Reproducibility in Social Vulnerability Models

Is reproducibility important for social vulnerability models?

It is critical for researchers to be able to replicate and reproduce social vulnerability models to validate original results and ensure their credibility. Replication can assist in assessing if the findings are applicable in other contexts and how broadly findings can be generalized in regard to scientific theory. Replication of studies can additionally assist in fostering a collaborative scientific environment.

Moreover, replication can help interrogate subjective researcher decisions that may have unintentionally biased the results to ensure that the results are credible. For instance, when reproducing Malcomb et al. (2014), there were numerous aspects of the methodology that were subjective; when the study was reproduced with slightly different inputs to subjective decisions, there were substantially different results (see a Reproduction of Malcomb (2014)). In contrast, when a study is able to be replicated and achieve the same or similar results, credibility is added to the model. Model creditability is further validated when a model is able to be replicated in a new location or with a different spatial extent than the original study and still achieve the same results, highlighting the efficacy and generalizability of the model. However, if similar results are not achieved, study replication can assist with the improvement in the conceptualization and execution of the model.

Assessing the generalizability of social and climate vulnerability models is critical since such models are often used to inform development interventions and have a direct impact on disaster risk preparedness of communities. Furthermore, ensuring the reproducibility of social vulnerability and other types of multi-criteria models is fundamental to improving their credibility and assessing the applicability of models across time and space and can have direct impacts on peoples lives.

In regard to my own research, this upcoming winter term I am planning on taking a course on vulcanology and geo-hazards of the Central American volcanic arc. For my independent research project for this course, I am interested in applying a multi-criteria model to assess differences in regional risks to geo-hazards based on soil type, seismic risk, and a number of other factors to assess volcanic risk for communities and to identify where prevention interventions should be prioritized.