<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[CUInfoScience - Medium]]></title>
        <description><![CDATA[The Information Science department at CU Boulder is powered by interdisciplinary research and teaching that allows us not only to imagine what today’s technology makes possible, but to invent what society will do with technology next. - Medium]]></description>
        <link>https://medium.com/cuinfoscience?source=rss----e0e808b9deda---4</link>
        <image>
            <url>https://cdn-images-1.medium.com/proxy/1*TGH72Nnw24QL3iV9IOm4VA.png</url>
            <title>CUInfoScience - Medium</title>
            <link>https://medium.com/cuinfoscience?source=rss----e0e808b9deda---4</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Wed, 18 Nov 2026 23:18:36 GMT</lastBuildDate>
        <atom:link href="https://medium.com/feed/cuinfoscience" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[An Exploration of the Twitter to Mastodon Migration]]></title>
            <link>https://medium.com/cuinfoscience/an-exploration-of-the-twitter-to-mastodon-migration-21c15c4336f2?source=rss----e0e808b9deda---4</link>
            <guid isPermaLink="false">https://medium.com/p/21c15c4336f2</guid>
            <category><![CDATA[twitter-x]]></category>
            <category><![CDATA[twitter]]></category>
            <category><![CDATA[online-community]]></category>
            <category><![CDATA[migration]]></category>
            <category><![CDATA[mastodon]]></category>
            <dc:creator><![CDATA[Matthew N. Nicholson]]></dc:creator>
            <pubDate>Thu, 05 Oct 2023 21:04:41 GMT</pubDate>
            <atom:updated>2024-12-04T13:22:01.478Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="A cartoon Mastodon with a knapsack over his shoulder is greeted by a group of other cartoon Mastodons holding a “Welcome!” sign" src="https://cdn-images-1.medium.com/max/1024/1*Y3ImuN7_JzId8x6SWXpHXA.png" /><figcaption>Welcome to Mastodon (via <a href="https://mastodon.social/about">https://mastodon.social/about</a>)</figcaption></figure><p><em>This post was updated on December 4, 2024 to reflect additional collaborators and support for this project.</em></p><h3>What we did and why</h3><p>On the eve of Elon Musk’s acquisition of Twitter in October 2022, he declared <a href="https://twitter.com/elonmusk/status/1585841080431321088?lang=en">“the bird is freed.”</a> In the name of creating <a href="https://variety.com/2022/digital/news/elon-musk-twitter-advertisers-open-letter-1235415748/">“a common digital town square” for the “future of civilization,”</a> Musk’s subsequent actions, like firing the <a href="https://www.washingtonpost.com/wellness/2022/11/18/disability-support-twitter-harassment/">accessibility engineering</a> and <a href="https://www.wired.com/story/twitter-ethical-ai-team/">ethical AI</a> teams, a <a href="https://www.washingtonpost.com/technology/2022/11/14/twitter-fake-eli-lilly/">botched rollout of a new verification system</a>, and the <a href="https://www.vice.com/en/article/n7zm9q/elon-musk-twitter-nazis-white-supremacy">reinstatement of numerous white supremacists</a> prompted hundreds of thousands of users to change their use of Twitter over the following weeks. Many former Twitter users have since adopted alternative social media platforms, like Mastodon.</p><p>While this user migration between competing products is not new or unique to life online, the scale of these movements beginning in late 2022 marked a key moment in the history of online social behavior. Moving quickly to capture empirical evidence of this ephemeral phenomenon, our team of researchers at the University of Colorado Boulder developed a survey instrument to capture (1) motivations and challenges of this platform migration, and (2) perceptions of emerging alternative social media platforms. We chose to focus our attention towards Mastodon in this survey since it was the most commonly mentioned Twitter alternative in Q4 2022 and Q1 2023. However, demonstrating the fluidity of this situation, not one, but two additional micro-blogging alternatives (Bluesky and Threads) have since rapidly gained popularity. We were able to add a section on Bluesky to our survey prior to launching, while Threads had not launched until after we had concluded our data collection.</p><p>Our survey ran from May 25 to June 14, 2023. In the three weeks that our survey was open, we recruited via posts to <a href="https://twitter.com/cfiesler/status/1661807702266499072">Twitter</a>, <a href="https://hci.social/@cfiesler/110430819925642186">Mastodon</a>, and <a href="https://bsky.app/profile/cfiesler.bsky.social/post/3jwlaqa6jm22x">Bluesky</a> and received well over 2,000 total responses. After removing responses that provided gibberish responses to free response questions or were flagged as probable spam by Qualtrics, we were left with 1,343 responses. The rest of this blog post reports numbers with respect to this sample.</p><p>Our survey respondents report that Musk’s takeover of Twitter (and accompanying changes to the platform) drove many users’ platform migration decisions. Despite common narratives that folks only want a safer version of Twitter, the technical and cultural features of Mastodon itself were also a draw for many users. Despite its attractiveness, the particulars of joining Mastodon and discovering relevant people and content was difficult for a large portion of our respondents. Finally, we report widespread differences in perception of both Twitter and Mastodon, with Mastodon being seen as safer, better for protecting privacy, and more in line with personal values, even among respondents that currently use Twitter more than Mastodon.</p><h3>What we found</h3><h4>Elon Musk’s acquisition and policy changes were the primary catalyst for respondents’ changing use of Twitter and Mastodon.