The current fiscal climate demands now, more than ever, that test and evaluation(T&E) provide relevant and credible characterization of system capabilities andshortfalls across all relevant operational conditions as efficiently as possible. Indetermining the answer to the question, “How much testing is enough?” it isimperative that we use a scientifically defensible methodology. Design ofExperiments (DOE) has a proven track record in Operational Test andEvaluation (OT&E) of not only quantifying how much testing is enough, but alsowhere in the operational space the test points should be placed. Over the last fewyears, the T&E community has made great strides in the application of DOE toOT&E, but there is still work to be done in ensuring that the scientificcommunity’s full toolset is utilized. In particular, many test programs have yet tocapitalize on the power of the test design when conducting the data analysis.Employing empirical statistical models (e.g., regression techniques, analysis ofvariance (ANOVA)) allows us to maximize the information from every data point,resulting in defensible analyses that provide crucial information about systemperformance that decision-makers and warfighters need to know. DOT&E willcontinue to work to ensure the highest technical caliber in every DOT&Eevaluation, and that Test and Evaluation Master Plans (TEMPs) are adequate tosupport these robust evaluations. As we improve in our use of these test designsand analysis methods, we need to ensure these practices are institutionalizedacross the entire T&E community and applied across all phases of DoD testing
Suggested Citation
Freeman, Laura, and V. Bram Lillard. “Taking the Next Step: Improving the Science of Test in DoD T and E.” The ITEA Journal of Test and Evaluation 35, no. 1 (March 2014). https://apps.dtic.mil/sti/citations/trecms/AD1123777.