本人用eviews做logit模型，结果如下，除系数外，标准差、统计量、P值都为NA，不知是什么原因，请大侠指点！！！

Dependent Variable: DEPENDENT    
Method: ML - Binary Logit (Quadratic hill climbing)    
Date: 03/29/09   Time: 22:15    
Sample: 1 131    
Included observations: 131    
Convergence achieved after 39 iterations    
WARNING: Singular covariance - coefficients are not unique    
Covariance matrix computed using second derivatives    
    
Variable Coefficient Std. Error z-Statistic Prob. 
    
C 1637.127 NA NA NA
EXPERIENCE1 -24.96572 NA NA NA
EXPERIENCE2 0.113727 NA NA NA
HCI 0.630150 NA NA NA
IT 3.449813 NA NA NA
GDP -0.068768 NA NA NA
PGDP -0.051291 NA NA NA
POP -0.104628 NA NA NA
RATIOFDI 42.75348 NA NA NA
YEAR -0.805496 NA NA NA
RESOURCE 1.267700 NA NA NA
    
Mean dependent var 0.824427     S.D. dependent var  0.381917
S.E. of regression 0.336748     Akaike info criterion  0.878288
Sum squared resid 13.60787     Schwarz criterion  1.119717
Log likelihood -46.52786     Hannan-Quinn criter.  0.976391
Restr. log likelihood -60.86431     Avg. log likelihood  -0.355174
LR statistic (10 df) 28.67291     McFadden R-squared  0.235548
Probability(LR stat) 0.001407   
    
Obs with Dep=0 23      Total obs  131
Obs with Dep=1 108