</h4><p>The majority (70.0%) of respondents to our survey reported that they use Twitter less than they did 12 months prior, with 40.0% reporting that they no longer use Twitter at all. While this decrease in use could be attributed to a number of factors, respondents were particularly alarmed by Elon Musk — 90.0% of respondents cited a dislike of the new ownership as one of the reasons that they now use Twitter less, a sizable difference between the second most selected option (71.8%) that reported dislike of policy changes.</p><h4>Twitter pushed. Mastodon pulled.</h4><p>While Musk’s takeover pushed users away from Twitter, Mastodon itself was a draw for many respondents in our survey. Nearly half (41.8%) of those with a Mastodon account joined prior to 2022; however, more than two thirds (68.1%) did not use their accounts regularly until 2022 or later. Though the most popular reason for joining Mastodon was to find an alternative to Twitter (70.9%), the next most popular reasons were liking the idea of federated social media (59.0%), and that Mastodon feels less toxic than other social media (58.1%).</p><p>Despite popular narratives that federation <a href="https://www.washingtonpost.com/opinions/2023/01/17/twitter-mastodon-replacement-social-media/">“solves problems that most users don’t care about,”</a> it was a salient factor for a majority of our respondents. Among those who reported “looking for an alternative to Twitter” as a reason that they joined Mastodon, an even higher percentage cited liking the idea of federated social media (70.3%). While these results do not generalize to all of Mastodon or the Fediverse, federation is not an afterthought for our respondents. Additionally, the comparative lack of ads on Mastodon was a moderately popular response (46.6%), and is another key point of differentiation from Twitter. Only 23.1% of respondents identified the presence of friends already on Mastodon as a reason to join.</p><h4>Joining and using Mastodon isn’t always easy.</h4><p>For a number of our survey respondents, joining Mastodon was more difficult in practice than in theory. Searching for an instance to join was the most commonly selected challenge (39.6%), and the next most popular initial challenge was actually knowing how to join a specific instance (21.1%). While these struggles are far from universal, it is worth exploring what could be done differently to better support newcomers when joining Mastodon.</p><p>User experience challenges continued for many respondents once on Mastodon. About half of respondents reported that finding friends or community members from other platforms was a challenge (48.9%). Difficulty finding accounts with similar interests (34.7%) and finding content about news or current events (34.6%) were common challenges as well. Each of these challenges are related to “discoverability” of people and content. While Mastodon instances often eschew friend recommendations and algorithmically ordered timelines to limit the outrage that comes with optimizing for user engagement, users often find these features useful in some respects (or at least notice when they are gone).</p><h4>Different places with different values.</h4><p>Our survey respondents view Twitter and Mastodon quite differently. Across all of our respondents we see large differences in value perception between these two micro-blogging platforms.</p><p>On average, our respondents were more likely to agree that Mastodon protects privacy, is safe, and is good for society compared to Twitter. Interestingly, these results hold <strong>even when filtering on respondents that currently use Twitter more than Mastodon</strong>, which might suggest that our respondents find value in this platform pluralism, beyond the utility of Mastodon alone. Among this same group, however, it seems that there are affordances of Twitter that are less readily translatable to the federated landscape as it stands. Respondents who use Twitter more than Mastodon tend to depend on Twitter and report that Twitter helps to keep them informed. One possible explanation for these differences is that many localities use Twitter to communicate emergency alerts, as interconnectedness of the platform has been <a href="https://www.pewresearch.org/short-reads/2013/10/28/twitter-served-as-a-lifeline-of-information-during-hurricane-sandy/">vital for on the ground information sharing during crises</a>. In the face of Twitter’s deteriorating reliability, <a href="https://fediversereport.com/dutch-government-officially-launches-mastodon-server/">certain governments are also considering Mastodon as an alternative</a>.</p><figure><img alt="A plot of statements down the y-axis that read. These are sorted in descending order of difference between Mastodon and Twitter. The x-axis ranges from Strongly Disagree to Strongly Agree from left to right.  All statements are marked significantly different at p = 0.05, and there is a footnote that reads that the valence of hate speech and misinformation has been swapped for clarity." src="https://cdn-images-1.medium.com/max/1024/1*in5aVnYqaJWxQ4ieXsRqNQ.png" /><figcaption>Differences in average perception of Twitter and Mastodon across all respondents, sorted by the size of the difference between Mastodon and Twitter. On average, our respondents were more likely to agree that Mastodon protects privacy, is safe, and is good for society compared to Twitter.</figcaption></figure><h3>Explaining our findings</h3><p>Two prominent theories offer a way to explain these findings — the push-pull-mooring framework and network effects.</p><p>The push-pull-mooring framework describes migration as being informed by a combination of factors that make people want to leave a space (pushes), factors that make people want to join a new space (pulls), and factors that make people want to stay (mooring). To our respondents, Musk’s takeover and accompanying policy changes “pushed” users away from Twitter. At the same time, aspects of Mastodon like federation, less toxicity, and fewer ads “pulled” users towards Mastodon.