Skip to main content

Table 2 Parameter estimates (standard errors) of provision networks

From: Exchanging and using research evidence in health policy networks: a statistical network analysis

  Child health Malaria HIV
  Structural effects model Attribute effects model Full model Structural effects model Attribute effects model Full model Structural effects model Attribute effects model Full model
Structural effects    
Edge intercept -4.52 (0.96)* -1.18 (0.70)* -4.65 (1.83)* -2.13 (0.32)* -2.20 (0.44)* -2.56 (0.52)* -1.03 (0.37)* -1.99 (1.03)* -1.79 (1.14)
Triangle closure 2.81 (0.63)* __ 2.36 (0.78)* 0.74 (0.47) __ 0.78 (0.85) 0.382 (0.79) __ 0.405 (1.45)
Reciprocity -5.10 (1.92)* __ -3.52 (2.20) -1.20 (1.15) __ -1.71 (1.29) 0.007 (0.93) __ 0.318 (1.40)
Layering 5.15 (1.44)* __ 6.16 (1.73)* 3.23 (0.59)* __ 2.87 (0.63)* 2.29 (0.83)* __ 2.31 (0.98)*
Attribute effects    
Researcher __ -0.23 (0.42) -0.93 (0.98) __ 1.43 (0.37)* 1.021 (0.44)* __ 0.505 (0.56) 0.154 (0.60)
Civil society org. __ No obs. No obs. __ -0.46 (0.32) -0.671 (0.40)* __ 0.150 (0.46) -0.120 (0.49)
Development partner org. __ 1.02 (0.45)* 0.31 (1.14) __ -1.10 (0.67) -0.813 (0.79) __ 3.37 (1.15)* 3.50 (1.16)*
Other __ 2.43 (1.19)* 5.04 (5.29) __ 0.81 (0.74) 0.217 (0.94) __ 2.10 (0.89)* 1.61 (1.04)
Manager/director __ 0.33 (0.35) 0.16 (0.80) __ nf nf __ nf nf
Homophily    
Researcher __ nf nf __ nf nf __ nf nf
Organization   nf nf __ nf nf __ nf nf
AIC 233.7 179.5 176.4 345.3 369.1 342.8 210.1 191.6 193.1
  1. *p < 0.05; nf = did not improve model fit.