predictmeans (version 1.0.9)

ci_mcp: Multiple Comparisons Based on the Confidence Intervals

Description

This function produces letter representations for a multiple comparison test by analyzing the confidence intervals associated with the mean values of different treatments. In particular, if the confidence intervals of two treatments overlap, it indicates that there is no significant difference between them. Conversely, if the confidence intervals do not overlap, it indicates that the treatments are significantly different from each other.

Usage

ci_mcp(LL, UL, trt_n=NULL)

Arguments

LL

Lower limits of treatments' confidence interval.

UL

Upper limits of treatments' confidence interval.

trt_n

Treatments' names.

Author

Dongwen Luo, Siva Ganesh and John Koolaard

References

Vanessa, C. (05 October 2022), Confidence tricks: the 83.4% confidence interval for comparing means, https://vsni.co.uk/blogs/confidence_trick.

Examples

  library(predictmeans)
  ci_mcp(LL=c(68.2566,  87.7566, 103.0899, 112.2566), UL=c(90.5212, 110.0212, 125.3545, 134.5212))
  
  data("Oats", package="nlme")
  Oats$nitro <- factor(Oats$nitro)
  fm <- lme(yield ~ nitro*Variety, random=~1|Block/Variety, data=Oats)
# fm <- lmer(yield ~ nitro*Variety+(1|Block/Variety), data=Oats)
  predictmeans(fm, "nitro", adj="BH", plot=FALSE)$mean_table
  predictmeans(fm, "nitro", pair=TRUE, level=0.166, letterCI = TRUE, plot=FALSE)$mean_table