series: “Section 1: Introduction - do the Design of Experiments”

We are now going to consider experiments to compare two conditions or treatments. The discussion will lead to a review of several statistical concepts. And the engineer aims to evaluate if the addition of a polymer emulsion in a cement mortar affects the tension bond strength of the product. The experimenter has collected 10 observations of the modified formulation and 10 observations of the unmodified one. Looking into the table of results, each observation is called a run, and the differences among the individual runs in the same treatment is called noise or experimental error or simply error. The noise or error arises from an uncontrolled variation that is generally unavoidable. By plotting the experimental data in a dot diagram, we can visualize the central trend of each treatment, with the average strength of then modified formulation being slightly higher than the one of the formulation of the polymer and also the simple variability, meaning how much that the data is spread around the average.

Note: 10 observations were not repeated runs.