Hi all,

I have been using the following method for quite a while to run lab sessions where different numbers of subjects could be used depending on how many show up: https://otree.readthedocs.io/en/latest/rooms.html#what-if-not-all-participants-show-up 

The question I have is, what should we do if more than one subject accidentally enters in the same participant label?  This can be identified if you start the session only after confirming the expected number of participants in the room.  However, once you have identified this, the only solution appears to be to check the URL of each subject, and edit and resubmit the incorrect URL.  This approach has two problems, 1.) it's a bit time consuming to do, 2.) it is likely not possible when using locked down lab computers with a 'kiosk mode' or similar.

Am I missing a simpler solution to this problem?

Why would one want participants in a lab session to enter their label manually?

We use URLs of the form

http://localhost:8000/room/live_demo?participant_label=Cubicle01,
http://localhost:8000/room/live_demo?participant_label=Cubicle02,
http://localhost:8000/room/live_demo?participant_label=Cubicle03,
etc.

Then there is no need to enter a label manually.

And if you open these URLs sequentially according to the cubicle number, the rows in the oTree database will be sorted by the cubicle number.

Usually we open the URLs via shortcuts that we put on participants’ PCs’ desktops that instruct the browser to launch in kiosk mode.

Yes, by the way regarding this:

> if you open these URLs sequentially according to the cubicle number, the rows in the oTree database will be sorted by the cubicle number.

You can use this technique of pre-assigning participant labels in creating_session: https://otree.readthedocs.io/en/latest/rooms.html#pre-assigning-participants-to-labels

Then they are guaranteed to be in the correct order in the admin interface, regardless of which computer you click first.

If I have say, 18 computers set up, but only 12 subjects show up, then I would like to be able to start the experiment using only the 12 computers that subjects are sitting in.  This is why I have them enter their participant label manually, as suggested here: https://otree.readthedocs.io/en/latest/rooms.html#what-if-not-all-participants-show-up .

There are of course alternatives depending on the lab setup.  But in order to get the experiment started quickly, having them enter in their label manually is a simple way to achieve this.

Of course, the trade-off here is the situation I originally described, which would take longer to fix. Given that it doesn't happen often, it might be worth the risk.  I guess there is no easy way to deal with this situation, as clearly I'm doing something that is a bit of an edge case.

Sure, I was also talking about a situation like “24 computers are set up, but only 23 subjects show up.” You asked for “a simpler solution to this problem” than having participants enter their label manually. I would consider opening URLs that already contain the participant labels on the different computers simpler: It is even less time-consuming (during the experiment) than asking participants to enter their labels, and it is less error-prone, as long as you can save the predefined URL on the respective computer.