Skip to contents

Add confidence intervals to Bland-Altman analysis object.

Source: R/calc_stats.R
add_confint.Rd

The confidence intervals are calculated with a percentile parametric bootstrap (see lme4::confint.merMod() for details).

Usage

add_confint(
  ba_obj,
  level = 0.95,
  nsim = 1999,
  boot.type = c("perc", "basic", "norm"),
  .progress = "txt",
  PBargs = list(style = 3)
)

Arguments

ba_obj

Bland-Altman analysis object

level

Confidence level (default is 0.95)

nsim

Number of bootstrap samples

boot.type

Type of interval (one of "perc" (default), "basic", "norm"). Passed to boot::boot.ci()

.progress, PBargs

see lme4::bootMer()

Value

Bland-Altman analysis object (x) with added confidence intervals

Details

Note that when calculating confidence intervals for percentage errors, the mean value is treated as a known value rather than an estimate. Any uncertainty in the estimate of the mean is not included in the confidence intervals for percentage errors. This is a technical limitation, since the bootstrap analysis is based on a single mixed model (diff ~ 1 + (1|id)) which has no information about the mean value.