NAO robot as experimenter: Social cues emitter and neutralizer to bring new results in experimental psychology

摘要

The aim of this study is to highlight the impact of social factors on the outcome of well-known experimental situations, as the exchange paradigm of Knetsch, used to measure the endowment effect. This bias in decision making brings individuals to assign a widely greater value to an object when they own it. Using a humanoid robot NAO taking the place of the experimenter and left alone with the participant permits to standardize the social factors and the pragmatics engaged within any participant-experimenter interaction. This feature allowed us to validate the hypothesis that the endowment effect could be produced by pragmatical factors, as politeness rules, inherent to all exchange situations between humans. As a continuation of our previous work presented for IDT 2016, we refined our choice of programming methods, keeping a “Wizard of Oz” method for any time the robot must appear as a social entity and applying a completely preprogrammed behavior when all pragmatic factors related to the presence of an experimenter should be annihilated. These methods were applying to the exchange paradigm. We based here on the same paradigm to measure the endowment effect, the exchange paradigm of Knetsch, and using a NAO robot to conduct the experiments. But in contrast with our previous works, a dissociation of programming technics was introduced: a wizard of Oz and fully pre-programmed behavior. After finding that results obtained do not differ regarding the programming technic chosen from previous experiments, we set up a new pilot study to find out if a specific nonverbal clue emitted by the experimenter and varying within a single fixed method (fully preprogrammed behavior) would be sufficient to produce an endowment effect. For this purpose, the first results obtained by only varying the voice intonations of the robot show that vocal features can constitute an important factor for a human individual to activate the application of social standards towards a robot, bringing him to produce an endowment effect.