Treatment assignment for one participant group

[Anna Shcherbiak]

Hi Chris and others,

I'm designing a study, where participants are randomly assigned to be either Player 1 or Player 2, and then Player 2 can be assigned different levels of the endowment. So we have the following structure: 

• Player 1
• Player 2
  • High endowment 
  • Mid endowment 
  • Low endowment

It seems like I know how to assign participates to be either Player 1 or Player 2, but not sure how to layer in this endowment component for Player 2. I thought maybe I could combine the factorial treatment approach, but it doesn't seem to be the correct way to think about this: 

def creating_session(subsession: Subsession):
       players = subsession.get_players()
for p in players:
       p.is_employee = p.id_in_group % 2 > 0
       p.is_boss = p.id_in_group % == 0 

import itertools
treatments = itertools.cycle(
        itertools.product([50, 100, 150])
for p.is_employee in subsession.get_players():
      p.endowment = next(treatments)

Thank you so much! Any guidance would mean the world to me!

Anna

[Chris @ oTree]

The for loop needs to be written like this:

for p in subsession.get_players():

    if p.is_employee:
      p.endowment = next(treatments)

By the way, you don't need 2 fields (is_employee and is_boss). Just use one of them, e.g. "if not p.is_boss" means it's an employee.

[Anna Shcherbiak]

Thank you so much, Chris!

If I have multiple rounds in my experiment, should I be worried about participants being assigned a different role in each round or does intertools.cycle take care of that?

Thanks!

Anna

[Chris @ oTree]

see here; https://otree.readthedocs.io/en/latest/treatments.html#treatment-groups-multiple-rounds