Let's say that I've been part of a research lab at a U.S. university for awhile but do not plan to continue on for my PhD. What does one typically do with all of their research work, e.g. experimental write-ups, code files and email correspondence with researchers external to the lab? Does one typically initiate some "handover process" and document all of the work carefully enough so that someone else in the lab gets to inherit the work and take over from that point on, or should one keep their own work theirs, without distributing it to the lab that they were a part of?
Ideally, all work would be carefully documented and discussed with co-workers in the group before someone leaves. Often times, though, departures are much more sloppily handled.
However, the research group obviously retains an interest the lab notebooks, data, and other work products created while someone is part of a research group. For data retention practices, I would say that "best practices" would include leaving the originals with the research group, while the person who created the data keeping a copy for their own personal records later down the line.
Considering documentation and informing coworkers about your work, is not different than in the industry. The research you create at work, where you were employed and paid to do the research with the employer's resources, is most often property of the team/department and the company. But usually, you can keep your research notes and some of the data - if it's not critical intellectual property.
Also, the correspondence to external people could be viewed as somewhat public (if they didn't have to sign a NDA) and you could keep it and use it in further correspondence to those externals.
Even during normal research projects, coworkers should know what you do and what experiments etc. are lined up in the near time frame and vice versa. In the industry, it's also very important that you constantly have notes that 1 or 2 coworkers that are familiar with the topic, can use if you're ill or on holiday.
So I would say it's normal to start a "handover process" at least 1 or 2 months before you leave. Even if the present coworkers won't (or can't, because time etc.) take over your research, as a supplement to the notes, they can help the next person who does.
The idea of writing the description like a paper seems a very good example. Maybe you can talk with some of your coworkers about your research and focus the "papers" on the parts where they had the most questions about.
This answer also applies to leaving a lab but staying in academia, just moving to a new lab, institution, etc.
Have a Talk About Future Plans : Is it intended that you will be an author on the projects you are leaving, because they're at the stage where your part was a significant contribution? Have you done enough that you should still be first authors (if you are), or are you willing to give that up for whoever carries things the final distance.
This is also a good time to make sure what you're planning to do and what your PI/boss wants you to do are the same.
This is also also a good time to make sure you know whether or not you can use code/data/etc. in the future at your new position or not.
Leave Your Contact Info : Preferably a relatively stable email address.
User accounts/Electronic Resources : Make sure that any electronic accounts that are tied to you (i.e. GitHub accounts, things where you're the admin/moderator/etc.) are transferred to someone else, and that their username and password combinations are findable.
Documentation : That thing you've always been meaning to do, but never gotten around to doing? Now's the time. Create a data dictionary with clear descriptions of any data you have, what the values mean, etc. Talk it over with whoever is taking over your project to make sure that they understand what you mean. If you have code, make sure it's documented, commented and easy to follow. Clean up any lingering #FIXME type comments, or at least make sure your successor knows they're there.
Contacts : If you've built a relationship with a stakeholder group, another collaborating team, someone in a core facility that helps with your work, etc. it's time to do your best to hand over that relationship - at the very least, make some introductions so that they know who you are.
We have a lab culture of having a lot of digital write ups. We spend a lot of time documenting everything and they look like research papers even if they do not get submitted. Since you seem to not have that:
I would take a bit of time to outline anything that you believe is valuable in the same way you would write an academic paper. The two reasons for this are 1) it transfers knowledge to your laboratory, and 2) it is something that you can show an employer if questions come up about your time in academia.
From anecdotal stories from people who have left the lab, these sort of documents were useful when they were finding an industry job and if they went back to academia in the future.