
Sample: 1986 2005    
Included observations: 20    
Cross-sections included: 29    
Null Hypothesis: No cointegration    
Trend assumption: No deterministic trend    
Lag selection: Automatic SIC with a max lag of 2    
Newey-West bandwidth selection with Bartlett kernel    
    
Alternative hypothesis: common AR coefs. (within-dimension)    
                                       Weighted
                     Statistic  Prob.  Statistic  Prob.
Panel v-Statistic    -0.241774  0.3875 -1.156392  0.2044
Panel rho-Statistic   3.870924  0.0002  4.087770  0.0001
Panel PP-Statistic    0.356500  0.3744 -0.647370  0.3235
Panel ADF-Statistic  -3.235889  0.0021 -3.991631  0.0001
    
Alternative hypothesis: individual AR coefs. (between-dimension)    
    
                      Statistic Prob.  
Group rho-Statistic   5.993404  0.0000  
Group PP-Statistic   -4.450792  0.0000  
Group ADF-Statistic  -5.287760  0.0000  

您这里Pedroni 7统计量协整检验，组间结论比较一致，组内结论不大一致，建议再做Kao的LM检验和FISHER-ADF

还是不懂，单位根检验和协整检验有关系吗