Do graduate students affect their supervisors’ research?

1. ATH panel at the 8th International Conference on Webometrics, Informatics and Scientometrics & 13th COLLNET Meeting
25th October 2012
Duckhee Jang(SKAI), Dong-Seong Han(NRF) and Ki-Seok Kwon (KIU)

2. Contents
1.Introduction
2.Size, Graduates and Productivity
3.Methodology and data
4.Results
5.Conclusion

3. 1. Introduction
A number of studies have been conducted until
recently about what factors bring the differences on
the research productivity (Hess, 1977; Stephan, 1996)
According to a variety of previous empirical
researches, the results show that different
environmental, organizational, and human
properties influence on the research productivity
of researchers
An effective procurement of more talented graduate
students makes more researches conducted in depth.
Moreover, research experience is critical for the career
management of graduate students

4. 2. Literature review: size and
productivity
The procurement of excellent human resources is
the deterministic factor having higher research
productivity (Stephan & Levin, 1992)
In terms of the correlation between the size and
research productivity, positive correlation was
noticed in the small scale at the personal level, and
the highest makes were shown at a certain level at
the departmental level, and the positive correlation
was witnessed in the very small-scale institution at
the institutional level (Von Tunzelmann et al., 2003)

5. 2. Literature review: graduates and
productivity
Graduate students play the most important roles
in the research results of universities among
human resources of research (Salter et
al., 2000)
The increase in the number of doctoral students
has significantly positive correlation with the
professors’ and universities’ productivity
(Song, 2001; Kwon, 2010; Ryu & Pae, 2007)

6. 2. Literature review
Recent studies on the relationship between the size
of research groups and the productivity show very
complicated aspects.
The most typical empirical research outcome is that
the research productivity in small-size groups
increase as the size increases, yet it shows the
opposite result in large-size groups.
Also, the optimal research group size is said to be
5-9 researchers as verified in various research
fields (Johnston, 1994)

7. 3. Methodology and data
Regression Model: OLS
Category Index Definition and Measurement
Pub_1 KCI-Level Research Outcomes
Dependent Variable
Pub_2 SCI-Level Research Outcomes
Independent Variable Number of Graduate Students Number of Graduate Students
Gender 1= Dummy of Females
Natural Logarithm of Government's scale
Grant_gov
of grants to research cost
Natural Logarithm of the scale of private g
Grant_pri
rants to research cost
Natural Logarithm of the scale of inner gra
Grant_inner
nts to research cost
Control Variable Age Age
Age2 Age Squared
Nation 1=Dummy of Domestic Doctorates
NU 1=Dummy of National/Public Universities
Local Regional Dummy Variable
Dummy of Medium-Scale Classification o
Cur
f Fields of Study

8. 3. Methodology and data
Data is from ‘A study on academic research and
development activities in Korea’ in 2009, which NRF
carries out every year aimed at four-year universities.
Descriptive statistics: academics in four fields
Discipline Frequency %
Humanities 8,955 23.47
Social Sciences 12,807 33.57
Natural Sciences 5,513 14.45
Engineering 10,874 28.5
Total 38,149 100

9. 4. Results
Descriptive statistics: size of graduate students
Variation
Discipline Frequency Average S.D.
Coefficient
Humanities 8,955 1.057 0.050 4.730
Social
12,807 1.920 0.054 2.813
Sciences
Natural
5,513 5.415 0.147 2.715
Sciences
Engineering 10,874 7.346 0.122 1.661

10. T-test results by gender and region
Gender Region
Discipline Group Obs. Mean Group Obs. Mean
Male 6,780 1.186(0.063) Non-Metropolitan Area 4,722 0.881(0.062)
Female 2,175 0.654(0.069) Metropolitan Area 4,233 1.253(0.082)
Humanities combined 8,955 1.057(0.050) combined 8,955 1.057(0.050)
diff 0.532(0.118) diff -0.372(0.101)
t-value 4.5275** t-value -3.6778**
Male 10,683 1.937(0.059) Non-Metropolitan Area 7,108 1.665(0.067)
Female 2,124 1.834(0.131) Metropolitan Area 5,699 2.238(0.087)
Social
combined 12,807 1.92(0.054) combined 12,807 1.92(0.054)
Sciences
diff 0.102(0.145) diff -0.573(0.108)
t-value 0.7069 t-value -5.2854**
Male 4,385 5.752(0.170) Non-Metropolitan Area 3,406 5.257(0.182)
Female 1,128 4.105(0.278) Metropolitan Area 2,107 5.671(0.248)
Natural
combined 5,513 5.415(0.147) combined 5,513 5.415(0.147)
Sciences
diff 1.648(0.363) diff -0.414(0.302)
t-value 4.5351** t-value -1.3709
Male 10,466 7.419(0.125) Non-Metropolitan Area 7,083 7.328(0.148)
Female 408 5.478(0.526) Metropolitan Area 3,791 7.38(0.214)
Engineering combined 10,874 7.346(0.122) combined 10,874 7.346(0.122)
diff 1.941(0.640) diff -0.053(0.255)
t-value 3.0336** t-value -0.2070

