The list of topics for this site on https://physics.stackexchange.com/help/on-topic includes this entry:

Experimental designs and results

Example: What is needed to claim the discovery of the Higgs boson ?

I think what is meant here is questions related to the design of experiments themselves, is that right?

If so, the existing wording gave me a somewhat different and I think unintended impression, something along the lines of: "designs that are uncertain". (Designs of what? Not clear...)

Should the wording be changed to something like "Design of Experiments"? I thought that expression was a more typical one in the scientific community and to me it seems a more precise description.

I do see that Wikipedia includes "experimental design" as an alternative to "design of experiments", so I guess people in relevant fields may be familiar with both terms. But using the clearest possible term for a wide often non-professional audience seems better for this site, IMO.

  • 1
    $\begingroup$ A complete answer to this question might include documentation about how, in the culture of working physicists, an "experimental" result or effect or process is generally being contrasted with a "theoretical" counterpart. A physicist talking about an "experimental observation" is generally expressing confidence, rather than doubt. This is somewhat different from the most common use of "experimental" by people from outside of the sciences. $\endgroup$ – rob Jan 26 at 16:36

I'm taking my mod hat off and speaking as a regular user here (and one who does not do theoretical physics work so I can't address the suggestion in rob's comment either, unfortunately).

From the perspective of somebody in the fluid dynamics field, my immediate reaction is to say that these are two different things, or perhaps that "experimental design" is the laymen terminology for a much broader concept at least.

"Design of Experiments" is a statistical technique to ensure that the conditions tested encompass the response of something to variations in its input. Wikipedia has a decent explanation of the process. In some sense, it is very similar to uncertainty quantification, although it happens before the system is tested rather than after.

"Experimental design," to me, is about how to set up and/or design an experiment to study the physical effect in question. It could be about setting up physical hardware, or how to set up an optics train, or how to calibrate equipment properly as part of an experiment. It can also include how to establish which conditions are required, i.e. it can include the statistical tool "Design of Experiments" as part of it.

So, I would recommend we don't change the name because I see them as different things. "Design of Experiments" is a specific statistical tool, whereas "Experimental Design" can be much broader.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .