I have found the hour restrictions in this thread to be quite interesting. I just wanted to chime in from the "user" side of the policies that folks have described or posted. I'm a former professor and my undergrad and grad school years were full of late night activity.

I think the reason a lot of the policies have "sketchy" or loose descriptions is because the nature of research is such that you can't shoehorn everything into one box. Few academic laboratory research projects ever finish in the amount of time that you think. There are some processes that you just can't shut down when 5 PM or 8 PM or the appointed Official Hour is reached. Even with good planning and an experiment you've done before, researchers routinely take a lot longer to finish tasks than originally planned - ask any grad student how many times they have had to call/text their significant other that they wouldn't make dinner/date/whatever on time.

I fear that a time that appears set in stone by definition or enforcement language will encourage folks to rush through finishing up, thereby increasing the chances of an accident. And, in my personal experience, most accidents happen when you're cleaning up, so hurrying up at the end when you're "all done" is even more dangerous.

If you do have specified "regular" or "preferred hours" then you also need to come up with policies that reflect the realities of research. It should address what to do if an experiment unexpectedly takes you beyond the normal hours, explain that no new experiments/procedures should be started once you're in the "overtime" period etc. In other words, some performance-oriented language rather than black and white one size fits-all for this one aspect.

Rob Toreki