2.2: Takeaway 2- Students use Defensive Practices to Protect their Privacy

Often out of frustration, some students had taken to “gaming the algos,” a combination of practical strategies to protect their privacy that we came to call defensive practices. These practices included using free apps and browser extensions to counter tracking along with a few retaliatory tactics they had invented or learned from friends to intentionally “confuse algorithms.” While some students were adamant that their strategies were effective, many more were unsure about the net effects their actions were having.

Most commonly, students were running ad blockers or had regularly cleared their browsers of cookies, two old-school tactics for protecting their privacy. A few said they used DuckDuckGo as a search tool or Firefox as a browser to protect their privacy, countering consolidation of personal information by Google, and trying to avoid being put into the filter bubble of personalized search results. Others said they ran Virtual Private Networks (VPNs) to shield their browsing activity. Still others created multiple accounts on platforms like YouTube, Google, and Instagram so they could avoid having all of their internet activities tied to a single identity. As one student explained, “I got irritated because it sees who I am, and then it funnels me into getting certain content.”

For some students being reduced to a profile based on the profligate collection of data created a problem of “context collapse” — they lost control of the ability to craft their public identity when different facets of their personality became jumbled in a single account. They solved this problem by deliberately curating multiple “selves,” or different accounts that could reflect their varied interests. This defensive practice was intriguing because it gave them control of their self-representation as they interacted with the digital communities that formed around each of those interests, engaging in a kind of digital code switching.

International travel emerged as a surprising catalyst for learning about the role of algorithms in personalizing content. While traveling abroad, students said they were forced to learn new tools and strategies like using VPNs to get around firewalls. In some cases, they described seeing firsthand how international events could be misrepresented in the mainstream Western news. It also brought home the reality that geographic location was a major factor in personalizing search results and other content.

But not all students in our sessions said they were taking action to protect their identity from prying algorithms. These were often the optimistic students who said algorithms did more good than harm. A trend that emerged was that those who discussed using one defensive practice in the focus group ended up talking about using several. The students who were most aggressive in their strategies were often STEM majors, or lived with roommates who were STEM majors and were a source of useful tactics. While some students were clearly more knowledgeable than others about how to counter online tracking attempts, there was consensus throughout the focus groups that using sites like Google, Facebook, YouTube, or Instagram had made them much more vulnerable to tracking.

No matter their major, it was clear that most students wanted to learn how to fight back against online surveillance. In fact, when the topic came up in the focus groups, it was not unusual to see students jot down a few notes about apps, like AdBlock, or server subscriptions, like NordVPN, that fellow focus participants said they were using. We found this tendency to learn from peers had likely developed earlier during adolescence as they evaded parental oversight through digitally mediated peer relationships,⁶⁹ and as they shared hints about navigating around the school’s imposed barriers. As one student explained, “Everyone just kind of shared this information around the entire school, like, ‘Oh, get the VPN and it’ll hack everything and you can access your Instagram during the middle of the day.’”

Comments like this one suggest that many students, depending on their socioeconomic status and high school resources, may enter college already knowing far more about navigating the internet than many might suspect. And yet, many students in this study were largely unaware that systems they used in their courses, like Canvas, the popular learning management system (LMS), had the potential to gather, aggregate and sell personal information. Once it was discussed in our sessions they were indignant to learn it might be happening. Though university administrations often claim that such surveillance is valuable for student retention and assessment, some educators are concerned about “learning analytics” programs that fail to offer students opportunities to provide informed consent or opt out.⁷⁰

Breaking the news bubble

Targeted ads were clearly an annoyance, but “filter bubbles”⁷¹ were even worse. Most students knew that algorithms showed them only part of the picture, especially on social media platforms like Instagram and YouTube. In turn, these personalization effects, these “silos,” or “echo chambers,” had trapped them in a narrow space of confirming and reinforcing beliefs. One student expressed frustration by saying, “Just because I watched a video on whatever subject doesn’t mean that I don’t want to see the opposing side, I want to be educated, I don’t want to be in my box with one opinion.” Another described concern about the silo effect of social media platforms:

I often worry about getting everything because I usually ‘like’ pages that I agree with or ‘follow’ pages that I agree with, but that actually worries me because I feel like it will put me in this bubble where I don’t have any exposure to different opinions, so I made a conscious effort to not unfollow pages I dislike.

Many students, like this one, tried to evade news traps, so they could see “other sides of news” and escape the perils of personalization. Often they used what academics call lateral reading, ⁷² a strategy for seeking out sources that would present different approaches to the same topic. For example, students might compare how the same news story was covered by one source, such as The New York Times, and by more conservative sources, like Breitbart News or Fox News. One student explained reading across content producers to get the complete story:

I don’t trust one source — I purposely follow the other side, I guess it’s weird, but I want to see how they’re thinking, too, because it gives me some insight on how they’re forming this article or opinion.

