Prevalencia (prevere) y Extensión (cpodbere) por ciudad, con IC y p valor de Wald test, creo que para extensión solo miramos los intervalos de confianza.
Regresión Logística Uni-variada para prevere: 1) país 2) sexo 3) Escolmaerecat23 4) tipoesc 5) freqescovacat 6) usocreme 7) fluorprof Regresión Poisson Uni-variada para cpodbere: 1) país 2) sexo 3) Escolmaerecat23 4) tipoesc 5) freqescovacat 6) usocreme 7) fluorprof Si todo es parecido prevoms y cpodoms La Multi-variada en ambos modelos irían: 1) país 3) escmaerecat23 4) tipoesc 5) freqescovat Van todas las que queden <0,10 en la uni-variada. Por eso va variable país. Espero, que puedas el finde…..Besos Anun
# 2 de Noviembre 2013
#
# load('~/Dropbox/odontologia/maestria_anunziatta/BancoUnidoBrasUru/BRASURU(28102013).RData')
load("~/Dropbox/odontologia/maestria_anunziatta/BancoUnidoBrasUru/BRASURU(02112013).RData")
library(survey)
##
## Attaching package: 'survey'
##
## The following object(s) are masked from 'package:graphics':
##
## dotchart
levels(disenio_urubra$variables$prevbere)
## [1] "1-LC" "2-NTL"
addmargins(table(disenio_urubra$variables$prevbere, disenio_urubra$variables$pais))
##
## Bra Uru Sum
## 1-LC 986 764 1750
## 2-NTL 542 390 932
## Sum 1528 1154 2682
addmargins(table(disenio_urubra$variables$cpodbere, disenio_urubra$variables$pais))
##
## Bra Uru Sum
## 0 542 390 932
## 1 272 131 403
## 2 228 104 332
## 3 154 98 252
## 4 129 107 236
## 5 77 100 177
## 6 50 64 114
## 7 23 56 79
## 8 18 27 45
## 9 14 23 37
## 10 11 22 33
## 11 4 11 15
## 12 2 10 12
## 13 1 1 2
## 14 0 4 4
## 15 0 1 1
## 17 1 1 2
## 18 2 1 3
## 20 0 1 1
## 21 0 1 1
## 26 0 1 1
## Sum 1528 1154 2682
round(svymean(~prevbere, disenio_urubra, deff = TRUE) * 100, 2)
## mean SE DEff
## prevbere1-LC 65.1000 0.0184 4.23
## prevbere2-NTL 34.9000 0.0184 4.23
round(confint(svymean(~prevbere, disenio_urubra, deff = TRUE) * 100), 2)
## 2.5 % 97.5 %
## prevbere1-LC 65.06 65.14
## prevbere2-NTL 34.86 34.94
round(svymean(~cpodbere, disenio_urubra, deff = TRUE), 2) * 1
## mean SE DEff
## cpodbere 2.440 0.145 7.43
round(confint(svymean(~cpodbere, disenio_urubra, deff = TRUE)), 2)
## 2.5 % 97.5 %
## cpodbere 2.15 2.72
round(svyby(~prevbere, ~pais, disenio_urubra, svymean, deff = TRUE)[, 2:4] *
100, 2)
## prevbere1-LC prevbere2-NTL se.prevbere1-LC
## Bra 63.53 36.47 2.06
## Uru 66.29 33.71 2.76
round(confint(svyby(~prevbere, ~pais, disenio_urubra, svymean, deff = TRUE)) *
100, 2)
## 2.5 % 97.5 %
## Bra:prevbere1-LC 59.49 67.56
## Uru:prevbere1-LC 60.89 71.70
## Bra:prevbere2-NTL 32.44 40.51
## Uru:prevbere2-NTL 28.30 39.11
round(svyby(~cpodbere, ~pais, disenio_urubra, svymean, deff = TRUE)[, 2:4] *
1, 2)
## cpodbere se DEff.cpodbere
## Bra 1.96 0.10 3.05
## Uru 2.80 0.22 6.12
round(confint(svyby(~cpodbere, ~pais, disenio_urubra, svymean, deff = TRUE)) *
1, 2)
## 2.5 % 97.5 %
## Bra 1.76 2.16
## Uru 2.37 3.24
TABLAS
tabla1 <- svyby(~prevbere, ~pais, disenio_urubra, svymean, na.rm = TRUE, deff = TRUE)
round(tabla1[, 2:4] * 100, 1)
## prevbere1-LC prevbere2-NTL se.prevbere1-LC
## Bra 63.5 36.5 2.1
## Uru 66.3 33.7 2.8
round(confint(tabla1) * 100, 2)
