Post: Reproducability and Replicability

Based on the readings, I think that reproducibility and replicability are important in understanding geography. These tools mediate institutional and systemic biases in research. This is key for science to operate as communal. R&R mediates uncertainty in research and helps test increasingly complex computational research methods. I believe black box AI, ML, and DL pose important challenges that are beginning to be addressed in trust research. However, I believe there are there are fundamental questions of access (eg. who has access to understanding scientific methods or publishing scientific work?) that are not addressed by R&R. What does it mean for a scientific community to look/act/think/write/sound a certain way? Moreover, R&R relies on narratives of linear scientific progress and objective truth which must be considered in their institutional and political contexts. To address this, I suggest further study of participatory action research methods that engage a wider array of epistemologies and methodologies and are accountable to the communities that produce and are impacted by scientific study. Survey findings of low engagement with R&R and contradictory definitions/uses of reproducibility and replicability present challenges in validating scientific conclusions and point to the arbitrary, institutional dependence of scientific research as a whole. In particular this presents challenges to interdisciplinary and mixed-method research where discrepancies in scientific standards vary across fields.