</p><p>“Network effects” occur when the value and motivation of adopting a service scales with the number of other adopters. In the context of a social networking site, there are more connections to be made and more potentially interesting or relevant content when there are more people on a site. This was notably not among the most popular reasons that our respondents joined Mastodon, but it was the second most cited reason for continued use of Twitter (29.4%, “it’s a habit” was the most selected at 32.9%). To many in our sample, these network effects helped “moor” users to Twitter, but did not “pull” users to Mastodon.</p><p>While our sample may be skewed towards those who felt strongly about changes to Twitter, these results imply that the migration from Twitter to Mastodon was driven less by network effects of following and re-building one’s social graph and instead was governed by other “push” factors. This could be evidence that the network effect-driven “growth hacking” strategies that dominated venture capital-backed platforms through the 2010s may be bumping up against users’ concerns about usability and safety.</p><h3>Conclusion and what’s next</h3><p>The social media landscape has changed profoundly since November 2022. Our survey provides early evidence into the motivations and challenges for users who are migrating between platforms. In the context of the migration(s) between Twitter and Mastodon in late 2022 and early 2023, we find that Elon Musk’s acquisition and accompanying changes were a common “push” factor driving users away from Twitter, but there are also features of Mastodon that “pulled” users toward it. Contrary to popular narratives that “users don’t care about federation,” a majority of our respondents named Mastodon’s federated affordances as distinguishing features that “pulled” them towards adoption. Perspectives about Twitter versus Mastodon differ significantly with substantially greater positive attributions made to Mastodon, even among active users of Twitter.</p><p>We thank our respondents for their time, attention and thoughtfulness in filling out our survey, and we look forward to sharing additional findings from this work as it progresses.</p><h3>Further readings</h3><p>Lots of thoughtful commentary on this ongoing migration from Twitter/X to Mastodon has been written. Check out this interview from Justin Hendrix of Tech Policy Press and Dr. Johnathan Flowers <a href="https://techpolicy.press/the-whiteness-of-mastodon/">on the obstacles to joining Mastodon, especially for people of color</a>, this piece from Drs. Chris Gilliard and Kishonna Gray on that gives more <a href="https://www.wired.com/story/black-twitter-social-media/">insights on the costs of leaving Twitter for Black folks in particular</a>, and also check out this piece from tech writer and editor Erin Kissane that gives <a href="https://erinkissane.com/mastodon-is-easy-and-fun-except-when-it-isnt">perspectives from other folks who switched social networking sites on “how should Mastodon be.”</a></p><p>For a brief sample of our theoretical foundations, check out some of the following resources:</p><ul><li>Fiesler, C., &amp; Dym, B. (2020). Moving Across Lands: Online Platform Migration in Fandom Communities. Proceedings of the ACM on Human-Computer Interaction, 4(CSCW1), 1–25. <a href="https://doi.org/10.1145/3392847">https://doi.org/10.1145/3392847</a></li><li>Hsieh, J.-K., Hsieh, Y.-C., Chiu, H.-C., &amp; Feng, Y.-C. (2012). Post-adoption switching behavior for online service substitutes: A perspective of the push–pull–mooring framework. Computers in Human Behavior, 28(5), 1912–1920. <a href="https://doi.org/10.1016/j.chb.2012.05.010">https://doi.org/10.1016/j.chb.2012.05.010</a></li><li>Hou, Avus; Shang, Rong-An; Huang, Chin-Chun; and Wu, Kuo-Long, “The Effects of Push-Pull-Mooring on the Switching Model for Social Network Sites Migration” (2014). PACIS 2014 Proceedings. 64. <a href="http://aisel.aisnet.org/pacis2014/64">http://aisel.aisnet.org/pacis2014/64</a></li></ul><h3>Collaborators and Acknowledgements</h3><p>This work has been a collaboration between Matthew N. Nicholson, Brian C. Keegan, Joanna Mendy, Blakeley Hoffman Payne, and Casey Fiesler.</p><p>This work is supported by NSF Awards #<a href="https://www.nsf.gov/awardsearch/showAward?AWD_ID=2309485">2309485</a> and <a href="https://www.nsf.gov/awardsearch/showAward?AWD_ID=2240196">#2240196</a>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=21c15c4336f2" width="1" height="1" alt=""><hr><p><a href="https://medium.com/cuinfoscience/an-exploration-of-the-twitter-to-mastodon-migration-21c15c4336f2">An Exploration of the Twitter to Mastodon Migration</a> was originally published in <a href="https://medium.com/cuinfoscience">CUInfoScience</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Black Mirror Writers Room: The Case (and Caution) for Ethical Speculation in CS Education]]></title>
            <link>https://medium.com/cuinfoscience/the-black-mirror-writers-room-the-case-and-caution-for-ethical-speculation-in-cs-education-5c81d05d2c67?source=rss----e0e808b9deda---4</link>
            <guid isPermaLink="false">https://medium.com/p/5c81d05d2c67</guid>
            <category><![CDATA[black-mirror]]></category>
            <category><![CDATA[computer-science]]></category>
            <category><![CDATA[tech-ethics]]></category>
            <category><![CDATA[cs-education]]></category>
            <category><![CDATA[science-fiction]]></category>
            <dc:creator><![CDATA[Casey Fiesler]]></dc:creator>
            <pubDate>Wed, 06 Apr 2022 18:22:26 GMT</pubDate>
