External validity of social psychological experiments is a concern, but these models are useful

International audienceAbstract We agree that external validity of social psychological experiments is a concern, we disagree these models are useless. Experiments, reconsidered from a situated cognition perspective and non-linearly combined with other methods (qualitative and simulations) allow grasping decision dynamics beyond bias outcomes. Dynamic (vs. discrete) insights regarding these processes are key to understand missing forces and bias in real-world social groups