CODEWITHSUNDEEP
imputationr-micemediator
Dale
Dale

Reputation: 1

Using pool.scalar() for multivariate analysis to pool data after multiple imputation

My study aims to examine multiple parallel mediators between a predictor and an outcome variable. I've used mice to run multiple imputations for the missing data. I would like to use the PROCESS function in bruceR package to conduct the multiple mediation.

Upon checking the relevant discussions, I understand that bruceR is not incorporated into broom so I received error messages after using the mice::pool function.

I also understand that I can use pool.scalar() function to extract the analysis myself. However, I can only find instructions for using pool.scalar() for univariate analysis. Since my analysis involves multiple mediators (5 in total), can anyone please advise on the steps for multivariate analysis or direct me to the resources on this? I'm afraid I can't find the relevant guidance online...

Upvotes: 0

Views: 324

Answers (1)

hanne
hanne

Reputation: 102

There's now a new function in mice (v3.16.0) to pool estimates of analyses that are not recognized by the broom workflow underlying the pool() function. See the pool.table() documentation on https://amices.org/mice/reference/pool.table.

# conventional mice workflow
imp <- mice(nhanes2, m = 2, maxit = 2, seed = 1, print = FALSE)
fit <- with(imp, lm(chl ~ age + bmi + hyp))
pld1 <- pool(fit)
pld1$pooled

# using pool.table() on tidy table
tbl <- summary(fit)[, c("term", "estimate", "std.error", "df.residual")]
tbl
pld2 <- pool.table(tbl, type = "minimal")
pld2

identical(pld1$pooled, pld2)

# conventional workflow: all numerical output
all1 <- summary(pld1, type = "all", conf.int = TRUE)
all1

# pool.table workflow: all numerical output
all2 <- pool.table(tbl)
all2

identical(data.frame(all1), all2)

Upvotes: 0

Related Questions