            <atom:updated>2023-04-01T13:04:15.989Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ne55FPIMbw3M23_7WJq4RQ.png" /><figcaption>A student group’s Black Mirror episode pitch. [Credit: Shamika Klassen and Ella Sarder, used with permission]</figcaption></figure><p>When the AI conference NeurIPS began requiring papers to include a <a href="https://arxiv.org/pdf/2105.04760.pdf">broader impacts statement</a> that includes “ethical aspects and future societal consequences,” there was a fair amount of pushback, and one social media comment I distinctly remember was along the lines of: How do they expect us to <em>see the future</em>?</p><p>Isaac Asimov once said, <strong>“Science fiction writers foresee the inevitable, and though problems and catastrophes may be inevitable, solutions are not.”</strong> It is true that science fiction fiction writers are very good at <em>seeing the future</em>, particularly when it comes to things like the unintended consequences of technology. However, it also doesn’t tend to be the job of science fiction writers to come up with solutions — that is a task that often falls to technologists. So something I’ve been thinking a lot about lately is: <strong>How can we help technologists think more like science fiction writers?</strong></p><p>Now to be clear, when I say “science fiction,” I don’t mean that technologists should be designing towards sentient AI or the robot wars. I mean the ability to consider future consequences at all — including (very) near future. Speculation is a skill that can be practiced, and science fiction can be an engaging way to do that. But: <strong>if technologists can get excited about imagining some harm caused by technology someone <em>might</em> create in 50 years, why can’t they do the same thing with what they create tomorrow?</strong></p><p>In 2018 <a href="https://cfiesler.medium.com/the-black-mirror-writers-room-teaching-technology-ethics-through-speculation-f1a9e2deccf4">I wrote about two teaching exercises</a> that I’d been using in my tech ethics and policy class, including the Black Mirror Writers Room, which I subsequently turned into <a href="https://tinyurl.com/blackmirrorwritersroom">a slide deck that others could use to run the exercise</a>. Over the next few years I heard from many instructors who had used this exercise with students from high school to PhD and in many different kinds of computing-related classes. Hoping to learn more about the benefits and challenges of this exercise, as well as using ethical speculation in the computing classroom generally, PhD student Shamika Klassen interviewed 12 of these instructors, and our paper based on this study was just published at the <a href="https://sigcse2022.sigcse.org/">ACM SIGCSE 2022</a> computer science education conference: <a href="https://cmci.colorado.edu/~cafi5706/SIGCSE2022_BlackMirror.pdf">“Run Wild a Little With Your Imagination”: Ethical Speculation in Computing Education with Black Mirror</a>.</p><p>Overall, the instructors we interviewed were extremely positive about the benefits of this teaching exercise for their classrooms. They spoke about the positive responses students had to the <strong>opportunity to be creative and collaborative, </strong>and that it was a fun way to make ethical topics engaging and salient. As one instructor said:</p><blockquote>“It was a really great way for them to express their creativity, plus their knowledge of our domain space that we had been talking about… It just shows this synthesis of the knowledge that they’ve built up over the quarter about these types of technologies in this context.”</blockquote><p>For the past few years my students and I have also been <a href="https://www.internetruleslab.com/responsible-computing">conducting research about pedagogical strategies for teaching ethics and related topics in computing</a>. Our interviews also revealed how this exercise could work well as a way to introduce ethical thinking in technical computing classes. One instructor also noted how students were <strong>“shocked to be asked to think about these issues and be creative,”</strong> and another elaborated:</p><blockquote>“They know how to fix your computers, they understand how to code, they understand everything that there is to know about software, but they don’t get any ethics most of them and they just really liked it. They really liked just thinking that way.”</blockquote><p>However, despite the overall positive sentiment towards this pedagogical strategy, instructors also pointed out a number of challenges, including the lack of scaffolding around the ethics instruction for them, and the lack of cultural context (around Black Mirror or science fiction generally) for some students. Based on this feedback, students and I are working on providing some a<strong>dditional resources for specific domain areas, as well as ethical speculation exercises that do not rely on Black Mirror.</strong> For example, as part of a <a href="https://www.colorado.edu/cmci/2021/05/26/ai-cybersecurity-coursework-be-developed-nsf-funding-cu-boulder">research project on AI+cybersecurity education</a>, PhD student Samantha Dalal and I created<a href="https://docs.google.com/presentation/d/1zIa6V4CjtTra6s_QWPdRDC1boSNlRZUayuGO2ePlV7c/edit#slide=id.p"> this exercise (please feel free to copy and adapt this slide deck for yourself!) </a>for a computer vision course that introduces types of threats and then has students write future news headlines.</p><p>Finally, an important insight from our interviews confirmed a caution around ethical speculation as an approach that had already begun to take shape for me. We asked instructors about justice-related themes in the Black Mirror episodes students came up with, and some instructors noticed such themes — most often around class inequality or gender discrimination, but none of the examples instructors recalled involved race. We also asked instructors about the demographics of the classes in which they used the activity, and most noted that there were few BIPOC students present in these classes. This led us to consider the <strong>limitation of how different types of diversity in the classroom might result in different kinds of stories,</strong> and one of our interview participants had this critical insight about the limitations of speculation:</p><blockquote>“I tend to be cautious about speculation in part because people tend to lean on their own experiences pretty heavily in speculation, and don’t, unless, they’re very carefully prompted, consider broader context. I worked as a user experience designer. . . and I would see use of like personas or other kinds of vaguely speculative exercises where you’re asked to imagine the ideal user or imagine ideal kind of use cases. And almost universally people would be like, the user is a middle-class white person and here’s what they’re going to do. . . you are just reifying all the stereotypes. And so I think that’s the danger with speculation done well: it’s a great opportunity to disrupt the stereotypes, but I think it takes a lot of work to do that right. . . In more diverse classes, you tend to get a lot of different viewpoints in a lot of discussions, although there is a tendency for people who are in kind of minority populations to suppress their viewpoints to some degree in order to not stick out in the majority. So that’s always a risk and that’s a discussion we have on the first day of class in my class.”