estimate_mdiff_contrast_bs: Estimate the mean difference for an independent groups...

In rcalinjageman/esci2: Effect Sizes and Confidence Intervals

Description

\loadmathjax

estimate_mdiff_contrast_bs returns the point estimate and confidence interval for the mean difference in a linear contrast: \mjdeqn \psi = \sum_i=1^ac_iM_i psi = sum(contrasts*means) Where there are a groups, and M is each group mean and c is each group weight; see Kline, equation 7.1

Usage

estimate_mdiff_contrast_bs(
  data = NULL,
  grouping_variable = NULL,
  outcome_variable = NULL,
  means = NULL,
  sds = NULL,
  ns = NULL,
  group_labels = NULL,
  grouping_variable_name = "My grouping variable",
  outcome_variable_name = "My outcome variable",
  contrast = NULL,
  conf_level = 0.95,
  assume_equal_variance = FALSE,
  save_raw_data = TRUE
)

Arguments

data	For raw data - a dataframe or tibble
grouping_variable	For raw data - The column name of the grouping variable, or a vector of group names
outcome_variable	For raw data - The column name of the outcome variable, or a vector of numeric data
means	For summary data - A vector of 2 or more means
sds	For summary data - A vector of standard deviations, same length as means
ns	For summary data - A vector of sample sizes, same length as means
group_labels	For summary data - An optional vector of group labels, same length as means
grouping_variable_name	Optional friendly name for the grouping variable. Defaults to 'My grouping variable' or the grouping variable column name if a data.frame is passed.
outcome_variable_name	Optional friendly name for the outcome variable. Defaults to 'My outcome variable' or the outcome variable column name if a data frame is passed.
contrast	A vector of group weights.
conf_level	The confidence level for the confidence interval. Given in decimal form. Defaults to 0.95.
assume_equal_variance	Defaults to FALSE
save_raw_data	For raw data; defaults to TRUE; set to FALSE to save memory by not returning raw data in estimate object

Value

Returnsobject of class esci_estimate

Details

This is a friendly version of CI_mdiff_contrast_bs

• This friendly version can handle raw data and summary data input.

• This friendly version returns an esci_estimate object which provides additional supporting information beyond the effect size and CI.

• All the computational details for this analayis are documented in CI_mdiff_contrast_bs

References

• Cumming, G., & Calin-Jageman, R. J. (2017). Introduction to the new statistics: Estimation, open science, and beyond. Routledge.

See Also

• plot_mdiff_contrast_bs to visualize the results

• CI_mdiff_contrast_bs for a version of this function focused just on the effect size and CI

Examples

# From Raw Data ------------------------------------
# Just pass in the data source, grouping column, and outcome column.
# You can pass these in by position, skipping the labels:

# Note... not sure if PlantGrowth dataset meets assumptions for this analysis
estimate_mdiff_contrast_bs(
 PlantGrowth, 
 group, 
 weight,
 contrast = c('ctrl' = -1, 'trt1' = 1)
)

rcalinjageman/esci2 documentation built on Dec. 22, 2021, 1:02 p.m.