## 2.5 % 97.5 %
## Bra:prevbere1-LC 59.49 67.56
## Uru:prevbere1-LC 60.89 71.70
## Bra:prevbere2-NTL 32.44 40.51
## Uru:prevbere2-NTL 28.30 39.11
svychisq(~prevbere + pais, disenio_urubra, statistic = "Chisq")
##
## Pearson's X^2: Rao & Scott adjustment
##
## data: svychisq(~prevbere + pais, disenio_urubra, statistic = "Chisq")
## X-squared = 2.218, df = 1, p-value = 0.4239
levels(disenio_urubra$variables$sexoinv) <- c("M", "F")
tabla2 <- svyby(~prevbere, ~sexoinv + pais, disenio_urubra, svymean, keep.var = TRUE,
na.rm = TRUE, deff = TRUE)
round(tabla2[, 3:5] * 100, 1)
## prevbere1-LC prevbere2-NTL se.prevbere1-LC
## M.Bra 60.6 39.4 2.5
## F.Bra 66.5 33.5 2.3
## M.Uru 67.4 32.6 3.9
## F.Uru 65.1 34.9 3.3
tabla3 <- svyby(~prevbere, ~socioecon4cat + pais, disenio_urubra, svymean, na.rm = TRUE,
deff = TRUE)
round(tabla3[, 3:5] * 100, 1)
## prevbere1-LC prevbere2-NTL se.prevbere1-LC
## 1-alto.Bra 52.0 48.0 3.5
## 2-medio-alto.Bra 54.6 45.4 2.7
## 3-medio-bajo.Bra 68.3 31.7 2.0
## 4-bajo.Bra 72.1 27.9 4.5
## 1-alto.Uru 47.2 52.8 6.0
## 2-medio-alto.Uru 65.2 34.8 4.2
## 3-medio-bajo.Uru 72.4 27.6 3.6
## 4-bajo.Uru 80.0 20.0 3.4
tabla4 <- svyby(~prevbere, ~socioecon3cat + pais, disenio_urubra, svymean, na.rm = TRUE,
deff = TRUE)
round(tabla4[, 3:5] * 100, 1)
## prevbere1-LC prevbere2-NTL se.prevbere1-LC
## 1-alto.Bra 52.0 48.0 3.5
## 2-medio.Bra 64.1 35.9 2.0
## 3-bajo.Bra 72.1 27.9 4.5
## 1-alto.Uru 51.4 48.6 4.6
## 2-medio.Uru 71.0 29.0 2.5
## 3-bajo.Uru 80.0 20.0 3.4
tabla5 <- svyby(~prevbere, ~escolmaerecat23cat + pais, disenio_urubra, svymean,
na.rm = TRUE, deff = TRUE)
round(tabla5[, 3:5] * 100, 1)
## prevbere1-LC prevbere2-NTL se.prevbere1-LC
## 1-college-university.Bra 51.9 48.1 4.1
## 2-high school.Bra 58.5 41.5 2.7
## 3-elementary school.Bra 71.2 28.8 2.2
## 1-college-university.Uru 53.0 47.0 5.7
## 2-high school.Uru 67.4 32.6 2.6
## 3-elementary school.Uru 78.9 21.1 3.2
tabla5a <- svyby(~prevbere, ~escolmae13cat + pais, disenio_urubra, svymean,
na.rm = TRUE, deff = TRUE)
round(tabla5a[, 3:5] * 100, 1)
## prevbere1-LC prevbere2-NTL se.prevbere1-LC
## 1-college-university.Bra 52.2 47.8 4.0
## 2-high school.Bra 53.8 46.2 2.7
## 3-Elementary School.Bra 70.9 29.1 2.1
## 1-college-university.Uru 44.6 55.4 4.2
## 2-high school.Uru 63.6 36.4 5.5
## 3-Elementary School.Uru 73.8 26.2 2.6
tabla6 <- svyby(~prevbere, ~freqescovacat + pais, disenio_urubra, svymean, na.rm = TRUE,
deff = TRUE)
round(tabla6[, 3:5] * 100, 1)
## prevbere1-LC prevbere2-NTL se.prevbere1-LC
## 1-menos de 1 vez al d\xeda.Bra 74.7 25.3 2.8
## 2-veces al d\xeda.Bra 61.5 38.5 2.5
## 3-veces al d\xeda.Bra 59.2 40.8 3.0
## 1-menos de 1 vez al d\xeda.Uru 73.8 26.2 4.0
## 2-veces al d\xeda.Uru 66.1 33.9 4.0
## 3-veces al d\xeda.Uru 63.3 36.7 3.3
tabla7 <- svyby(~prevbere, ~usocreme + pais, disenio_urubra, svymean, na.rm = TRUE,
deff = TRUE)
round(tabla7[, 3:5] * 100, 1)
## prevbere1-LC prevbere2-NTL se.prevbere1-LC
## 1-Si.Bra 63.3 36.7 2.0
## 2-No.Bra 75.8 24.2 9.0
## 1-Si.Uru 66.1 33.9 2.8
## 2-No.Uru 73.8 26.2 10.9
tabla8 <- svyby(~prevbere, ~visidentcatonde + pais, disenio_urubra, svymean,
na.rm = TRUE, deff = TRUE)
round(tabla8[, 3:5] * 100, 1)
