The assumption is more that people could form/use them to different degrees. Then based on their behaviour when the environment is misaligned we can calculate how much they rely on allocentric cues. We can compare this "weight" across conditions/experiments

Of the first environment yes

Thanks! I'll check this out.

We tried something like that in exp 4 with our baseline. They see the real door, put on the headset and see nothing, turn and point to the real door. They could do this accurately. After seeing the visual environment they then pointed to the visual door. So it did not seem to affect anything.

Exactly yes! If you haven't read it yet, I think you would really enjoy the General Discussion section of the paper :)

If you like to do that you should stay tuned for our follow up :D

It sure does! But it is a bit nuanced. In the case of vr, you might want to keep track of the previous real world environment (so you don't walk into a wall). Instead people seem to just rely on the current enviro. We suggest that maybe you can hold multiple allo but maybe only 1 ego representation.