This dissertation examines the influence of normality perceptions and shifts of normality perceptions on individuals attitudes and behaviors. A social psychological concept of normality was developed and stated in a model of normality and a model of shifts of normality, based on diverse research lines (such as norm theory, research on the status quo, naive realism, shared reality and social norms research). The underlying mechanisms of the models were examined in seven quantitative studies in heterogenous study fields. The model of normality illustrates the process of emergence and development of stimuli and events as normal; a key aspect of the model being the deductions drawn by individuals from normality to normativity and objective reality. The model of shifts of normality contributes to explaining how the scope of normality can gradually move and be extended towards the extreme. In Study 1 and 2, the emergence, shift, and influence of normality in social settings were examined. Study 1 was embedded in the simulation of a UN General Assembly, and Study 2 was set in the context of the German federal election 2013. In Study 3, shifts of perceptions of groups normality were examined. The study was implemented via an integration-fostering mentorship program and the ingroup projection model was used as a theoretical framework. Studies 4 to 7 form the core element of the empirical section, examining the effect of social norms on reality perceptions in the domain of prejudice. The studies show that whichever prejudice is congruent to the group norms is perceived as true, objective and real, and hence not identified as prejudice. Nonnormative prejudices, in contrast, are perceived to be faulty and resulting from biased perceptions. These results question the widespread concept of prejudice as reflecting bias and underline the necessity to examine the basic processes. All seven studies are discussed with regard to the two models and their practical implications.