Creativity in data analysis through confrontation with anomalous data

Creativity is perceived to be one of the core competences to succeed in the modern world. It is connected to art, music, dancing, etc., but there is just little insight into the role of creativity in science. As data analysis plays a major role in science, creative thinking has to be used to form theories out of observations. This study uses models-of-data to represent the mental model of the students. Within models-of-data different observations and previous knowledge are linked through different types of links, e.g. causal links. The goal of this study was to examine the creative aspect of data evaluation in presence of anomalous data. The students were confronted with two experiments which contradicted each other with regard to the underlying theory. Afterwards the students were asked to decide about their favored theory and reasoning about their choice in their lab reports. The model-of-data, which was reconstructed from the lab reports of the students, was used to extract creative aspects in the mental modelling process. Furthermore, a pre-post-questionnaire on epistemological beliefs of the students was conducted. Students mostly acquired an unscientific view on epistemology in school. The question was whether the confrontation with anomalous data and the triggering of creative modelling processes have any influence on this. The results show that most students did not take anomalous data into account in their modelling processes. They did either not recognise or just ignore the data. Just a few students worked on their theory because of the new, contradictory data. The students working on their theory scored higher in the post-questionnaire, so a positive effect of creativity on scientific epistemology can be assumed. Thus, creativity obviously gets a part of science and should be taught in school science in order to make students aware of their own creative potential.