我用SFA做成本效率分析,用Frontier4.1运算出来的效率值都大于1,第一次做这个,很多不懂,求教各位大神,哪里有问题?
下面是运算结果。
the final mle estimates are :
coefficient standard-error t-ratio
beta 0 0.12573106E+01 0.11422503E+01 0.11007312E+01
beta 1 0.35018363E-01 0.51865415E-01 0.67517753E+00
beta 2 -0.12668889E-01 0.71139278E-02 -0.17808571E+01
beta 3 0.14828278E+00 0.85853089E-01 0.17271689E+01
beta 4 0.22122124E+00 0.76374937E-01 0.28965161E+01
beta 5 0.93003455E-01 0.47324582E-01 0.19652251E+01
beta 6 -0.78653266E-02 0.25099403E-01 -0.31336708E+00
beta 7 0.11383865E+00 0.58328392E-01 0.19516850E+01
beta 8 0.11396525E-01 0.32928802E-02 0.34609597E+01
beta 9 -0.47843597E-02 0.28213281E-02 -0.16957828E+01
beta10 0.23252984E-01 0.14358372E-01 0.16194721E+01
beta11 -0.72793886E-01 0.88892026E-01 -0.81890232E+00
beta12 0.92904229E-02 0.70102484E-01 0.13252630E+00
beta13 -0.54950883E-01 0.66922934E-01 -0.82110690E+00
beta14 -0.16689328E-01 0.15201074E-01 -0.10979046E+01
beta15 0.36759334E+00 0.74940009E-01 0.49051681E+01
beta16 0.19698219E+00 0.19576060E-01 0.10062402E+02
sigma-squared 0.13036583E+00 0.16837994E-01 0.77423609E+01
gamma 0.17286246E+00 0.87981553E-01 0.19647581E+01
mu is restricted to be zero
eta 0.19787723E+00 0.82589703E-01 0.23959067E+01
log likelihood function = -0.15841153E+03
LR test of the one-sided error = 0.14660113E+02
with number of restrictions = 2
[note that this statistic has a mixed chi-square distribution]
number of iterations = 30 迭代次数
(maximum number of iterations set at : 100) (迭代最大次数)
number of cross-sections = 68
number of time periods = 6
total number of observations = 408
thus there are: 0 obsns not in the panel