## prevbere1-LC prevbere2-NTL se.prevbere1-LC
## 1-Convenio particular.Bra 57.9 42.1 2.1
## 2-Publico.Bra 71.5 28.5 2.6
## 3-Nunca fue al dentista.Bra 65.3 34.7 3.1
## 1-Convenio particular.Uru 58.3 41.7 3.4
## 2-Publico.Uru 78.5 21.5 2.7
## 3-Nunca fue al dentista.Uru 76.3 23.7 5.1
tabla9 <- svyby(~prevbere, ~visiquando + pais, disenio_urubra, svymean, na.rm = TRUE,
deff = TRUE)
round(tabla9[, 3:5] * 100, 1)
## prevbere1-LC prevbere2-NTL se.prevbere1-LC
## 1- menos de 1 a\xf1o.Bra 63.4 36.6 2.4
## 2- m\xe1s de 2 a\xf1os.Bra 62.1 37.9 2.8
## 3- nunca fue.Bra 65.3 34.7 3.1
## 1- menos de 1 a\xf1o.Uru 63.6 36.4 3.0
## 2- m\xe1s de 2 a\xf1os.Uru 70.2 29.8 5.5
## 3- nunca fue.Uru 77.4 22.6 5.1
tabla10 <- svyby(~prevbere, ~fluorprof + pais, disenio_urubra, svymean, na.rm = TRUE,
deff = TRUE)
round(tabla10[, 3:5] * 100, 1)
## prevbere1-LC prevbere2-NTL se.prevbere1-LC
## 1-Si.Bra 62.7 37.3 2.1
## 2-No.Bra 65.0 35.0 2.9
## 1-Si.Uru 65.6 34.4 3.6
## 2-No.Uru 65.8 34.2 3.9
tabla11 <- svyby(~prevbere, ~isg20 + pais, disenio_urubra, svymean, na.rm = TRUE,
deff = TRUE)
round(tabla11[, 3:5] * 100, 1)
## prevbere1-LC prevbere2-NTL se.prevbere1-LC
## 1- <=20.Bra 60.5 39.5 7.4
## 2- 20 a 60.Bra 57.9 42.1 2.2
## 3- >=60.Bra 75.6 24.4 2.1
## 1- <=20.Uru 65.0 35.0 4.3
## 2- 20 a 60.Uru 66.5 33.5 3.4
## 3- >=60.Uru 73.2 26.8 5.2
tabla12 <- svyby(~prevbere, ~isg45 + pais, disenio_urubra, svymean, na.rm = TRUE,
deff = TRUE)
round(tabla12[, 3:5] * 100, 1)
## prevbere1-LC prevbere2-NTL se.prevbere1-LC
## 1- <=45.Bra 52.9 47.1 2.8
## 2- 45 a 60.Bra 63.1 36.9 2.5
## 3- >=60.Bra 75.6 24.4 2.1
## 1- <=45.Uru 64.3 35.7 3.3
## 2- 45 a 60.Uru 74.4 25.6 4.6
## 3- >=60.Uru 73.2 26.8 5.2
tabla13 <- svyby(~prevbere, ~tipoesc + pais, disenio_urubra, svymean, na.rm = TRUE,
deff = TRUE)
round(tabla13[, 3:5] * 100, 1)
## prevbere1-LC prevbere2-NTL se.prevbere1-LC
## 1-Particular.Bra 47.9 52.1 2.9
## 2-Publica.Bra 67.9 32.1 1.9
## 1-Particular.Uru 53.8 46.2 4.0
## 2-Publica.Uru 70.9 29.1 2.7
tabla13 <- svyby(~cpodbere, ~pais, disenio_urubra, svymean, na.rm = TRUE, deff = TRUE)
round(tabla13[, 2:4] * 1, 1)
## cpodbere se DEff.cpodbere
## Bra 2.0 0.1 3.0
## Uru 2.8 0.2 6.1
round(confint(tabla13) * 1, 2)
## 2.5 % 97.5 %
## Bra 1.76 2.16
## Uru 2.37 3.24
tabla14 <- svyby(~cpodbere, ~sexoinv + pais, disenio_urubra, svymean, keep.var = TRUE,
na.rm = TRUE, deff = TRUE)
round(tabla14[, 3:5] * 1, 1)
## cpodbere se DEff.cpodbere
## M.Bra 1.9 0.1 2.3
## F.Bra 2.0 0.1 1.7
## M.Uru 2.8 0.3 4.3
## F.Uru 2.8 0.2 4.2
tabla15 <- svyby(~cpodbere, ~socioecon4cat + pais, disenio_urubra, svymean,
na.rm = TRUE, deff = TRUE)
round(tabla15[, 3:5] * 1, 1)
## cpodbere se DEff.cpodbere
## 1-alto.Bra 1.2 0.1 0.9
## 2-medio-alto.Bra 1.5 0.1 0.9
## 3-medio-bajo.Bra 2.2 0.1 2.0
## 4-bajo.Bra 2.4 0.2 1.1
## 1-alto.Uru 1.7 0.4 3.3
## 2-medio-alto.Uru 2.7 0.4 7.1
## 3-medio-bajo.Uru 3.5 0.4 3.4
## 4-bajo.Uru 3.2 0.3 2.8
tabla16 <- svyby(~cpodbere, ~socioecon3cat + pais, disenio_urubra, svymean,
na.rm = TRUE, deff = TRUE)
round(tabla16[, 3:5] * 1, 1)