11. Regression results: the estimation of KCI papers by OLS
Variable Total Humanities Social Sciences Natural Sciences Engineering
0.002** 0.002 0.004** 0.002** 0.003**
# of Graduate Students
(0.0001) (0.001) (0.001) (0.001) (0.0001)
0.008** 0.018** 0.013** 0.001 0.007**
lncen
(0.001) (0.002) (0.001) (0.001) (0.001)
0.007** 0.013** 0.009** 0.015** 0.005**
lnpri
(0.001) (0.004) (0.002) (0.002) (0.001)
0.026** 0.036** 0.027** 0.016** 0.019**
lninner
(0.001) (0.001) (0.001) (0.001) (0.001)
-0.043** -0.036** -0.039** -0.079** -0.015
Gender(1=Male)
(0.007) (0.011) (0.012) (0.019) (0.025)
0.081** 0.088** 0.079** 0.066** 0.066**
age
(0.003) (0.005) (0.005) (0.008) (0.006)
-0.001** -0.001** -0.001** -0.001** -0.001**
age2
0.000 0.000 0.000 (0.0001) (0.0001)
0.025** 0.008** 0.011 0.063** 0.030**
nation(1=Domestic Doctorates)
(0.005) (0.011) (0.009) (0.013) (0.010)
NU(1=National/Public 0.075** 0.075** 0.071** 0.059** 0.070**
Universities) (0.006) (0.012) (0.010) (0.014) (0.010)
LocalDum(1=Metropolitan 0.001 0.005 0.030** -0.045** -0.017**
Area) (0.005) (0.010) (0.009) (0.014) (0.011)
Sub-Discipline Included
-1.500** -1.744** -1.315** -1.263** -1.213**
Intercept
(0.072) (0.119) (0.133) (0.205) (0.153)
The value of ( ) is Std. Err.
**: p<0.01, *: p<0.05

12. Regression results: the estimation of SCI papers by OLS
Variable Total Humanities Social Sciences Natural Sciences Engineering
0.003** 0.0001* 0.001* 0.005** 0.002**
Number of Graduate Students
(0.0001) (0.0001) (0.0001) (0.001) (0.0001)
0.012** 0.002** 0.004** 0.020 0.015**
lncen
(0.0001) (0.0001) (0.001) (0.001) (0.001)
0.007** 0.001 0.003** 0.004** 0.009**
lnpri
(0.001) (0.001) (0.001) (0.002) (0.001)
0.005** 0.001* 0.002** 0.013** 0.009**
lninner
(0.0001) (0.0001) (0.0001) (0.001) (0.001)
0.025** -0.002 0.017** 0.080** 0.080**
Sex(1=Male)
(0.005) (0.002) (0.005) (0.018) (0.020)
0.009** 0.000 0.004 0.002** 0.028**
age
(0.002) (0.001) (0.002) (0.008) (0.005)
-0.0001** -0.0001 -0.0001 -0.0001** -0.0001**
age2
(0.00001) (0.0001) (0.0001 ) (0.00001) (0.00001)
-0.046** -0.005* -0.026** -0.087** -0.090**
nation(1=Domestic Doctorates)
(0.003) (0.002) (0.004) (0.012) (0.008)
0.027** 0.006** 0.007 0.061** 0.051**
NU(1=National/Public Universities)
(0.004) (0.002) (0.004) (0.013) (0.009)
0.040** 0.003 0.033** 0.063** 0.072**
LocalDum(1=Metropolitan Area)
(0.003) (0.002) (0.004) (0.013) (0.009)
0.382** 0.005 0.172** 0.458** 0.921**
Sub-Discipline
(0.047) (0.023) (0.058) (0.194) (0.126)
The value of ( ) is Std. Err.
**: p<0.01, *: p<0.05

13. 5. Summary and discussion
Male researchers managed more graduate students
than females, and this may be due to the fact that
the impact of research funding factors which had
focused on male researchers and the social
tendency that males’ career is more recognized
Capital-area researchers had higher research
supporting factors compared to non-metro area
researchers. Also, in terms of managing graduate
students, metro-area researchers managed more
graduate students

14. 5. Summary and discussion
Most importantly, the size of graduate students was
proved to be the predictor which causes the significant
differences in the researcher’s performance (both in
terms of domestic and int’l publications)
Research policies are needed to be implemented by
managing the research projects mostly aiming at
researchers of underrepresented groups, and these
groups are females, junior professors, and regional
researchers