I recently read a post about a social experiment from UN Women, claiming to show that actual Google searches are evidence of a high prevalence of sexism. I felt compelled to address the absence of context and serious flaws in this post and in the methodology of this experiment.
- The creators of this experiment and post assume that Google suggests search terms based on the most popular ones entered into the search bar. They fail to consider that other factors also play a role in these suggestions, such as a high level of engagement. This commenter explains it well.
- It's illogical to assume that all people who enter sexist search terms like "Women shouldn't have rights" and "Women should be put in their place," actually harbour these beliefs. If you, as someone opposed to the hate and ignorance, want to analyze sexism and other forms of hate online, one way is to put yourself in the position of the person spewing the hate and search what they would search, in their own language. This way you can get more results straight from the source of the hate. I have done this myself on numerous occasions. It's incredible the number of hate forums and hate blogs that come up or people who share hate anywhere -- and their ilk.
Another issue I have with this post/experiment is that it, as commenter nic points out, "confuses things." It also perpetuates misconceptions about research and the web. This can have serious implications.
I am sure many hateful people do search such hateful, vile things in Google and I think it's fair to assume that this is a factor in UN Women's examples of Google's suggestions, but this experiment does not at all prove the prevalence of this issue.
Of course, hate is a huge problem and I support drawing attention to its prevalence and presenting solutions, but there is no excuse for using an illogical research methodology. There are other ways. If we are going to use the web as a research tool, we must understand how the technology works and employ sound logic, as we should for all research.