## cpodbere se DEff.cpodbere
## 1-alto.Bra 1.2 0.1 0.9
## 2-medio.Bra 2.0 0.1 2.2
## 3-bajo.Bra 2.4 0.2 1.1
## 1-alto.Uru 2.0 0.3 3.5
## 2-medio.Uru 3.2 0.2 3.4
## 3-bajo.Uru 3.2 0.3 2.8
tabla17 <- svyby(~cpodbere, ~escolmaerecat23cat + pais, disenio_urubra, svymean,
na.rm = TRUE, deff = TRUE)
round(tabla17[, 3:5] * 1, 1)
## cpodbere se DEff.cpodbere
## 1-college-university.Bra 1.2 0.1 1.8
## 2-high school.Bra 1.7 0.1 0.8
## 3-elementary school.Bra 2.4 0.1 2.1
## 1-college-university.Uru 2.0 0.4 4.7
## 2-high school.Uru 2.9 0.2 2.0
## 3-elementary school.Uru 3.5 0.4 4.7
tabla17a <- svyby(~cpodbere, ~escolmae13cat + pais, disenio_urubra, svymean,
na.rm = TRUE, deff = TRUE)
round(tabla17a[, 3:5] * 1, 1)
## cpodbere se DEff.cpodbere
## 1-college-university.Bra 1.2 0.1 0.9
## 2-high school.Bra 1.5 0.1 1.1
## 3-Elementary School.Bra 2.4 0.1 2.0
## 1-college-university.Uru 1.6 0.3 2.3
## 2-high school.Uru 2.6 0.4 4.5
## 3-Elementary School.Uru 3.2 0.2 3.6
tabla18 <- svyby(~cpodbere, ~freqescovacat + pais, disenio_urubra, svymean,
na.rm = TRUE, deff = TRUE)
round(tabla18[, 3:5] * 1, 1)
## cpodbere se DEff.cpodbere
## 1-menos de 1 vez al d\xeda.Bra 2.5 0.1 0.8
## 2-veces al d\xeda.Bra 1.9 0.1 2.5
## 3-veces al d\xeda.Bra 1.6 0.1 1.5
## 1-menos de 1 vez al d\xeda.Uru 3.4 0.3 2.0
## 2-veces al d\xeda.Uru 2.8 0.3 3.9
## 3-veces al d\xeda.Uru 2.6 0.3 3.8
tabla19 <- svyby(~cpodbere, ~usocreme + pais, disenio_urubra, svymean, na.rm = TRUE,
deff = TRUE)
round(tabla19[, 3:5] * 1, 1)
## cpodbere se DEff.cpodbere
## 1-Si.Bra 1.9 0.1 2.9
## 2-No.Bra 2.7 0.6 1.4
## 1-Si.Uru 2.8 0.2 5.9
## 2-No.Uru 2.6 0.8 1.8
tabla20 <- svyby(~cpodbere, ~visidentcatonde + pais, disenio_urubra, svymean,
na.rm = TRUE, deff = TRUE)
round(tabla20[, 3:5] * 1, 1)
## cpodbere se DEff.cpodbere
## 1-Convenio particular.Bra 1.7 0.1 1.7
## 2-Publico.Bra 2.4 0.2 1.9
## 3-Nunca fue al dentista.Bra 2.1 0.2 2.0
## 1-Convenio particular.Uru 2.4 0.3 7.1
## 2-Publico.Uru 3.3 0.3 2.7
## 3-Nunca fue al dentista.Uru 3.6 0.4 2.9
tabla21 <- svyby(~cpodbere, ~visiquando + pais, disenio_urubra, svymean, na.rm = TRUE,
deff = TRUE)
round(tabla21[, 3:5] * 1, 1)
## cpodbere se DEff.cpodbere
## 1- menos de 1 a\xf1o.Bra 1.9 0.1 2.3
## 2- m\xe1s de 2 a\xf1os.Bra 1.9 0.1 1.4
## 3- nunca fue.Bra 2.1 0.2 2.0
## 1- menos de 1 a\xf1o.Uru 2.6 0.3 7.0
## 2- m\xe1s de 2 a\xf1os.Uru 3.1 0.5 4.1
## 3- nunca fue.Uru 3.6 0.4 2.8
tabla22 <- svyby(~cpodbere, ~fluorprof + pais, disenio_urubra, svymean, na.rm = TRUE,
deff = TRUE)
round(tabla22[, 3:5] * 1, 2)
## cpodbere se DEff.cpodbere
## 1-Si.Bra 1.83 0.10 2.08
## 2-No.Bra 2.20 0.16 2.15
## 1-Si.Uru 2.80 0.27 4.81
## 2-No.Uru 2.77 0.29 3.79
tabla23 <- svyby(~cpodbere, ~isg20 + pais, disenio_urubra, svymean, na.rm = TRUE,
deff = TRUE)
round(tabla23[, 3:5] * 1, 1)
## cpodbere se DEff.cpodbere
## 1- <=20.Bra 1.7 0.4 1.4
## 2- 20 a 60.Bra 1.6 0.1 2.4
## 3- >=60.Bra 2.8 0.2 1.6
## 1- <=20.Uru 2.6 0.3 7.5
## 2- 20 a 60.Uru 2.9 0.2 2.4
## 3- >=60.Uru 3.7 0.4 1.3
tabla24 <- svyby(~cpodbere, ~isg45 + pais, disenio_urubra, svymean, na.rm = TRUE,
deff = TRUE)
round(tabla24[, 3:5] * 1, 1)