</blockquote><p><strong>In other words: It is inevitably easier for people to imagine futures for and harms to people like them.</strong> So given our diversity problems in the computer science classroom, there is a real risk of a student-created season of Black Mirror that speculates about technological harms to middle class white people. And in fact I realized my own blind spot here. The Black Mirror episodes I was using as examples were indeed all about middle to upper class white people:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*RWjGE6n4bQ0BFucJgBBE2A.png" /><figcaption>The episodes “Be Right Back,” “Nosedive, “Arkangel,” “The Entire History of You” all have white protagonists.</figcaption></figure><p>We are working on continuing to create more scaffolding for this and related speculation exercises to bring in e.g. principles of <a href="https://design-justice.pubpub.org/">design justice</a> for learning from and involving people who are different than you.<strong> </strong>In the short term I have added slides to the current exercise to point out this issue along with the reminder that technological harms disproportionately impact marginalized groups. But in general in using speculation in the classroom, <strong>we caution instructors to ensure that they avoid reinforcing stereotypes and consider ways to encourage students to think broadly about injustice.</strong></p><p>I have written before about the <a href="https://cfiesler.medium.com/what-our-tech-ethics-crisis-says-about-the-state-of-computer-science-education-a6a5544e1da6">importance of touching on ethics throughout computing curriculum</a> in addition to students learning in standalone classes from experts, and I hope that this exercise helps computing students understand that thinking through potential consequences and harms is <em>part of</em> what it means to design and build technology. I am also thinking about the potential power of this approach beyond the classroom, and my <a href="https://www.colorado.edu/cmci/2021/04/09/intervention-how-ethical-speculation-could-prevent-future-tech-trouble">recently funded NSF CAREER grant</a> has as an outcome an “ethical speculation toolkit” that can help practitioners reduce <a href="https://www.wired.com/story/opinion-ethical-tech-starts-with-addressing-ethical-debt/">ethical debt</a> in the design process. And again, the point here is not to encourage students or practitioners to actually think like they’re writing for Black Mirror — but instead to pull back and use the same skills to think about <em>tomorrow</em> just as easily as fifty years from now.</p><p>Finally, I want to point out that beyond commentary about the substance of the exercise itself, a number of our participants also offered (unprompted) how helpful it was to have come across a teaching exercise that was openly shared, and could be “plug and play” or adapted to fit a particular instructor’s needs. The majority of participants said that they heard about the exercise from Twitter, either as originally shared by me or by someone else who had used it. Multiple participants mentioned how useful it was to have instructions, examples, and a template, both in terms of running the exercise and in students using the slide template to create artifacts. I hope that this is encouraging for instructors who might consider sharing teaching materials so that others can use and build upon them!</p><p>Thank you to the Responsible Computer Science Challenge for partially funding this research, to NSF for supporting my work in this space moving forward, and especially to the instructors who generously gave us their time and insights. You can read Shamika’s and my full paper here:</p><p>Klassen, Shamika, and Casey Fiesler. <a href="https://cmci.colorado.edu/~cafi5706/SIGCSE2022_BlackMirror.pdf">“Run Wild a Little With Your Imagination”: Ethical Speculation in Computing Education with Black Mirror</a>. In <em>Proceedings of the 53rd ACM Technical Symposium on Computer Science Education</em>, pp. 836–842. 2022.</p><p><a href="https://www.internetruleslab.com/black-mirror-writers-room">At this link you will also find more resources on the Black Mirror Writers Room</a>, and for more on my work related to ethics in computing education, check out my lab’s <a href="https://www.internetruleslab.com/responsible-computing">webpage of writings and resources</a> or the video below:</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FLYRKqnLeIDo%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DLYRKqnLeIDo&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FLYRKqnLeIDo%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/5db6e479b0d4154e24b691aa3c02acb5/href">https://medium.com/media/5db6e479b0d4154e24b691aa3c02acb5/href</a></iframe><p>And for more examples of the amazing things that students in my classes have come up with over the years, some TikToks. :)</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.tiktok.com%2Fembed%2Fv2%2F6953305075607407877&amp;display_name=tiktok&amp;url=https%3A%2F%2Fwww.tiktok.com%2F%40professorcasey%2Fvideo%2F6953305075607407877&amp;image=https%3A%2F%2Fp16-sign-va.tiktokcdn.com%2Ftos-maliva-p-0068%2Ff8ea266bf3e047119d677629c1a9f2a6_1618942502%7Etplv-tiktok-play.jpeg%3Fx-expires%3D1647014400%26x-signature%3DgzbjqdMocYKTTuHaSlDjiMGScuE%253D&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=tiktok" width="340" height="700" frameborder="0" scrolling="no"><a