Skip to main content

Table 7 Comparison of two-level random-intercept linear regression models for duration of epidemics by network size

From: A model for the co-evolution of dynamic social networks and infectious disease dynamics

 

Combined

N=10

N=15

N=20

N=25

N=50

Fixed effects (observed effects)

 Intercept

21.23*** (0.19)

18.41*** (0.29)

20.43*** (0.23)

21.48*** (0.18)

22.00*** (0.16)

20.97*** (0.08)

Main effects

 I. CIDM parameters

  Benefit distance 2 (\(\beta\))

0.36*** (0.04)

0.59*** (0.10)

0.71*** (0.08)

0.45*** (0.06)

0.19*** (0.05)

− 0.16*** (0.03)

  Cost increase for infected ties (\(\mu\))

− 3.42*** (0.44)

− 10.68*** (1.15)

− 6.26*** (0.94)

− 2.83*** (0.73)

0.11 (0.65)

2.58*** (0.31)

  Disease severity (\(\frac{\sigma }{50}\))

− 1.81*** (0.26)

− 3.64*** (0.68)

− 3.08*** (0.56)

− 2.42*** (0.44)

− 1.38*** (0.39)

1.50*** (0.19)

  Risk perception (r)

− 2.81*** (0.27)

− 5.06*** (0.70)

− 4.35*** (0.57)

− 3.62*** (0.45)

− 2.46*** (0.40)

1.45*** (0.19)

  Network size (\(\frac{N}{50}\))

− 0.11 (1.15)

     

 II. Network properties (start of epidemics)

  Density (\({\text{den}}_{\text{start}}\))

− 19.93*** (3.20)

− 22.55*** (2.81)

− 33.53*** (4.87)

− 25.41*** (6.60)

− 18.31* (8.19)

− 41.03*** (11.45)

  Degree of patient-0 (\({\text{deg}}^{0}_{\text{start}}\))

3.31*** (0.62)

2.24* (0.98)

7.35*** (1.38)

1.71 (1.66)

2.57 (1.88)

− 1.67 (2.11)

Interaction effects

 \(\beta\) x \(\mu\)

1.05*** (0.15)

1.71*** (0.38)

2.01*** (0.31)

1.45*** (0.24)

0.56** (0.22)

− 0.47*** (0.10)

 \(\beta\) x N

− 1.06*** (0.13)

     

 \(\mu\) x \(den_{start}\)

− 38.17*** (5.00)

2.64 (10.54)

− 1.11 (18.65)

30.21 (25.49)

31.33 (31.70)

− 27.52 (45.03)

 \(\frac{\sigma }{50}\) x r

− 4.16*** (0.64)

− 6.74*** (1.67)

− 6.84*** (1.37)

− 6.38*** (1.07)

− 3.98*** (0.95)

3.12*** (0.45)

 \(\frac{\sigma }{50}\) x N

6.48*** (0.93)

     

 r x N

8.20*** (0.96)

     

 \(\frac{N}{50}\) x \({\text{den}}_{\text{start}}\)

32.97*** (9.35)

     

Random effects (unobserved effects)

 \(s^{2}\)

3.96

5.59

3.57

2.05

1.60

0.30

 Log likelihood (\(\ell\))

− 113,274.74

− 23,109.20

− 23,355.84

− 23,203.31

− 22,761.55

− 19,558.51

 Intraclass correlation (\(\rho\))

0.114

0.138

0.087

0.054

0.048

0.022

 Observations

36,000

7200

7200

7200

7200

7200

 Groups: parameter combinations

360

72

72

72

72

72

  1. *** p < 0.001, **p < 0.01, *p < 0.05, SEs in parentheses