## cpodbere se DEff.cpodbere
## 1- <=45.Bra 1.4 0.1 1.7
## 2- 45 a 60.Bra 1.8 0.1 1.7
## 3- >=60.Bra 2.8 0.2 1.6
## 1- <=45.Uru 2.6 0.3 7.4
## 2- 45 a 60.Uru 3.3 0.4 1.4
## 3- >=60.Uru 3.7 0.4 1.3
tabla25 <- svyby(~cpodbere, ~escolmae13cat + pais, disenio_urubra, svymean,
na.rm = TRUE, deff = TRUE)
round(tabla25[, 3:5] * 1, 1)
## cpodbere se DEff.cpodbere
## 1-college-university.Bra 1.2 0.1 0.9
## 2-high school.Bra 1.5 0.1 1.1
## 3-Elementary School.Bra 2.4 0.1 2.0
## 1-college-university.Uru 1.6 0.3 2.3
## 2-high school.Uru 2.6 0.4 4.5
## 3-Elementary School.Uru 3.2 0.2 3.6
tabla26 <- svyby(~cpodbere, ~tipoesc + pais, disenio_urubra, svymean, na.rm = TRUE,
deff = TRUE)
round(tabla26[, 3:5] * 1, 1)
## cpodbere se DEff.cpodbere
## 1-Particular.Bra 1.2 0.1 1.0
## 2-Publica.Bra 2.2 0.1 2.0
## 1-Particular.Uru 1.9 0.1 0.8
## 2-Publica.Uru 3.1 0.3 6.0
RL Univariada
1) país 2) sexo 3) Escolmaerecat23 4) tipoesc 5) freqescovacat 6) usocreme 7) fluorprof
modelo1.bin <- svyglm(prevbere ~ pais, disenio_urubra, family = quasibinomial())
summary(modelo1.bin)
##
## Call:
## svyglm(formula = prevbere ~ pais, disenio_urubra, family = quasibinomial())
##
## Survey design:
## svydesign(id = ~numer_esc, strata = ~pais, weights = ~round(weight.rec,
## 1), data = brasuru, nest = TRUE)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -0.5548 0.0888 -6.25 1.9e-08 ***
## paisUru -0.1216 0.1521 -0.80 0.43
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for quasibinomial family taken to be 1)
##
## Number of Fisher Scoring iterations: 4
confint(modelo1.bin)
## 2.5 % 97.5 %
## (Intercept) -0.7288 -0.3808
## paisUru -0.4196 0.1765
modelo2.bin <- svyglm(prevbere ~ pais + sexoinv, disenio_urubra, family = quasibinomial())
summary(modelo2.bin)
##
## Call:
## svyglm(formula = prevbere ~ pais + sexoinv, disenio_urubra, family = quasibinomial())
##
## Survey design:
## svydesign(id = ~numer_esc, strata = ~pais, weights = ~round(weight.rec,
## 1), data = brasuru, nest = TRUE)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -0.5293 0.1120 -4.73 9.8e-06 ***
## paisUru -0.1221 0.1519 -0.80 0.42
## sexoinvF -0.0519 0.1334 -0.39 0.70
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for quasibinomial family taken to be 1)
##
## Number of Fisher Scoring iterations: 4
confint(modelo2.bin)
## 2.5 % 97.5 %
## (Intercept) -0.7489 -0.3098
## paisUru -0.4198 0.1756
## sexoinvF -0.3134 0.2096
modelo3.bin <- svyglm(prevbere ~ pais + escolmaerecat23cat, disenio_urubra,
famil3y = quasibinomial())
## Error: unused argument(s) (famil3y = list(family = "quasibinomial", link = "logit", linkfun = function (mu)
## .Call(C_logit_link, mu), linkinv = function (eta)
## .Call(C_logit_linkinv, eta), variance = function (mu)
## mu * (1 - mu), dev.resids = function (y, mu, wt)
## 2 * wt * (y * log(ifelse(y == 0, 1, y/mu)) + (1 - y) * log(ifelse(y == 1, 1, (1 - y)/(1 - mu)))), aic = function (y, n, mu, wt, dev)
## NA, mu.eta = function (eta)
## .Call(C_logit_mu_eta, eta), initialize = expression({
## if (NCOL(y) == 1) {
## if (is.factor(y)) y <- y != levels(y)[1]
## n <- rep.int(1, nobs)
## if (any(y < 0 | y > 1)) stop("y values must be 0 <= y <= 1")
## mustart <- (weights * y + 0.5)/(weights + 1)
## } else if (NCOL(y) == 2) {
## n <- y[, 1] + y[, 2]
## y <- ifelse(n == 0, 0, y[, 1]/n)
## weights <- weights * n
## mustart <- (n * y + 0.5)/(n + 1)
## } else stop("for the quasibinomial family, y must be a vector of 0 and 1's\n", "or a 2 column matrix where col 1 is no. successes and col 2 is no. failures")
## }), validmu = function (mu)
## all(mu > 0) && all(mu < 1), valideta = function (eta)
## TRUE))
summary(modelo3.bin)
