This thesis deals with the implications of the filter bubble to unbiased and equal information access for users from different countries. Thus, it investigates how personalisation algorithms on search engines might affect access to information. Consequently, this thesis asks: If any differences are found, what are their possible effects? Can a filter algorithm potentially influence users by framing search results differently? Overall, this paper aims to analyse whether the Internet can provide equal and true information, even when using filtering algorithms.
This paper follows constructivism and interpretivism for its epistemology and ontology. The data collection uses an experiment. Two participants that were recruited based on convenience took part in the experiment. They are both female and international Master’s programme students. The participants collected data from search result pages from two different search engines, Google and Brave. Their query was “Ukraine conflict” and they took screenshots of the top three search results on the first search result page for both search engines.
The headlines of the search results did not include special typography or language that is untypical of serious news outlets. However, most search results highlighted Russia’s part as the initiator in the war. One participant saw images related to the query at the top of the page on Google, while the second one did not. There was additional information displayed around the search results in both cases and for both search engines. The participants were presented with differing content, both on Google and on Brave. In some cases, they saw information from the same news sources but their display, as well as the overall composition of the search page, varied. In general, the structure of the Google search result page showed a stark contrast between the first and second participants.
The findings of the analysis show that the participants do not have equal access to information. Despite using the same search query and the same search engines, they were presented with differing information, not only content-wise but also with regard to how this content was framed. It seems like the filter algorithm has some kind of influence on which information users can see, possibly also considering the display of search results.
Keywords: Search engine, filter bubble, Google, Brave, information online, filter algorithms, news online, frame analysis, multimodal analysis