I'm wonder from your experiences what people seem to forget or can't get their head around about statistical tests and models.
In my, yet short, experience I see people totally bypass randomization, independence and the correct interpretation of power and confidence.
I work in a regulated industry (pharmaceutical industry) and I'm thinking that we could use R in our analysis with GxP regulated data. I was asked wheater we could lock the R instalation and packages so that: 1) users cannot install unauthorized packages; 2) unauthorized users cannot access it and run scripts; and 3) we can restore the exact instalation of R and packages in case we need in future.
I think we could do this but not sure how or if anyone has achieved something close to it any where so I turn to reddit for help.
Edit: I know of initiatives like Rinpharma, R Validation Hub, pharmaverse, Posit paid solutions and the paper from R Foundation about R meeting FDA's CFR Part 11 but I have not been able to take much advantages from those
I'm looking for anyone that has tried R+RStudio in a pharmaceutical Quality Control lab and is willing to share the experience. I'm talking things like Analytical Method Validation, Stability Stydies, managing Control Charts and so on. In our company, we, the stats team, face several questions when introducing the ideia and for some of those we have hard time finding an argument/solution. Yes, I'm aware of initiatives like rinpharma, we just have not been able to build from those alone.
Also, in case anyone wonders, at the current moment we don't have any sort of statistical software approved for use. We do any analysis that need to be done in R (ggplot2 for plots, quarto for reports, and on, and on...) in such non formal way.