Please specify whether your issue is about:

• a possible bug
• a question about package functionality
• a suggested code or documentation change, improvement to the code, or feature request

It seems that there is an issue with calculation of confidence intervals in mm_diffs. Instead of returning a 1 - alpha confidence interval, a 1 - 2*alpha confidence interval is currently returned.
In lines 71-72 of mm_diffs(), the confidence intervals are calculated so:

        # CIs
        mm_split[[i]][["lower"]] <- mm_split[[i]][["estimate"]] - (stats::qnorm(1-alpha) * mm_split[[i]][["std.error"]])
        mm_split[[i]][["upper"]] <- mm_split[[i]][["estimate"]] + (stats::qnorm(1-alpha) * mm_split[[i]][["std.error"]])

But the function documentation specifies that alpha is a "A numeric value indicating the significance level at which to calculate confidence intervals for the MMs (by default 0.95, meaning 95-percent CIs are returned)." The default of alpha = 0.05 gives a 90-percent confidence interval as of right now. I'm sure the solution is obvious to you; changing alpha to alpha/2 in both cases above would fix the issue. Or adding + (alpha/2) within qnorm(), which seems to be what is done in mm() and amce_diffs()

This is mainly a problem for subsequent plotting, I think.

Thank you for this package, it's awesome!

Put your code here:

## load package
library("cregg")

data("immigration")
alpha <- .05
# Differences in MMs by Gender feature
temp <- mm_diffs(immigration, 
                 ChosenImmigrant ~ LanguageSkills + Education,
                 ~ Gender, id = ~ CaseID,
                 alpha = alpha)

#confidence interval from mm_diffs()
temp[, c("lower", "upper")]

#95% confidence interval calculated "by hand"
data.frame(lower = temp$estimate - qnorm(1 - alpha/2) * temp$std.error,
           upper = temp$estimate + qnorm(1 - alpha/2) * temp$std.error)


## session info for your system
sessionInfo()