Big Data Techniques in Science Education and What Story Google Trends Tells Us About Science?

Resumo

The intention of this work is to provide a quick overview of what Big Data is and present a few examples of techniques through which it can contribute to Science Education. Google offers the Google Trends (GT) free analysis tool that allows users to sort through several years of Google search queries from around the world to get a graphical plotting showing the popularity of chosen search terms both over region and time. According to the time, region, and frequency of search, three kinds of data are evaluated in terms of compatibility with a sort of “correlation analysis”. A few techniques of extracting meaning from them are exemplified through geographical searches for ‘Solar Eclipse’ in USA and through temporal searches of the term ‘research’ in the period 2013-2017. In addition, and as the main study, an experiment was conducted to replicate with Big Data and GT Taşdere, Özsevgeç, and Turkmen’s survey on the Nature of Science (NoS). To that end, the same nine concepts they selected were searched in GT. Two-way correlation analysis was performed on these words, and those pairs with a Pearson Correlation of 0.8 and higher were used to build a conceptual network. Three main levels emerge in our hierarchical conceptual network and, as a result of this structuring, a storytelling can be built: Science is seen, in a more publicly understandable level, as associated to ‘laws’, followed by a less-visible level of research being associated to ‘building theories’, and then, in a even lesser understanding level, the scientists doing experiments to test hypotheses, which are confirmed or not by observation – an image of scientists’ work shaped in a large degree by popular media.