## Error: object 'modelo3.bin' not found
confint(modelo3.bin)
## Error: object 'modelo3.bin' not found
modelo4.bin <- svyglm(prevbere ~ pais + tipoesc, disenio_urubra, family = quasibinomial())
summary(modelo4.bin)
##
## Call:
## svyglm(formula = prevbere ~ pais + tipoesc, disenio_urubra, family = quasibinomial())
##
## Survey design:
## svydesign(id = ~numer_esc, strata = ~pais, weights = ~round(weight.rec,
## 1), data = brasuru, nest = TRUE)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.0387 0.1145 0.34 0.74
## paisUru -0.1645 0.1333 -1.23 0.22
## tipoesc2-Publica -0.7747 0.1296 -5.98 6.3e-08 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for quasibinomial family taken to be 1)
##
## Number of Fisher Scoring iterations: 4
confint(modelo4.bin)
## 2.5 % 97.5 %
## (Intercept) -0.1856 0.26312
## paisUru -0.4256 0.09671
## tipoesc2-Publica -1.0288 -0.52055
modelo5.bin <- svyglm(prevbere ~ pais + freqescovacat, disenio_urubra, family = quasibinomial())
summary(modelo5.bin)
##
## Call:
## svyglm(formula = prevbere ~ pais + freqescovacat, disenio_urubra,
## family = quasibinomial())
##
## Survey design:
## svydesign(id = ~numer_esc, strata = ~pais, weights = ~round(weight.rec,
## 1), data = brasuru, nest = TRUE)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -0.983 0.133 -7.41 1.3e-10 ***
## paisUru -0.148 0.156 -0.94 0.34774
## freqescovacat2-veces al d\xeda 0.490 0.139 3.52 0.00073 ***
## freqescovacat3-veces al d\xeda 0.596 0.142 4.20 7.0e-05 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for quasibinomial family taken to be 1.001)
##
## Number of Fisher Scoring iterations: 4
confint(modelo5.bin)
## 2.5 % 97.5 %
## (Intercept) -1.2435 -0.7235
## paisUru -0.4545 0.1589
## freqescovacat2-veces al d\xeda 0.2169 0.7636
## freqescovacat3-veces al d\xeda 0.3178 0.8740
modelo6.bin <- svyglm(prevbere ~ pais + usocreme, disenio_urubra, family = quasibinomial())
summary(modelo6.bin)
##
## Call:
## svyglm(formula = prevbere ~ pais + usocreme, disenio_urubra,
## family = quasibinomial())
##
## Survey design:
## svydesign(id = ~numer_esc, strata = ~pais, weights = ~round(weight.rec,
## 1), data = brasuru, nest = TRUE)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -0.5471 0.0877 -6.24 2e-08 ***
## paisUru -0.1191 0.1528 -0.78 0.44
## usocreme2-No -0.4658 0.3971 -1.17 0.24
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for quasibinomial family taken to be 1.002)
##
## Number of Fisher Scoring iterations: 4
confint(modelo6.bin)
## 2.5 % 97.5 %
## (Intercept) -0.7190 -0.3752
## paisUru -0.4185 0.1804
## usocreme2-No -1.2441 0.3126
modelo7.bin <- svyglm(prevbere ~ pais + fluorprof, disenio_urubra, family = quasibinomial())
summary(modelo7.bin)
##
## Call:
## svyglm(formula = prevbere ~ pais + fluorprof, disenio_urubra,
## family = quasibinomial())
##
## Survey design:
## svydesign(id = ~numer_esc, strata = ~pais, weights = ~round(weight.rec,
## 1), data = brasuru, nest = TRUE)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -0.5377 0.0964 -5.58 3.3e-07 ***
## paisUru -0.0933 0.1569 -0.59 0.55
## fluorprof2-No -0.0488 0.1280 -0.38 0.70
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for quasibinomial family taken to be 0.9991)
##
## Number of Fisher Scoring iterations: 4
confint(modelo7.bin)
## 2.5 % 97.5 %
## (Intercept) -0.7267 -0.3488
## paisUru -0.4009 0.2143
## fluorprof2-No -0.2997 0.2021
R POISSON UNIVARIADA
modelo1.poi <- svyglm(cpodbere ~ pais, disenio_urubra, family = quasipoisson())
summary(modelo1.poi)
##
## Call:
## svyglm(formula = cpodbere ~ pais, disenio_urubra, family = quasipoisson())
##
## Survey design:
## svydesign(id = ~numer_esc, strata = ~pais, weights = ~round(weight.rec,
## 1), data = brasuru, nest = TRUE)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.6716 0.0516 13.02 < 2e-16 ***
## paisUru 0.3590 0.0946 3.79 0.00029 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for quasipoisson family taken to be 3.184)
##
## Number of Fisher Scoring iterations: 5
confint(modelo1.poi)
## 2.5 % 97.5 %
## (Intercept) 0.5706 0.7727
## paisUru 0.1735 0.5445
modelo2.poi <- svyglm(cpodbere ~ pais + sexoinv, disenio_urubra, family = quasipoisson())
summary(modelo2.poi)
##
## Call:
## svyglm(formula = cpodbere ~ pais + sexoinv, disenio_urubra, family = quasipoisson())
##
## Survey design:
## svydesign(id = ~numer_esc, strata = ~pais, weights = ~round(weight.rec,
## 1), data = brasuru, nest = TRUE)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.66724 0.06367 10.48 < 2e-16 ***
## paisUru 0.35905 0.09456 3.80 0.00029 ***
## sexoinvF 0.00891 0.07063 0.13 0.89989
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for quasipoisson family taken to be 3.185)
##
## Number of Fisher Scoring iterations: 5
confint(modelo2.poi)
## 2.5 % 97.5 %
## (Intercept) 0.5425 0.7920
## paisUru 0.1737 0.5444
## sexoinvF -0.1295 0.1473
modelo3.poi <- svyglm(cpodbere ~ pais + escolmaerecat23cat, disenio_urubra,
famil3y = quasipoisson())
## Error: unused argument(s) (famil3y = list(family = "quasipoisson", link = "log", linkfun = function (mu)
## log(mu), linkinv = function (eta)
## pmax(exp(eta), .Machine$double.eps), variance = function (mu)
## mu, dev.resids = function (y, mu, wt)
## 2 * wt * (y * log(ifelse(y == 0, 1, y/mu)) - (y - mu)), aic = function (y, n, mu, wt, dev)
## NA, mu.eta = function (eta)
## pmax(exp(eta), .Machine$double.eps), initialize = expression({
## if (any(y < 0)) stop("negative values not allowed for the quasiPoisson family")
## n <- rep.int(1, nobs)
## mustart <- y + 0.1
## }), validmu = function (mu)
## all(mu > 0), valideta = function (eta)
## TRUE))
summary(modelo3.poi)
## Error: object 'modelo3.poi' not found
confint(modelo3.poi)
## Error: object 'modelo3.poi' not found
modelo4.poi <- svyglm(cpodbere ~ pais + tipoesc, disenio_urubra, family = quasipoisson())
summary(modelo4.poi)
