DoE module (Experimental design builder). I haven’t found any description/examples in the Statistica help how to analyze a D-optimal / I-optimal custom design prepared in Experimental design builder

[Marek Skowronek]

Let’s assume that the user of Statistica has prepared a custom design, collected data according to this plan and ?.

[Heleen Snelting]

Hi Marek, hope you are well - I'm sending a message on behalf of Amanda Shankle Knowlton who has reviewed your question with a few other colleagues: Greetings from the Statistica engineering team. The intention behind the Experimental Design Builder is to provide assistance in designing the structure of experiments, taking into account all of the various constraints of the research and desired optimization approach. The end result is intended to be a spreadsheet with columns for all factors and their required settings for each experimental run as well as columns where you can put values for your dependent/measurement variables. The next steps would be for the researcher to use this spreadsheet for data collection and then return to one of our other modules for analysis of the data in the spreadsheet.

To summarize a reply from a colleague: GLM, ANOVA, Variance Components, VEPAC can be used to analyze and in some specific cases (e.g. surface, central, Latin squares) there is some special support in the DoE module. We rely on the statistical knowledge of the user to choose the appropriate analysis.

If you have specific guidance as to what would be helpful to include in our documentation here to get our users on the right track, we welcome your suggestions on the ideas portal. This is listed in the top menu of the community, and if you do not use this yet, please find some instructions here. Thank you for your question and your feedback!

[Amanda Shankle]

Hi again Marek,

Our colleague Michael O'Connell also passed along the following comment:

There are a number of books that cover analysis of D-optimal designs eg

https://www.amazon.com/Experimental-Design-Biotechnology-Statistics-Monographs/dp/0824778812

Have a great week!