Refer to 'project' folder as 'experiment' folder?

[JoeZiminski]

In the docs for here and datashuttle I think in the general the root folder is referred to as the 'project' folder. Maybe this is a little ambigious, because really it is expected that it will be 1 'project' per experiment. I think it is unlikely people will merge across projects, though maybe this is not true. In general systems neuroscience experiments are so big that an experiment is basically a project. But, for example a short pilot followed by main experiment, I would (I think) imagine the pilot would be split out into a separate project. I wonder what you think?

[niksirbi]

I would favour keeping the current name. That's what BIDS uses, and besides, the definition of "experiment" can be equally ambiguous and will differ across fields/labs.

Instead, we could perhaps add a note emphasising that the users can choose what constitutes a "project" based on pragmatic needs (and maybe provide some sensible use cases / examples).

[adamltyson]

I like project. Naively I think of:

• Project - the whole thing, including multiple modalities
• Experiment - something smaller. Maybe a single modality.

[JoeZiminski]

Okay this sounds good I agree, that's a nice idea to add a note, maybe if people raise questions about this on the course next week we can add this. Cheers!

[adamltyson]

Is it worth starting a glossary (as discussed yesterday)?

[niksirbi]

BIDs has a definitions section as well as an extensive glossary of schema objects.

I think what we need right now is the "definitions" part.

The animal electrophysiology BEP also defines some basic terms here.