Blog #1: Is There Algorithmic Bias in Edtech Software?

Edtech software can be advantageous in relation to distance learning, however, there are many ethical concerns regarding managing students’ personal data in algorithmic software programs. The main issue I would like to address is a concern brought up in Regan and Jesse’s (2019) article surrounding algorithmic bias when using student tracking. Student tracking involves the tracking of test scores, keystrokes, reading patterns, ability groups, learning levels, racial and gender groups (Regan & Jesse, 2019). The authors argue that the tracking done through big data applications perpetuates a stratified student population (Regan & Jesse, 2019). Since the data is aggregated and cross-analyzed through multiple software systems, there is an ability to re-identify students through their personal information (Regan & Jesse, 2019). Tory Chen suggested inputting my opinion on how we could reduce discrimination in Edtech. I believe that if we clear students’ identity factors and other personal information from the big data applications, we can readily furnish chances for equal opportunity. To make this feasible, should students be taking charge of their own data?

Supporting this argument, in a prior learning experience, I read an article on the University of Texas’ machine-learning system called GRADE, which illustrates how educational software can be inherently biased (Burke, 2020). GRADE is used to predict how likely the graduate admissions committee is to approve an applicant (Burke, 2020). The article explains how this technology has a gender and racial bias embedded into the algorithm (Burke, 2020). The system analyses the keywords in letters of recommendation which can distinguish whether it was written from a female or male student (Burke, 2020). These equity and equality issues in educational software are pertinent issues as there has been a switch to an online/open learning environment, due to the COVID-19 pandemic. The main purpose of open learning, to broaden learning opportunities for all, is inherently against the GRADE system. Open learning involves the use of open educational resources in which their content and digital artifacts are freely used and acceptable to all (“Open Learning,” 2021).

The remaining issue is the subject of the default White male which I learned prior to this class (Pawlowska, 2011). The default White male is a hegemonic framework that influences search engine algorithms and perpetuates racial and sexist biases (Pawlowska, 2011). For example, when “black girls” is typed into Google, Eurocentric conceptualizations of people of colour are embedded in the algorithm, resulting in racist Google images and word searches (see photo below) (Noble, 2018). Congruent with the historical oppression of people of African descent, Google’s algorithm was constructed to portray black students as criminals (Noble, 2018). Therefore, it is important to acknowledge how the default White male is also embedded into Edtech software. Congruently, Regan and Jesse (2019) shed light on this, noting that although the Edtech software reduces educator bias, there is bias in the humans who design the systems, which inevitably results in discrimination against students. As a woman and a person of colour myself, I am deeply concerned with this issue. I then ask you, how can we reduce algorithmic bias?

Photocreds: https://www.theguardian.com/technology/2016/jun/09/three-black-teenagers-anger-as-google-image-search-shows-police-mugshots

References

Regan, P. & Jesse, J. (2019). Ethical Challenges of Edtech, Big Data and Personalized Learning: Twenty-First Century Student Sorting and Tracking. Ethics and Information Technology, 21(3), 167-179. https://doi.org/10.1007/s10676-018-9492-2

Noble, S. (2018, March 26) Google Has a Striking History of Bias Against Black Girls. Time. https://time.com/5209144/google-search-engine-algorithm-bias-racism/

Burke, L. (2020, December 14) U of Texas Will Stop Using Controversial Algorithm to evaluate  Ph.D. Applicants. Inside Higher Ed.  https://www.insidehighered.com/admissions/article/2020/12/14/u-texas-will-stop-using-controversial-algorithm-evaluate-phd

Pawlowska, M. (2011, November 1) Why Is ‘White Male’ the Default?. The Good Men Project. https://goodmenproject.com/featured-content/why-is-white-male-the-default/

Open Learning (2021, September 27). in Wikipedia. https://en.wikipedia.org/w/index.php?title=Open_learning&oldid=1046723524