How to choose the appropriate design for an experiment. Hello, everyone. I wrote a little summary about how to choose the appropriate design for an experiment. I hope that you can help me with some corrections. I would thank you. How to choose the appropriate design. First, you must begin by selecting an approprate model. This model must summarize the relationship between response variable and explanatories variables. Examples of models are first order model and second order model. We use first order model when we consider that the relationship is summarized by a flat plane. We use second order model when we consider that the relationship is summarized by a surface with curvature. Also, the selected model must have a appropiate number of parameters. If it has too few parameters, then it may not capture all the features. If it has too many paramenter, then it may reduce the precision. There are two common objectives when we are looking for an appropiate desing: • To be able to estime the parameter of the model • To predict the response on any location in the design space. Second, we need to assess the quality of model precision. We asses the quality of precision by D-optimality and A-optimality. D-optimality corresponds to finding small variances for model parameters without collecting any data.