Collaborative Automation Reliably Remediating Erroneous Conclusion Threats (CARRECT)

Epidemiologists and public health officers face considerable challenges in optimizing the efficacy of their analyses as they are susceptible to bias from missing data, confounding variables, and selecting the appropriate statistical models. Analysts can benefit from assistance in mitigating these issues as possessing the expertise to consider the multitude tests, corrections, and selection rules is extremely burdensome. To embed an expert system in analysts"workflow, Aptima proposes Collaborative Automation Reliably Remediating Erroneous Conclusion Threats (CARRECT) to support statistical bias reduction and improve analyses and decision making. The system will engage the user in a collaborative process in which the technology is transparent to the user. An intuitive data wizard will guide the user through analysis processes by analyzing key features of a given dataset, as well as prompt the user to input additional substantive knowledge to improve the handling of imperfect datasets and the selection of the most appropriate algorithms and models. Reduced bias in the analysis fosters a more veridical assessment for decision makers.