The relationship between social capital and self-rated health: a multilevel analysis based on a poverty alleviation program in the Philippines

Table 4 Multilevel mixed-effects linear regression testing the impact of social capital on self-rated health, post-Transform (Endline data)

Variables	Unstandardised estimate (S.E.)
Variables	Model 1	Model 2	Model 3	Model 4
Intercept	79.71 (0.74)***	31.81 (10.02)***	31.78 (5.53)***	34.31 (10.12)**
Fixed effects
Bonding SC		−4.44 (3.31)	−3.25 (2.24)	−10.15 (4.20)*
Bridging SC		6.20 (3.16)*	5.38 (2.19)*	−7.80 (12.46)
Cognitive SC		4.22 (2.66)	5.42 (2.77)*	10.14 (4.12)*
Bonding SC x Bridging SC				27.26 (13.82)*
Bridging SC x Cognitive SC				−13.94 (6.99)*
Random effects
Level 1 variance	375.87 (37.90)	361.71 (36.09)	326.28 (10.76)	323.96 (31.17)
Level 2
Variance of random intercepts (RI)	14.82 (5.89)	15.65 (6.21)	196.60 (53.09)	195.74 (72.06)
Variance of random slopes (RS)			256.37 (71.25)	256.41 (82.10)
Covariance between the RI and RS			−216.55 (59.85)	− 215.56 (75.40)
ICC^a	0.038	0.041
Model information criteria
AIC	19,041.18	16,917.78	16,796.48	16,788.62

* p < 0.05; ** p < 0.01; *** p < 0.001
^a A single estimate for ICC is not available for Models 3 and 4 as it is a function of the variable for which random slopes are specified (i.e. conditional on the values of cognitive social capital)
The following covariates were also controlled for in the models: age, sex, marital status, number of people in household, highest educational attainment, whether the respondent is in work, religion, food security and hygiene. The estimates for these variables can be found in Additional file 1: Table S7A

ISSN: 1471-2458