href="https://medium.com/media/83f5156d539d5e9ce3250a73d825f44e/href">https://medium.com/media/83f5156d539d5e9ce3250a73d825f44e/href</a></iframe><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.tiktok.com%2Fembed%2Fv2%2F6960337725958458630&amp;display_name=tiktok&amp;url=https%3A%2F%2Fwww.tiktok.com%2F%40professorcasey%2Fvideo%2F6960337725958458630&amp;image=https%3A%2F%2Fp16-sign-va.tiktokcdn.com%2Ftos-maliva-p-0068%2Fb3586e0b3abd42b1a6e22991537543b9_1620579919%7Etplv-tiktok-play.jpeg%3Fx-expires%3D1647014400%26x-signature%3D0V%252FMLtqJ%252B0d9PZje5H2XuXUBdpk%253D&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=tiktok" width="340" height="700" frameborder="0" scrolling="no"><a href="https://medium.com/media/b8dc0750133a12749f5d40bb5bcb5304/href">https://medium.com/media/b8dc0750133a12749f5d40bb5bcb5304/href</a></iframe><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=5c81d05d2c67" width="1" height="1" alt=""><hr><p><a href="https://medium.com/cuinfoscience/the-black-mirror-writers-room-the-case-and-caution-for-ethical-speculation-in-cs-education-5c81d05d2c67">The Black Mirror Writers Room: The Case (and Caution) for Ethical Speculation in CS Education</a> was originally published in <a href="https://medium.com/cuinfoscience">CUInfoScience</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Wait, how do these algorithms work?]]></title>
            <link>https://medium.com/cuinfoscience/wait-how-do-these-algorithms-work-b2cabdb3108f?source=rss----e0e808b9deda---4</link>
            <guid isPermaLink="false">https://medium.com/p/b2cabdb3108f</guid>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[research]]></category>
            <category><![CDATA[algorithms]]></category>
            <category><![CDATA[ethics]]></category>
            <dc:creator><![CDATA[Jessie J. Smith]]></dc:creator>
            <pubDate>Wed, 08 Sep 2021 15:26:55 GMT</pubDate>
            <atom:updated>2021-09-08T15:26:55.654Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*oPgQUR_XGaaMvOM4" /><figcaption>Photo by <a href="https://unsplash.com/@markuswinkler?utm_source=medium&amp;utm_medium=referral">Markus Winkler</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><h4>Improving transparency in fairness-aware recommendations</h4><p>Have you ever scrolled through social media and noticed advertisements that show you exactly what you’ve been looking for? Have you noticed how platforms like TikTok and Instagram often show you content that keeps you scrolling on the platform for <em>much </em>longer than you were anticipating? Does it sometimes seem like Spotify just ‘gets you’ in a way that Pandora’s music recommendations never did?</p><p>If you’ve ever experienced any of this, then you have interacted with a <strong>Recommender System. </strong>In fact, most people who participate in the digital world have interacted with recommender systems — whether they realize it or not.</p><p>In this article, I will be explaining some of the basics of how these artificially intelligent recommender systems work, as well as how they can be designed to treat their users better.</p><p>This is a part of ongoing research that I am conducting as a part of my PhD, where my goals are (1) to improve Artificial Intelligence (AI) and Machine Learning (ML) systems to make them more ethical; and (2) to help users better understand how these systems work. Ideally, this research will also help improve the general public’s understanding of AI/ML to help empower everyone in their digital lives.</p><p><em>This article was written as a public-facing summary of an academic paper that was published in The 29th Conference on User Modeling, Adaptation and Personalization (UMAP 2021). </em><a href="https://dl.acm.org/doi/10.1145/3450613.3456835"><em>The full paper and a presentation by the authors can be found on the proceedings website.</em></a></p><h3>How do Recommendation Systems Work?</h3><p>In this section I’ll explain some of the key components of recommender systems as they exist today, and how ethical concepts like “fairness” alter traditional recommendation algorithms.</p><h4>Personalized algorithms know us better than we know ourselves</h4><p>One of the key ingredients of a recommender system is that it collects information about you in order to recommend relevant content to you.</p><p>For example: on a platform like YouTube, the videos that you have watched, the channels you have subscribed to, the likes, dislikes, comments, and your general activity on the internet can all be used in combination by the recommendation algorithm to try to understand what types of videos you like to watch. Once the algorithm has a good idea of what you like, it can recommend videos to you that you might be interested in.</p><p>This is called <strong>personalization</strong>.</p><p>Personalization is a great way to keep users interested or engaged in a platform, because it transforms every individual feed into something that is catered to your unique, personal profile.</p><p>Platforms like Spotify, Netflix, Hulu, YouTube, Amazon, Facebook, Instagram, and TikTok can thank a large amount of their success to personalized recommender systems that keep their unique users entertained and coming back for more.</p><h4>Recommendation algorithms trying to be more “fair”</h4><p>In a personalized recommender system, one of main objectives is to attempt to <em>accurately</em> represent users’ interests. In other words, the objective is <strong>accuracy</strong>. This means that the recommendation algorithm is trained and tested to try to be as accurate as it can be.</p><p>However, there are other objectives besides accuracy that we can try to optimize for in recommendation — one of which is “fairness.” I put “fairness” in quotes here, because there are dozens of definitions of the term, and <a href="https://www.youtube.com/watch?v=jIXIuYdnyyk">just as many algorithmic implementations of these definitions</a>. This is helpful to keep in mind while exploring algorithmic fairness, because there is no ‘one size fits all’ here . <a href="https://arxiv.org/pdf/2003.06461.pdf">What is ‘fair’ for some may be ‘unfair’ for others</a>.