IT Best Practices

Search Engines Reinforce Social Stereotypes

When we type an inquiry into Google or Bing, do we sit back and think deeply about the result? Or rather, do we hope we find the quickest, easiest answer and keep it moving? The latter is more likely. Without much thought, most fail to realize how search engines are reinforcing social stereotypes. Jahna Otterbacher discusses for Harvard Business Review separate studies that show how this is so and the way it is accomplished.

Skewed Results

In April of 2016, MBA student Bonnie Kamona noted that a Google image search for “unprofessional hair for work” showed pictures dominated by that of colored women. On the flip side, predominantly images of white women came up when searching for “professional hair.” Kamona’s study is part of the growing body of research on search engine bias.

Another study performed by Matthew Kay, Cynthia Matuszek, and Sean Munson had to deal with professions in Google image search and how gender was distributed. When “doctor” was searched, more images of men were present, and for “nurse” there were more women, even though this is not statistically correct. This showed how search engines also exaggerate gender stereotypes concerning professions.

Otterbacher and her colleagues from the University of Sheffield’s Information School have gone even further to distinguish through research how search engines perpetuate more generalized gender stereotypes. Research of “person perception” has determined two categories by which people judge others upon interaction: “warmth” and “agency”:

Warm people are seen as having pro-social intentions and do not threaten us, although warm traits can be either positive or negative; for example, adjectives such as “emotional,” “kind,” “insecure,” or “understanding” would all count as denoting “warmth.” Agentic people are those who have the competency to carry out their goals and desires and are ascribed traits such as “intelligent,” “rational” and “assertive.”

Bing was used for this experiment, and after searching 68 character traits and analyzing the first thousand images for each, the conclusion came that warm images are primarily depicted by women, while men show up in images of agentic character traits. What can we do about such biases in search engines? Well, not a lot at the moment, other than try to be mindful of it.

You can access the original article here:

Show More

Leave a Reply