skip to Main Content
Abstract ID: 463785

Abstract Title: An Analytics Toolbox Facilitating Reproducible Interventional and Observational Science in Aging

Presentation Type: Accept as Symposia

Date: Saturday, November 17, 2018

Time: 8:00 AM to 9:30 AM

Program Overview
Reproducibility and replicability are essential differentiators between unfounded assertion and valid scientific claim. The striking lack of replicability of major research findings in translational research, though broadly attributable to structural biases afflicting the broader research enterprise, has
underscored the need for better design, execution, documentation, and review of quantitative analyses. With the development of technologies permitting the joint production of computational results and research narratives, reproducible analyses, defined as the integrated combination of analytic methods, tools and replicable results – has become plausible in everyday practice. The adoption of open coding and data standards in academic publishing underscores the urgency of engaging these approaches in clinical research practice as a day-to-day standard. To date, few applications have been developed to permit easy adaptation of reusable or modular tools in clinical research. In this symposium we will describe our experience developing a customizable library of reusable and shareable modules for study monitoring, quality control, exploratory data analysis, regulatory reporting, and research dissemination in aging. We focus specifically on the development of platform-independent, web-based, user-friendly, and extensible monitoring and analytic software tools that permit collaboration and version-control among members of a translational research team, including quantitative scientists, clinical and nonclinical investigators, and skilled project staff. We will demonstrate the potential for these systems to be used to answer questions in real-time through the use of interactive graphics and text, and to facilitate quality assurance, exploratory analysis, hypothesis generation, and sensitivity studies by independent parties using ‘synthetic’ simulated data mimicking published results.

Back To Top