A Short Guide to Experimental Design and Analysis for Engineers

An experiment is a test under controlled conditions to investigate the validity of a hypothesis. Experimentation is the basis for creating new knowledge – determining whether some factor causes an effect. Well-designed experiments provide a systematic approach to observe relations among variables while ruling out alternative explanations. This report has examined a number of experimental designs including the simple experiment, matched-pairs, repeatedmeasures and the single group design. The relevant statistical techniques are also discussed to help identify key quantitative methods for data processing and analysis. While not intending to replace any textbook, this guide summarises the resources and some worked examples that have helped the authors gain a basic understanding of design, measurement and statistical analysis to support military experiments. RELEASE LIMITATION Approved for public release