From the results of SDs of TE, PTE and SE, in 2007–2009, we find the
between-university variations of TE, PTE, or SE scores are not enlarged obviously,
The SD of TE scores is in the downtrend, and that of PTE scores is in the uptrend,
while that of SE scores is relatively stable. The SDs of three efficiency scores are
around 0.23, 0.27, and 0.19 respectively, a little smaller than the results in Table4.9.
The means of SDs are relatively large, reflecting that the between-university dif-
ferences of TE, PTE and SE are considerably large among 211Us.
By using original data and MA data to run the DEA models and do the analysis,
wefind that the results of the original data are much morefluctuant than those of the
MA data, which demonstrates that MA method with one-period lag can smooth the
noises of the yearly data to some extent. Meanwhile, even there are some differ-
ences between the results of the two methods, the distributions of efficiency scores
and their overall patterns are basically consistent. Therefore, the researchfindings in
this section are robust.
by University Location............................ 4.2.6 Status and Change of Research Production Efficiency
in 985Us and Non-985Us
Through the descriptive analysis on the research inputs and outputs of Chap. 3 ,we
can observe that though the sampled universities are all titled“211 Project”, there
are substantial differences between 985Us and non-985Us in research inputs and
outputs: 985Us outperform non-985Us with great advantage in every input and
output indicator. However, this advantage in absolute amount does not mean that
the 985Us are much more technically efficient than non-985Us. In fact, a research
finding that we have in Chap. 3 is that on some per unit indicator, the performance
of 985Us is not greater than non-985Us. For example, in HSS disciplines, the
averages of per capita monographs and per¥10 million monographs of 985Us are
lower than those of non-985Us. In order to clearly reveal the differences between
the two types of 211Us, in this section, we divide the full sample 211Us into 985Us
and non-985Us, and compare their efficiency differences.
4.2.4 NEAM Disciplines...............................
Based on the full sample results of NEAM disciplines in previous sections, we
aggregate the data by university type, and compare the differences of NEAM
research production efficiency between the two types.
Table4.12presents the numbers and proportions of DEA-efficient units in 985Us
and non-985Us. It can be seen from Table4.12, the number of DEA-efficient units of
985Usfirst increases and then decreases. More specifically, the DEA-efficient
number of 985Us reaches its minimum in 2008, with a proportion of merely 24%.
4.2 Overall Status and Change of 211Us’Research Production... 129