Eye Gaze Diversion and Dissociation in External and Internal Shame: A script-driven procedure

The current study examined Gilbert's (1998) concept of internal and external shame, and the degree to which eye gaze diversion is associated with the activation of shame generally, or is more isolated to when either external (i.e., perceived negative judgements from others) or internal (when shame-evoking evaluations come from the self) shame is experienced. The study also examined experiences of dissociation to ascertain whether there is a relationship between shame and dissociation. It was hypothesised that if shame is associated with internal and external observations, more gaze diversion will occur when participants are looking at themselves in the mirror or at the experimenter than when looking at a blank board, during the shame induction. Secondly, if eye gaze diversion is more strongly related to external shame, participants will look away from the experimenter more, during the shame induction. Thirdly, if eye gaze diversion is more strongly related to internal shame, more gaze diversion will be evident when they view themselves in the mirror during the shame induction. It is also thought that experiences of dissociation will increase after the shame induction. Psychology students from the University of Canterbury (n = 78) completed four measures assessing trait and state shame and dissociation, and listened to audio clips of shame-inducing and neutral scenarios while either viewing themselves in a mirror (internal shame), looking at the researcher (external shame) or looking at a blank board (control). Eye gaze diversion was recorded across scripts and conditions. Although initial analyses did not reach significance, simple effects analyses do indicate that eye gaze diversion is more associated with external shame than internal shame or shame more generally. Furthermore, state dissociation was significantly higher following shame inductions, when compared to the neutral induction. Taken together, results indicate support for hypotheses two and four and do not support hypotheses one and three