Skip to main content

Table 5 Jimma Longitudinal Family Survey of Youth

From: Modeling overdispersed longitudinal binary data using a combined beta and normal random-effects model

Effect

 

Logistic

Beta-binomial

  

Mean(s.d.)

Mean(s.d.)

Intercept

ξ0

1.185(0.624)

1.151(0.731)

Age

ξ1

0.039(0.049)

0.047(0.057)

Place urban

ξ2

0.977(0.148)

1.134(0.183)

Place semi-urban

ξ3

0.987(0.161)

1.161(0.202)

Gender:Female

ξ4

−1.113(0.123)

−1.266(0.148)

Work

ξ5

0.133(0.122)

0.154(0.140)

Round

ξ6

0.343(0.142)

0.404(0.165)

Std. dev. random effect

d

Ratio

α / β

0.0111(0.0029)

DIC

 

2002.0

2001.0

pD

 

6.97

13.77

Effect

 

Logistic-normal

Combined

  

Mean(s.d.)

Mean(s.d.)

Intercept

ξ0

1.452(0.732)

1.272(0.953)

Age

ξ1

0.047(0.057)

0.077(0.078)

Place urban

ξ2

1.107(0.180)

1.427(0.270)

Place semi-urban

ξ3

1.104(0.192)

1.382(0.269)

Gender:Female

ξ4

−1.247(0.149)

−1.528(0.214)

Work

ξ5

0.155(0.145)

0.199(0.184)

Round

ξ6

0.401(0.157)

0.521(0.203)

Std. dev. random effect

d

1.148(0.203)

1.417(0.266)

Ratio

α / β

0.013(0.003)

DIC

 

1943.0

1915.0

pD

 

241.5

211.9

  1. Estimated posterior mean and standard deviation in (1) the logistic model, (2) the beta-binomial model, (3) the logistic-normal model, and (4) the combined model.