In a first study, the impact of expression of opinion and salience of arguments on participants' argument repertoire, opinion quality, and attitudes towards nanotechnology was tested in a 2x2 experimental design. Expression of opinion was revealed to have an impact on the argument repertoire but was not sufficient for formation of high-quality opinions. In contrary, asking for an overall judgement only seems to trigger top-down processes of opinion formation, that is, opinions and attitudes are formed that are highly dependent from prior attitudes and beliefs (confirmation bias, belief bias). Salience of arguments, however, was shown to be important for the formation of well-founded opinions and attitudes that are independent from prior attitudes. However, the Ss still showed a myside bias in their essays on their personal opinion, that is, they were indeed enabled to generate a valid rationale to support their personal opinion but failed to integrate counterarguments and arguments to refute these counterarguments. This ability, however, is considered as major indicator for good informal reasoning and critical thinking. Thus, study 2 aims at specifying the potential of discussion-based installations to reduce the myside bias in judgement and thereby to further enhance critical thinking and reflective judgement.

61 participants explored an exhibition about nanotechnology and stated their personal opinion at a discussion terminal. Following opinion expression, three levels of feedback about others' judgement were varied (agree, disagree, no information control). The results showed that participants who encountered disagreement by means of feedback about others' judgement after their opinion expression experienced this feedback negatively (they were disturbed, astonished and felt less comfortable than participants who did receive feedback about an agreeing majority or no feedback at all). These participants were also likely to change their opinion towards the majority opinion to reduce dissonance. Disagreement was shown to reduce the myside bias in the participants' argument repertoire, that is, Ss who encountered disagreement recalled as much othersided arguments as mysided arguments, whereas Ss of the control and the agreement condition provided more supportive arguments than counterarguments to their position. Furthermore, disagreement led to higher scores in reflective judgement: Ss of the disagreement condition were able to state their personal opinion, to give a valid rationale, and to integrate and elaborate on counterarguments in their statements about nanotechnology. Quality of counterargument generation and rebuttal construction was also higher for Ss of the disagreement condition indicating that these participants deliberately elaborated on the reasons why other people disagreed and were thereby enabled to construct valid rebuttals of others' counterarguments.