margot

MARGinal Observational Treatment-effects.¹

Causal inference requires balance across the treatments to be compared. In observational studies, such balance is not guaranteed; quantifying causality therefore requires careful, multi-step workflows.

The goal of margot is to enhance the accessibility of these workflows for causal inference. Its primary audience includes psychological scientists, although it may benefit other social scientists.

The package offers functions for:

• evaluating causal assumptions
• modelling time-series data
• reporting results
• performing sensitivity analyses

margot focuses on streamlining the estimation of (Marginal) Average Treatment Effects, but it also supports workflows for Conditional Average Treatment Effects and exploring Heterogeneous Treatment Effects, as well as Modified Treatment Policies.

Installation

You can install the development version of margot like so:

if (!require(devtools, quietly = TRUE)) {
  install.packages("devtools")
  library(devtools)
}

devtools::install_github("go-bayes/margot")

Example

library("margot")

# create transition table to evaluate the positivity assumption
transition_matrix <- create_transition_matrix(df_nz, "religion_believe_god", "id")

# create table and table explanation
table_change_belief <- transition_table(transition_matrix)
table_change_belief

Code

Go to:https://github.com/go-bayes/margot

License

The code in this package is licensed under the Creative Commons Attribution 4.0 International (CC BY 4.0) license. You are free to share and adapt the code, even for commercial purposes, provided that you attribute the original author(s) appropriately. For more information, see CC BY 4.0.

The margot package is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the Disclaimer of Warranties and Limitation of Liability section in the licensing information for more details.

Citing margot

If you use the margot package in your research, you may cite it as follows:

bibentry(
  bibtype = "Manual",
  title = "margot: MARGinal Observational Treatment-effects",
  author = person("Joseph A", "Bulbulia"),
  year = "2024",
  note = "R package version 0.3.0.2, Functions to obtain MARGinal Observational Treatment-effects from observational data.",
  url = "https://go-bayes.github.io/margot/",
  doi = "10.5281/zenodo.10907724"
)

Doi

Footnotes

1. The logo is a Single World Intervention Template (SWIT). We use a SWIT to generate Single World Intervention Graphs (SWIGs) – causal diagrams for which identification assumptions can be read separately for each treatment (regime) to be compared. The name margot reflects the contents and aims of this package; it is also the name of my daughter, Margot. ↩