3. ๏ก Growth of MultidisciplinaryCollaboration
(MDC).
๏ก Shift from traditional research organizations to
networked research.
๏ก Top-down arrangement of networked research
๏ก How does the top-down arrangement of MDC
shape real multidisciplinary interactions among
resarchers?
3
4. ๏ก Social capital theory on diversity
1. Diversity at the dyadic level (Lin 1999;
Erickson 2003)
2. Diversity at the network level (Reagans. 2004)
๏ก Two types of diversity:
Compositional Diversity: the degree of variation
among the attributes of network members, e.g.
gender, ethnicity, age at the network level.
Interactional Diversity: the extent of actual
boundary-spanning interactions at the individual
level.
4
5. ๏ก Rational choice theory (Becker 1974)
1. Preference (of compositional diversity)
2. Action (of interactional diversity)
3. Rewards (e.g. status, productivity)
๏ก A multilevel model of diversity including
compositional diversity at the project level,
interests in MDC and interactional diversity
at the individual level.
5
7. ๏ก H1:The greater compositional diversity a project exhibits, its
members show the greater interactional diversity.
๏ก H2:The more researchers want to involved in
multidisciplinary collaborations, the greater their
interactional diversity.
๏ก H3: Compositional diversity moderates the relationship
between rationale and interactional diversity. When
compositional diversity increases, those who have higher
rationale for multidisciplinary collaborations show greater
interactional diversity.
๏ก H4:The higher researchers status, the lower their
interactional diversity.
๏ก H5:When compositional diversity increases, those who have
higher status have lower interactional diversity. 7
9. 44% of NIs are working with 3 projects.
24% of NIs are working with more than 3 projects.
75% of CRs are working with 1 projects
3
10
18
7
3
45
12
2
1
0
1
2
3
4
5
NI
CR
Individual N=144, Project N =144
9
11. ๏ก Online survey on collaboratorsโ social networks
in GRAND (N=101)
1. Rationale (want-to-meet)
2. Interaction (comm., support, work)
3. Attributes (status, age, gender, location)
๏ก GRAND roster
Compositional diversity of projects
Heterogeneity index Blau (1977)
1-โ ๐๐๐๐
2
where p is the proportion of the number of
people in the ith disciplinary category.
11
13. Mean S.D.
Individual level variable
DependentVariable
Interactional Diversity 0.31 0.22
IndependentVariable
Interests in MDCs 0.46 0.22
Age 46.59 8.98
Role in GRAND 0.41 0.49
Academic ranking 2.03 0.82
GIndex 51.74 60.46
ControlVariable
Gender 0.27 0.45
% of researchers in same city 0.10 0.06
Project-LevelVariable
Compositional Diversity 0.51 0.27
Note: Unless otherwise noted, individual N=101, Project N=32 13
14. Model 1 Model 2 Model 3 Model 4
Intercept -0.004 -0.129 -0.359 -0.356
Individual level variable
Interests in MDCs 0.657*** 0.644*** 0.749***
Age 0.064 0.060 0.064
Role in GRAND .217** 0.224*** -0.026
Academic ranking 0.010 0.014 0.015
GIndex 0.058 0.074 0.058
Gender 0.269แถง 0.277แถง 0.227
% of researchers in same city 0.563 0.626 0.743
Project-Level Variable
Disciplinary Diversity 0.456แถง 0.496แถง
Cross-Level Interactions
Compositional DiversityXRationale -0.201
Compositional DiversityXStatus 0.517แถง
Variance Estimates
GroupVariance 0.342*** 0.048*** 0.392*** 0.026***
Individual Variance 0.732*** 0.288*** 0.284*** 0.280***
Individual N=101, Project N=32
แถงp < .1 (one-tailed test), *p < .05; ** p<.01; *** p<.001 (two-tailed tests) 14
15. ๏ก Compositional diversity fosters interactional
diversity (H1 supported).
๏ก Researchers who are more willing to be involved
in multidisciplinary collaborations show greater
interactional diversity. (H2 supported)
๏ก Researchers in a higher position in GRAND show
higher interactional diversity. (H4 refuted)
๏ก Compositional diversity moderates the effects of
researchersโ roles on interactional diversity. In
projects of high compositional diversity, those in
higher positions indicates greater interactional
diversity. (H5 refuted)
15