</p><p>It has recently become more common for AI platforms to optimize for fairness, because of different types of users feeling like they have been treated unfairly by an AI platform. In recommender systems alone, many scandals have left users feeling unheard or mistreated by these platforms.</p><p>In 2016, many conservative Facebook users complained that FB’s news recommendations were perpetuating liberal / left-wing ideals and suppressing more conservative political views [<a href="https://www.nytimes.com/2016/05/10/technology/conservatives-accuse-facebook-of-political-bias.html">read the story</a>].</p><p>In 2018, many parents on YouTube complained that the recommendation algorithm was sending their children down conspiracy theory rabbit holes [<a href="https://www.businessinsider.com/youtube-suggested-conspiracy-videos-to-children-using-its-kids-app-2018-3">read the story</a>].</p><p>In 2020, many LGBTQ+ content creators on TikTok complained that the recommendation algorithm was suppressing their content because of their identity [<a href="https://dl.acm.org/doi/abs/10.1145/3432951">read the research</a>].</p><p>It has become obvious that recommendation algorithms that only optimize for <em>accuracy objectives </em>can lead to some unintended consequences, which is why new research about incorporating ethical objectives like <em>fairness </em>is becoming necessary.</p><p>In this research done by myself and my colleagues, we focused on one specific definition and implementation of fairness as it relates to recommendation: <strong>provider fairness.</strong></p><h4>What is a “provider” and what is provider fairness?</h4><p>Recommender systems are <em>multistakeholder systems, </em>which means that there are sometimes a variety of individuals or groups that can benefit from the delivery of recommendations on a platform.</p><p>For example: on a platform like Spotify, the recommendation algorithm caters simultaneously to those who <em>listen </em>to the music (those who <em>consume </em>the recommended music) and also those who <em>create </em>the music (those who <em>provide </em>the content to be recommended). Recommendation algorithms on a platform like Spotify want to make both of these types of stakeholders as happy as possible, even in situations where it is difficult to please everyone.</p><p>As someone who <em>consumes </em>recommendations (otherwise known as the <em>consumer</em>), we often forget about other stakeholders and their needs. As a Spotify listener, I usually only care about getting recommended music that matches my interests, and not much else. If I’m someone who loves Beyoncé and no other music, I only care about getting recommended Beyoncé’s music — even if it means that I am not supporting any new, indie artists.</p><p>However, if I was a new indie artist who had just published my first album on Spotify, I would hope that the algorithm would recommend my content to listeners in the same way that it recommends Beyoncé’s content.</p><p><strong>I would hope that Spotify’s recommendation algorithm would treat me fairly as a content provider. This is provider fairness.</strong></p><h4>What is consumer fairness?</h4><p>Now consider a different platform, like LinkedIn. On LinkedIn one of the many recommendation platforms that they host is the <em>job recommender</em>. Employers post new job listings that they have, and LinkedIn users get recommended the job listings that are most relevant to them. If I am a female user on LinkedIn searching for a job, I would hope that it would give me the same recommendations as an equally qualified male user searching for a job.</p><p>But that hasn’t always happened.</p><p>In 2018 it was discovered that a job recommendation platform was recommending more higher paying jobs to men (titles like “CEO” or “Executive”) and recommending lower paying jobs to women (titles like “Secretary” or “Assistant”) [<a href="https://www.hrw.org/report/2018/04/23/only-men-need-apply/gender-discrimination-job-advertisements-china#">read the story</a>].</p><p>In this kind of system, the job recommender ideally wants to treat all users fairly and to give them content that is personalized to them, while also giving them the content they deserve to see.</p><p><strong>I would hope that LinkedIn’s recommendation algorithm would treat me fairly as a female user. This is consumer fairness.</strong></p><h4>How is “fairness” coded into recommendation algorithms?</h4><p>At this point, I have introduced both provider and consumer fairness. There are other types of fairness too, but I won’t introduce those in this article.</p><p>All of these “ethical objectives” can be implemented in many different ways into recommendation algorithms. There are dozens of algorithms that can be used to try to optimize for various definitions of fairness. Sometimes these implementations come into conflict with one another (e.g., when optimizing for provider fairness makes consumer fairness worse, or when optimizing for consumer fairness makes the accuracy of the entire system worse, etc.).</p><p>There is no perfect way to implement fairness into an AI/ML algorithm. There is no perfect implementation of fairness in recommender systems that will truly treat <em>everyone </em>fairly. Don’t forget: “what is fair for some may be unfair for others.”</p><p>If a system claims that it is treating everyone fairly, it can mislead users into a false sense of trust. On the flip side, if a system is <strong>transparent </strong>about how it is <em>trying</em> to be fair, and if it <strong>explains </strong>how its algorithm works to its various stakeholders, then it gives greater agency to the users to come to their own <em>educated</em> conclusions about the fairness of the system.</p><p>In our research, we explored what transparency <em>could </em>look like in fairness-aware recommender systems. We asked real users of real systems to share their experiences, their opinions, and their needs on these kinds of platforms.</p><p>This is what we discovered.