Mass psychogenic illness (MPI) is more frequently observed in females than males. The proposed research investigates whether MPI symptoms can be elicited via expectation and modelling, and whether modelling is responsible for the gender effect (as suggested in a gender x modelling interaction observed in pilot data). In a 2x2x2x2 (inhale x modelling x participant gender x model gender) half of the participants will be told that they have been assigned to the experimental condition and will be asked to inhale a placebo substance (actually plain ambient air), presented as a suspected environmental toxin. Half of the participants will observe a confederate inhale a substance and display symptoms, and half will observe a confederate who does not inhale the substance or display symptoms. Confederates supposedly assigned to the experimental condition, will model the specified symptoms behaviourally and via their verbal ratings. Gender of participant and confederate will be balanced across cells. Verbal ratings of four specified symptoms and four additional symptoms will be elicited at 10-minute intervals over the course of one hour. At the conclusion of the study, all participants will be fully debriefed.

The observation units were 120 university students, 60 males and 60 females who were asked to inhale an alleged environmental toxin and to fill out a form just before inhalation, and six times afterwards, every ten minutes. On the form they had to report the experienced severity of each of eight symptoms, four of which were suggested to them as being connected to the inhaled substance. Participants inhaled from an inhaler used to train patients with asthma, which did not contain any active ingredients. Participants we run in pairs, each pair made of one participant and one confederate. In half of the cases confederates were of the same gender of the participant. Confederates were instructed and trained to simulate (verbally and behaviorally) the experience of suggested symtoms increasing in severity over the first 30 minutes, and then declining in severity after that. Participants were run in the Behavioral Medicine Unit Lab at the University of Hull.