##
## Call:
## svyglm(formula = cpodbere ~ pais + tipoesc, disenio_urubra, family = quasipoisson())
##
## Survey design:
## svydesign(id = ~numer_esc, strata = ~pais, weights = ~round(weight.rec,
## 1), data = brasuru, nest = TRUE)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.2191 0.0750 2.92 0.0046 **
## paisUru 0.3816 0.0817 4.67 1.2e-05 ***
## tipoesc2-Publica 0.5502 0.0776 7.09 5.0e-10 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for quasipoisson family taken to be 3.05)
##
## Number of Fisher Scoring iterations: 5
confint(modelo4.poi)
## 2.5 % 97.5 %
## (Intercept) 0.07206 0.3662
## paisUru 0.22146 0.5416
## tipoesc2-Publica 0.39806 0.7023
modelo5.poi <- svyglm(cpodbere ~ pais + freqescovacat, disenio_urubra, family = quasipoisson())
summary(modelo5.poi)
##
## Call:
## svyglm(formula = cpodbere ~ pais + freqescovacat, disenio_urubra,
## family = quasipoisson())
##
## Survey design:
## svydesign(id = ~numer_esc, strata = ~pais, weights = ~round(weight.rec,
## 1), data = brasuru, nest = TRUE)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.8705 0.0661 13.16 < 2e-16 ***
## paisUru 0.3779 0.0933 4.05 0.00012 ***
## freqescovacat2-veces al d\xeda -0.2190 0.0766 -2.86 0.00543 **
## freqescovacat3-veces al d\xeda -0.3189 0.0851 -3.75 0.00034 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for quasipoisson family taken to be 3.096)
##
## Number of Fisher Scoring iterations: 5
confint(modelo5.poi)
## 2.5 % 97.5 %
## (Intercept) 0.7409 1.00015
## paisUru 0.1951 0.56080
## freqescovacat2-veces al d\xeda -0.3691 -0.06892
## freqescovacat3-veces al d\xeda -0.4857 -0.15209
modelo6.poi <- svyglm(cpodbere ~ pais + usocreme, disenio_urubra, family = quasipoisson())
summary(modelo6.poi)
##
## Call:
## svyglm(formula = cpodbere ~ pais + usocreme, disenio_urubra,
## family = quasipoisson())
##
## Survey design:
## svydesign(id = ~numer_esc, strata = ~pais, weights = ~round(weight.rec,
## 1), data = brasuru, nest = TRUE)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.6703 0.0511 13.11 < 2e-16 ***
## paisUru 0.3570 0.0949 3.76 0.00032 ***
## usocreme2-No 0.0725 0.2102 0.35 0.73097
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for quasipoisson family taken to be 3.188)
##
## Number of Fisher Scoring iterations: 5
confint(modelo6.poi)
## 2.5 % 97.5 %
## (Intercept) 0.5701 0.7705
## paisUru 0.1710 0.5429
## usocreme2-No -0.3395 0.4845
modelo7.poi <- svyglm(cpodbere ~ pais + fluorprof, disenio_urubra, family = quasipoisson())
summary(modelo7.poi)
##
## Call:
## svyglm(formula = cpodbere ~ pais + fluorprof, disenio_urubra,
## family = quasipoisson())
##
## Survey design:
## svydesign(id = ~numer_esc, strata = ~pais, weights = ~round(weight.rec,
## 1), data = brasuru, nest = TRUE)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.6500 0.0553 11.75 < 2e-16 ***
## paisUru 0.3523 0.0990 3.56 0.00063 ***
## fluorprof2-No 0.0604 0.0754 0.80 0.42531
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for quasipoisson family taken to be 3.183)
##
## Number of Fisher Scoring iterations: 5
confint(modelo7.poi)
## 2.5 % 97.5 %
## (Intercept) 0.54158 0.7584
## paisUru 0.15836 0.5462
## fluorprof2-No -0.08738 0.2083