# Do we need an [evaluation] tag?

We have a new tag, :

For questions regarding the evaluation of experiments, including metrics.

Is this something we should keep? It seems to me that probably most of statistics could be construed to be about the evaluation of experiments. I can't help but wonder if this will really help organize the information on the site, or if it is inevitable that it will be used inconsistently over time and lead to more confusion.

If we shouldn't keep it, what should we do? Should it be deleted, or made a synonym of something (what)?

• As I understand it, "evaluation" for many people is about evaluating the success or failure of a programme/program [non-computing sense], policy, law, or intervention in society, medicine, the environment, etc. That often has a statistical flavour. (Indeed, it is hard to think of good evaluations that don't.) – Nick Cox Mar 1 '18 at 18:59

Yeah the description was made in a hurry sorry about that. Here was my thinking: I created a question about aggregating metrics, and the closest tag was model-evaluationwhich is close but not exactly right.
• We do have a metric tag. It isn't clear to me that your question needs any of that. Ultimately, you have a [good] question about aggregating over means in a particular circumstance. My 1st thought would be that the tags [mean], [aggregation], & perhaps [geometric-mean] would be appropriate & sufficient. – gung Mar 1 '18 at 19:17