diff --git a/bibliography.bib b/bibliography.bib
index 2b4854a..5d8aaff 100644
--- a/bibliography.bib
+++ b/bibliography.bib
@@ -489,3 +489,90 @@ @dataset{knight_2025_17250038
   doi          = {10.5281/zenodo.17250038},
   url          = {https://doi.org/10.5281/zenodo.17250038},
 }
+
+@article{open_science_collaboration2015,
+  title={Estimating the reproducibility of psychological science},
+  author={{Open Science Collaboration}},
+  journal={Science},
+  year={2015},
+  volume={349},
+  number={6251},
+  pages={aac4716},
+}
+
+@article{camerer2016,
+  title={Evaluating replicability of laboratory experiments in economics},
+  author={Camerer, Colin F. and others},
+  journal={Science},
+  year={2016},
+  volume={351},
+  number={6280},
+  pages={1433--1436},
+}
+
+@article{camerer2018,
+  title={Evaluating the replicability of social science experiments in Nature and Science between 2010 and 2015},
+  author={Camerer, Colin F. and others},
+  journal={Nature Human Behaviour},
+  year={2018},
+  volume={2},
+  pages={637--644},
+}
+
+
+@article{silberzahn2018,
+  title={Many analysts, one dataset: Making transparent how variations in analytical choices affect results},
+  author={Silberzahn, Raphael and others},
+  journal={Advances in Methods and Practices in Psychological Science},
+  year={2018},
+  volume={1},
+  number={3},
+  pages={337--356},
+}
+
+@article{breznau2022,
+  title={Observing many researchers using the same data and hypothesis reveals a hidden universe of uncertainty},
+  author={Breznau, Nate and others},
+  journal={PNAS},
+  year={2022},
+  volume={119},
+  number={44},
+  pages={e2203150119},
+}
+
+@article{sandve2013,
+  title={Ten simple rules for reproducible computational research},
+  author={Sandve, Geir Kjetil and others},
+  journal={PLoS Computational Biology},
+  year={2013},
+  volume={9},
+  number={10},
+  pages={e1003285},
+}
+
+@article{wilson2017,
+  title={Good enough practices in scientific computing},
+  author={Wilson, Greg and others},
+  journal={PLoS Computational Biology},
+  year={2017},
+  volume={13},
+  number={6},
+  pages={e1005510},
+}
+
+@article{stodden2018,
+  title={Enhancing reproducibility for computational methods},
+  author={Stodden, Victoria and Seiler, Jennifer and Ma, Zhaokun},
+  journal={PNAS},
+  year={2018},
+  volume={115},
+  number={11},
+  pages={2561--2570},
+}
+
+@book{turingway2022,
+  title={The Turing Way: A Handbook for Reproducible, Ethical and Collaborative Data Science},
+  author={{The Turing Way Community}},
+  year={2022},
+  note={Zenodo. DOI: 10.5281/zenodo.3233853}
+}
diff --git a/paper/main.pdf b/paper/main.pdf
index f89c4e1..00d387b 100644
Binary files a/paper/main.pdf and b/paper/main.pdf differ
diff --git a/paper/main.tex b/paper/main.tex
index 7bfc37e..a2c818a 100644
--- a/paper/main.tex
+++ b/paper/main.tex
@@ -151,11 +151,26 @@ \section*{Main}\label{sec:introduction}
 strategies against small, unrepresentative sets of opponents. Such practices
 bias conclusions and weaken claims about the relative performance of new
 strategies.
+
+These challenges are not limited to the \IPD{} literature.
+Reproducibility failures have been widely documented across the social sciences
+and economics, with large-scale replication projects revealing that only around
+half of published findings hold up under independent
+scrutiny~\cite{open_science_collaboration2015, camerer2016, camerer2018}. Computational
+research adds further complexity, as analytic flexibility and non-transparent
+workflows can yield highly variable conclusions even from identical 
+data~\cite{silberzahn2018, breznau2022}. Within game theory and related modelling work,
+the challenge of reproducibility intersects with simulation code, algorithmic
+implementation, and data provenance. 
 An important step toward addressing this issue has been the
 \texttt{Axelrod-Python} project~\cite{AxelrodProject}, an open-source Python
 package that provides a comprehensive framework for implementing and testing
 \IPD{} strategies. The library includes a wide variety of strategies from the
-literature, together with detailed documentation and usage examples. By
+literature, together with detailed documentation and usage examples. 
+This project illustrates best practice by providing fully open, tested, and version-controlled
+artifacts, embodying community principles outlined in reproducibility 
+guides~\cite{wilson2014, sandve2013, wilson2017, stodden2018, turingway2022}.
+By
 providing open, executable implementations, \AXL{} makes it possible to test
 strategies under common conditions and compare their performance systematically,
 and it has therefore been used in ongoing research~\cite{Harper2017,
@@ -713,7 +728,7 @@ \section*{Conclusion}\label{sec:discussion}
 the first effort to package and reproduce, according to contemporary best
 practices, code originally written in the 1980s. The archived materials~\cite{knight_2025_17250038} (at
 \url{https://doi.org/10.5281/zenodo.17250038})
-are curated to high standards of reproducible research~\cite{wilson2014} and
+are curated to high standards of reproducible research~\cite{wilson2014, sandve2013, wilson2017, stodden2018, turingway2022} and
 accompanied by a fully automated test suite. All changes to the original code
 were made systematically and transparently, with complete records available at
 (\url{https://github.com/Axelrod-Python/axelrod-fortran}).