If this is not the case, participants must attend the training session Fundamental Tools in Statistics. This training course is held the day before « Introduction to the Design of Experiments ».
Statistics is the art of gathering, presenting, analyzing and utilizing data in order to facilitate decision making and problem solving. Statistics play an important role in all steps of the experimental process and too often is the use of statistics confined to the post-experimental phase, namely the analysis of results. Statistical involvement at pre and post stages of the experiment actually facilitates the research process, while ensuring the reliability and the precision of the results as well as maximizing the power of the statistical tests. This 2-day training provides an overview of Design of Experiments (DOE) and the issues involved in implementing a design. The purpose and applications of the DOE are reviewed. The different sources of error and variability as well as the tools available to minimize this variability – repetition, randomization, blocking and controls – are discussed. Experimental designs, from simple to complex designs, are reviewed -comparison of two samples (independent or paired), factorial designs for more than two samples – along with the corresponding statistical tool – Student t-test, Analysis of Variance (ANOVA) F-test. The notion of breaking down the components of variability when performing an ANOVA is emphasized. Moreover, the notion of an interaction between two factors and how to test for one will also be explained in detail. The principle of multiple comparisons will also be covered, illustrated by the Least Significant Difference method. Following this training session, the design and implementation of studies with several experimental variables will be facilitated by the use of powerful and efficient statistical tools. Consequently, the analysis of the data collected will provide faster, more reliable and accurate results.
