| dc.description.abstract | The aim of the current study was to test whether initial biases about social groups would be maintained during an extensive sampling period. Specifically, the study examined whether initial biases regarding the gender-science stereotype, that is, the bias that science is associated with men over women, were upheld or attenuated in different environments. The current research assumes stereotypes to occur due to associations between attributes and social groups and it was expected that an initial pseudocontingency-inference (PC) between gender and success or failure in solving math problems would be upheld in an environment where participants predominantly experienced success, while the same PC-inferences would be attenuated in an environment where participants experienced mostly failure. Results show that an initial bias, regarding gender and math ability, can be induced. However, contrary to the expectations, during and after the sampling phase, the initial bias was attenuated in both environments. Moreover, no difference between the pre- and the post-measure was found. Alternative explanations explore the process of subtyping and the complexity of the social world. Recommendations for future research emphasize the importance of unknown mediators or moderators and an improvement of the current study design. Practical implications regard the notion to show variations to the initial gender-science stereotype, such that women are associated with the ability to perform sciences too, in order to overcome this damaging stereotype. | |