Though many agreed that lateral reading was the best strategy to ensure a balanced view and check for accuracy, others were clearly exasperated with the amount of work that they had to put into this process:

I see something on Facebook and I try to get the true information. I’m like, ‘Okay, how do I know which websites have not just posted a bunch of bull, how many websites do I need to scroll through to find what I’m looking for?’

A few students said they were taught in college to trace scientific research reported in the news to its source. Still others, critical of the incompleteness and inaccuracy of breaking news coverage, used local news sites to get “the real story,” with some waiting days — even weeks — to read a breaking news account so they could find out what had really happened.⁷³ As one student argued, it was the responsibility of news consumers to dig deeper and go beyond social media news feeds:

People shouldn’t just rely on social media to get news, they should hold themselves to a higher standard, and if they see something, then they should look deeper and try to find out if it’s true or false. So, I think that people should just make their own decisions based on their own research, instead of just looking at social media posts and just agreeing with it, without really thinking too much.

Regardless of how they got news and information, most students were especially concerned about how algorithms tailored information to the individual in ways that reinforce beliefs, biases, and prejudices. For some, this is how algorithms went beyond simply being code and preyed on human nature. Still other students acknowledged they were willing to be categorized so they could have their thoughts, opinions, and news preferences confirmed.⁷⁴

As one student pointed out, this is nothing new: “We surround ourselves with people we agree with and then we have conversations with them that reinforce our own ideas, so in a lot of ways, this is now just getting a third party to do all of that for us.”

References

69. danah boyd (2014), It’s complicated: The social lives of networked teens, Yale University Press.
70. Some students on campuses outside our sample, however, are becoming aware of data-gathering by institutions and are increasingly advocating resistance. See Zak Vescera (27 March 2019), “Canvas is tracking your data. What is UBC doing with it?” The Ubyssey, www. ubyssey.ca/features/double-edged-sword/; Tom Nash (November 2019), “Freedom of Information Request: Ram attend opt out,” Muckrock, https://www.muckrock.com/foi/virgini...-optout-83757/; Sharon Slade and Paul Prinsloo (2014), “Student perspectives on the use of their data: Between intrusion, surveillance and care,” European Distance and E-Learning Network, 18(1), 291–300, https://www.eurodl.org/ index.php?p=special&sp=articles&inum=6&abstract=672&article=679. Also see Kyle M.L. Jones (2 July 2019), “Learning analytics and higher education: a proposed model for establishing informed consent mechanisms to promote student privacy and autonomy,” International Journal of Educational Technology in Higher Education 16(24) DOI: https://doi.org/10.1186/s41239-019-0155-0; Kyle M.L. Jones and Dorothea Salo (April 2018), “Learning analytics and the academic library: Professional ethics commitments at a crossroads,” College & Research Libraries 79(3), 304-323, DOI: https://doi.org/10.5860/crl.79.3.304; Deborah West, Ann Luzeckyj, Danny Toohey, Jessica Vanderlelie, and Bill Searle (1 January 2020), “Do academics and university administrators really know better? The ethics of positioning student perspectives in learning analytics,” Australasian Journal of Educational Technology, DOI: https://doi.org/10.14742/ajet.4653
71. Eli Pariser popularized this concept in 2011 in The filter bubble: How the new personalized web is changing what we read and how we think (Penguin). However, critics have since called the significance of filter bubbles into question. See for example Mario Haim, Andreas Graefe, and Hans-Bernd Brosius (2017), “Burst of the filter bubble? Effects of personalization on the diversity of Google News,” Digital Journalism 6(3) DOI: https://doi.org/10.1080/21670811.2017.1338145 and Frederik J. Zuiderveen, Damien Trilling, Judith Möller, Balázs Bodó, Claes H. de Vreese, and Natali Helberger (31 March 2016), “Should we worry about filter bubbles?” Internet Policy Review 5(1), https://papers.ssrn.com/sol3/papers. cfm?abstract_id=2758126
72. For a discussion of lateral reading, see op. cit. Sam Wineburg and Sarah McGrew 2017, 38-40.
73. For more about growing issues with the veracity of breaking news, see op. cit., Head, et al. 2019, “Across the great divide,” p. 28.
74. Jeff Passe, Corey Drake, and Linda Mayger (July 2018), “Homophily, echo chambers, & selective exposure in social networks: What should civic educators do?” The Journal of Social Studies Research 42(3), DOI: https://doi.org/10.1016/j.jssr.2017.08.001

Contributors and Attributions

• Template:ContribInfoLiteracyAgeofAlgorithmPIL