Our goal in this SBIR effort is to produce a web-based suite of tools and services to help researchers design methodologically sound experiments for improved research outcomes and reproducibility of results. DesignAssist provides a promising approach that connects animal researchers to experimental design expertise, delivering a cost-effective solution, alleviating the need for in-house design expertise (an expensive proposition) and improving the reliability of research findings. Our research and development goal in this proposed effort is to develop a conduit between animal researchers and design support, providing user-friendly web-based tool for researchers to (a) develop experimental study goals and translate them into appropriate statistical structure, (b) create experimental plans for randomization and blinding, (c) examine sample size, detectable effect size, and statistical power, (d) undertake data analysis with insights on outlier detection, inclusion/exclusion/stopping criteria, (e) obtain expert advice on all of the above through an on-line chat interface, and (f) obtain a detailed report of experimental design recommendations and their justifications.
Our goal in this SBIR effort is to produce a web-based suite of tools and services to help researchers design methodologically sound experiments with reproducible results. Experiment reproducibility has become a crucial problem in biomedical research, and improving reliability can have a significant financial impact for the government and industry.