Data Visualization & Gender Studies
Modern feminist movements, also known as ‘the fourth wave,’ make use of the digital era by spreading information about feminism and human rights via social media. This process of moving feminist theory out of academia and into the public sphere may sound phenomenal, but can actually be quite harmful towards reaching feminist objectives. Using modern tools such as data visualizations, organizations have been taking data driven approaches to gender equality. This can do more harm than good as data visualizations don’t provide neutral nor objective representations of reality, but rather a series of judgments and selections that influence the ways in which we interpret certain topics. It can also be difficult, as women tend to be categorized with other ‘oppressed’ groups; in turn, leaving a gap in research. A second problem that arises with the use of data visualizations is the simplistic and straight forward appearances causing individuals to forget about or ignore where the information is coming from or how reliable it may be. For that reason, it is important to throughly analyze the rhetorical work of these visualizations. The following blog post will go over three data visualizations that have the same goal of “gathering data and analyzing the gains made for women and girls over the last two decades, as well as the gaps that remain” (No Ceilings, 2016).
The first visualization “Gender Pay Gaps in the US and UK”, has a clear approach to expanding one’s understanding of the current existing pay gaps between genders of different professions. In some ways the visualization can be seen as trying to fulfill an idea, touched upon by Donna Haraway in “Situated Knowledges: The Science Question in Feminism and the Privilege of Partial Perspective/Feminist Studies” (1988), known as “the God trick”. It is the hopes of seeing data from the perspective of no particular individual, thus removing all possible biases. In the case of the “Gender Pay Gaps” visualization, it does it’s job of clearly defining a problem women are facing with regards to equal pay, along with expanding one’s knowledge towards which professions specifically pay women less then men (or vice versa). Additionally, it does not create any noticeably disruptive bias through the colors or layout of the information, refraining from the use of any gender stereotypes with the use of neutral colors such as purple and green. However, it does have several hidden flaws with regards to the reliability of the information. What the bar chart does not reveal says a lot more about what is actually included. For example, no where in the visualization does it actually state the number of individuals used to come up with an average salary per gender. In addition, we do not know how many men are being compared to how many women. For all we know, the information could be completely unreliable. The information that is made visible then becomes these possibly biased averages of salaries which are used to represent individuals of different genders. At first glance, this visualization is simple and shocking. It has the tendency of drawing one’s attention to the female’s salaries, in order to help fight gender inequality and therefore speaking from an epistemological perspective.
The second visualization, “Film Dialogue”, has a more complex way of representing the gender inequality between the number of words men and women are given in films. One positive factor of the analysis was the way in which the data was collected. Using 2,000 films, the individuals collected real data, in a census format. After formatting the data into a visualization is where the results lose their effectiveness as too much information can confuse a viewer. Rather than compiling the information into one visualization, they decided to design three separate charts, each representing the same idea: men being given more dialogue in films then women. Additionally, some of the labels can be misleading, when looking solely at the data visualization and not reading into the topic for more context. One example is the label “100% of words are male”; the label can be highly problematic as it is unclear what a “male word” is referring to and could be easily misunderstood as a word referring to masculinity. In actuality, however, the label is used to represent the number of words given to male characters in the film versus those given to females. Another problem that can be seen in this particular visualization is the colors used to represent the different genders. Rather than deviating from the stereotypical colors associated with gender, the designer chose to stick with the recognizable blue for boys and faded red for girls. By choosing to replace the color pink with red, but also fading the colors out to represent more or less words spoken by each character, the male depiction always stays blue, but the females change to pink. This making the ideological perspective of gender as “situated knowledge” quite obvious. This only means that what is made visible is what we have already been forced to call “general knowledge”, with regards to male supremacy. Which, in turn, does not necessarily add to or help close gaps in gender equality.
The third and final visualization I will be discussing, “Female Entrepreneurs are on the Rise” is a break down of different countries and the percentage of females making large contributions of entrepreneurial activity in those regions. Using an interactive bar chart and different colored key codes, the visualization again, tries to see the data from a god’s eye view. In attempt to remove all biases the visualization makes sure to include all countries. One particular interesting factor is the unique representation of uncertainty for countries in which data was not made available. For example, when I searched Lebanon in the “search countries” engine i was brought to a page that specifically explains that no data is available for this region. The visualization does a good job at educating individuals on information they might not have known before without including any obvious biases towards gender, race or ethnicity within the design. Another interesting factor is the basis they used in categorizing a woman as entrepreneur. By calculating an average number of women who received loans from friends and family rather than from an institution, they come up with a number of female entrepreneurs. Whether or not this information is reliable is unclear, especially with regards to the individuals who got loans from their friends and families. Additionally, by choosing to disregard (or make invisible) the number of men in comparison to that over women around the world, this can be considered a gender bias alone. However, this visualization is probably the most beneficial with regards to “closing gaps” in feminist research as it offers new information from a minimally biased perspective.
After viewing the following visualizations, i think it is safe to say that until we are capable of understanding the power of “inclusion and exclusion” and the consequences that accompany choosing what is to be considered as “general knowledge”, “we must acknowledge data visualization as one more powerful and flawed tool of oppression” (D’Ignazio, 2017).
References:
Anderson, H., & Daniels, M. (2016, April). The Largest Analysis of Film Dialogue by Gender, Ever. Retrieved from https://pudding.cool/2017/03/film-dialogue/
D’Ignazio, C. (2017, January 22). What would feminist data visualization look like? — Catherine D’Ignazio — Medium. Retrieved from https://medium.com/@kanarinka/what-would-feminist-data-visualization-look-like-aa3f8fc7f96c
Female Entrepreneurs are on the Rise. (2013). Retrieved from http://www.noceilings.org/entrepreneurs/
McCandless, D. (2019, January 11). Gender Pay Gap. Retrieved from https://informationisbeautiful.net/visualizations/gender-pay-gap/
