Skip to main content

Table 3 Jimma Longitudinal Family Survey of Youth

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

Effect

Parameter

Logistic

Beta-binomial

  

Estimate (s.e.,p)

Estimate (s.e.,p)

Intercept

ξ0

1.171(0.626, 0.061)

1.155(0.702, 0.099)

Age

ξ1

0.039(0.049, 0.414)

0.044(0.055, 0.421)

Place urban

ξ2

0.971(0.148, 0.001)

1.089(0.266, 0.001)

Place semi-urban

ξ3

0.979(0.159, 0.001)

1.104(0.284, 0.001)

Gender:Female

ξ4

1.111(0.123, 0.001)

1.226(0.237, 0.001)

Work

ξ5

0.134(0.122, 0.274)

0.146(0.138, 0.288)

Round

ξ6

0.341(0.141, 0.016)

0.390(0.178, 0.029)

Std. dev. random effect

d

Ratio

α / β

0.009(0.014, 0.528)

2log-likelihood

 

1987.7

1987.4

Effect

Parameter

Logistic-normal

Combined

  

Estimate (s.e., p )

Estimate (s.e., p )

Intercept

ξ0

1.443(0.719, 0.045)

1.463(0.888, 0.099)

Age

ξ1

0.046(0.056, 0.408)

0.058(0.070, 0.408)

Place urban

ξ2

1.098(0.178, 0.001)

1.379(0.393, 0.001)

Place semi-urban

ξ3

1.092(0.189, 0.001)

1.339(0.368, 0.001)

Gender:Female

ξ4

1.241(0.147, 0.001)

1.499(0.339, 0.001)

Work

ξ5

0.153(0.144, 0.287)

0.189(0.182, 0.296)

Round

ξ6

0.398(0.155, 0.010)

0.519(0.237, 0.028)

Std. dev. random effect

d

1.138(0.188, 0.001)

1.342(0.318, 0.001)

Ratio

α / β

0.013(0.013, 0.293)

−2log-likelihood

 

1972.9

1972.1

  1. Parameter estimates, standard errors, and p-values for the regression coefficients in (1) the logistic model, (2) the beta-binomial model, (3) the logistic-normal model, and (4) the combined model.
  2. Estimation was done by maximum likelihood using numerical integration over the normal random effect, if present.