I would like to know how to compute a confidence interval for a "geese" object, the regression output for an ordinal generalized estimating equation. This type of object is not accepted by tidy() or conf.int().
See the reproducible example below.
n2 <- 3
n1 <- 7
data <- data.frame(id= c("101", "102", "103", "104", "105", "106", "107", "108", "109", "110",
"101", "102", "103", "104", "105", "106", "107", "108", "109", "110",
"101", "102", "103", "104", "105", "106", "107", "108", "109", "110",
"101", "102", "103", "104", "105", "106", "107", "108", "109", "110",
"101", "102", "103", "104", "105", "106", "107", "108", "109", "110"),
exposure= c("group1", "group3", "group6", "group2", "group4", "group5", "group1", "group3", "group6", "group2",
"group4", "group5", "group7", "group7", "group1", "group3", "group4", "group5", "group7", "group7",
"group6", "group2", "group6", "group2", "group6", "group2", "group6", "group2", "group6", "group2",
"group6", "group2","group6", "group2" , "group6", "group2", "group6", "group2","group6", "group2",
"group6", "group2", "group6", "group2", "group6", "group2", "group6", "group2", "group6", "group2"),
outcome=c("1", "2", "3", "2", "1", "3", "1", "1", "2", "3",
"1", "3", "1", "1", "3", "1", "1", "3", "1", "1",
"1", "3", "1", "1", "3", "1", "1", "3", "1", "1",
"1", "3", "1", "1", "3", "1", "1", "3", "1", "1",
"1", "3", "1", "1", "3", "1", "1", "3", "1", "1"))
data$outcome <- factor(data$outcome, ordered = TRUE)
data <- data %>%
arrange(id)
model1 <- ordgee(outcome ~ exposure,
id = id,
data = data,
corstr= "independence")
summary(model1)
I tried to use conf.int(), tidy() and they do not accept a "geese" object.
I also tried the solutions suggested here
But I got an error when I ran that.
