add ref

vignettes/bibliography.bib

@@ -64,3 +64,18 @@ @article{dickerman_counterfactual_2020
 	year = {2020},
 	pages = {615--617},
 }
+
+
+@misc{ho_matchit_2005,
+	title = {{MatchIt}: {Nonparametric} {Preprocessing} for {Parametric} {Causal} {Inference}},
+	shorttitle = {{MatchIt}},
+	url = {https://CRAN.R-project.org/package=MatchIt},
+	doi = {10.32614/CRAN.package.MatchIt},
+	abstract = {Selects matched samples of the original treated and control groups with similar covariate distributions -- can be used to match exactly on covariates, to match on propensity scores, or perform a variety of other matching procedures.  The package also implements a series of recommendations offered in Ho, Imai, King, and Stuart (2007) {\textless}DOI:10.1093/pan/mpl013{\textgreater}. (The  'gurobi' package, which is not on CRAN, is optional and comes with  an installation of the Gurobi Optimizer, available at  {\textless}https://www.gurobi.com{\textgreater}.)},
+	language = {en},
+	urldate = {2025-02-23},
+	author = {Ho, Daniel and Imai, Kosuke and King, Gary and Stuart, Elizabeth and Greifer, Noah},
+	month = jan,
+	year = {2005},
+	note = {Institution: Comprehensive R Archive Network}
+}

vignettes/practical_causality.Rmd

@@ -42,7 +42,7 @@ Estimating causal effects with observational data, where treatment and control g
 
 ## Propensity scores and G-computation
 
-Two primary methods exist for addressing the lack of randomization in observational data: propensity scores and g-computation. Regarding propensity scores, this vignette focuses on _inverse probability weighting_ (IPW), a common technique for estimating propensity scores [@chatton_causal_2024; @gabriel_inverse_2024].
+Two primary methods exist for addressing the lack of randomization in observational data: propensity scores and g-computation. Regarding propensity scores, this vignette focuses on _inverse probability weighting_ (IPW), a common technique for estimating propensity scores [@chatton_causal_2024; @gabriel_inverse_2024]. Other established techniques involve _matching_ [@ho_matchit_2005], which are, however, beyond the scope of this discussion.
 
 IPW assigns weights to individual observations to reflect their contribution to the outcome under the assumption of exchangeability between groups. When specific characteristics are over-represented in the treatment group, IPW assigns lower weights to these individuals, thereby adjusting for the imbalanced distribution of confounders between treatment and control groups.