</p><h3>Fairness and Transparency in Recommendation: The Users’ Perspective</h3><h4>Methods</h4><p>In this exploratory work, we interviewed 30 people who had interacted with a recommender system online. We began by asking them to explain to us how they thought that recommender systems work. Then, we asked them how they define “fair” treatment on these kinds of platforms, and how they feel about provider fairness versus consumer fairness.</p><p>We introduced them to a platform called <a href="https://www.kiva.org/">Kiva</a>, a microlending platform that seeks to raise financial inclusion globally. On this platform, borrowers who need money can get their loans funded through crowdsourcing, where multiple lenders can contribute to one single loan. Kiva has recently begun to play with adding a recommender system on their platform, where potential lenders can get recommended specific loans that are more relevant to them. On Kiva’s recommendation platform, the “recommended items” are actually people’s loans that they are hoping to get funded. It looks like this:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*6Ti0GnxBPO09_-I2RHJWkw.png" /><figcaption>Screenshot of several recommended loans on the <a href="https://www.kiva.org/">Kiva.org</a> website</figcaption></figure><p>For this recommender system, someone who needs money for a loan (a borrower) is the <em>provider</em>, and someone who lends money (a lender) is the <em>consumer</em> of recommendations<em>.</em></p><p>After introducing our interview participants to Kiva, we asked them how they thought that a recommender system could treat providers more fairly. For example, should Kiva recommend loans from countries that are consistently underfunded over countries that are consistently funded? Or should Kiva recommend loans from people who need money for food and water over loans from people who need money for art projects?</p><p>Finally, we asked users to describe to us how they would like to be <em>educated </em>about these fairness-aware recommendation algorithms, regardless of which fairness implementation was chosen.</p><h4>Results</h4><p>After asking participants to explain to us how they thought recommender systems worked, we received mixed responses.</p><p>While some users understood recommender systems fairly well, others thought of them as “black-box” systems that didn’t make any sense to them. Additionally, many participants shared with us that they had never thought of provider fairness, but after they had been introduced to the topic, they expressed concern for the providers of these systems.</p><p>We quickly learned that many participants were uncomfortable with the idea of fair recommendations entirely. Many participants had different definitions of what it meant for themselves or others to be treated fairly on a platform. Because of this, many interviewees desired greater transparency on these platforms.</p><p>Overall, users expressed that when a recommender system is fairness-aware, they want to be informed about <strong><em>how</em></strong><em> </em>fairness constraints change their recommendation list and <strong><em>why</em></strong><em> </em>this change is occurring<em>.</em> This is what led us to pursue explanations as a means to educate users about these things.</p><h4>Explanations as Education</h4><p>After we showed the Kiva loan recommendation example to our participants, we asked them how they would like recommendations on Kiva to be explained to them in a way that made them trust the organization and the algorithm better. Here were our three biggest takeaways:</p><ol><li><strong>First, explanations should define the system’s fairness objective for users.</strong> This was based off of the majority of our participants indicating that if a recommender system uses fairness as an objective rather than just accuracy — they would like to be aware of this.</li><li><strong>Second, explanations should not nudge/manipulate users into making a decision, even if the goal is fairness. </strong>This was based off of many participants expressing fear that organizations can choose any fairness metric to optimize for, and what is fair to some people might be considered unfair for others.</li><li><strong>Finally, explanations should disclose the motivation for using fairness as a system objective</strong>. This was in line with our goal to use explanations to <em>educate </em>users so that they understand what the fairness goals are of the organization that is hosting the recommender system, and why this organization prioritizes these goals.</li></ol><h3>Conclusion</h3><p>As recommender systems become more ubiquitous in online spaces, their impact on various stakeholders can no longer be ignored. Ethical objectives like fairness are promising improvements of these algorithms, but only if they genuinely improve the experience of <em>all </em>users of a system — including both consumers and providers.</p><p>In this work, we sought to explore potential directions for various ways to increase user’s trust and agency on a platform that seeks to be more fair to its users. We plan to conduct similar work in the future with other types of stakeholders of these systems, including providers and employees of organizations that use recommendations in their platforms.</p><p>We hope that this work sparks conversations about ways that we can make AI/ML better. We also hope that future work continues to explore best practices for transparency in these systems, so that we can better understand how algorithms impact us as users.</p><p><a href="https://dl.acm.org/doi/abs/10.1145/3450613.3456835">Read the full peer-reviewed paper here</a></p><p><em>Nasim Sonboli and Jessie J. Smith, Florencia Cabral Berenfus, Robin Burke, and Casey Fiesler. 2021. Fairness and Transparency in Recommendation: The Users’ Perspective. In Proceedings of the 29th ACM Conference on User Modeling, Adaptation and Personalization (UMAP ‘21). Association for Computing Machinery, New York, NY, USA, 274–279. DOI:https://doi.org/10.1145/3450613.3456835</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b2cabdb3108f" width="1" height="1" alt=""><hr><p><a href="https://medium.com/cuinfoscience/wait-how-do-these-algorithms-work-b2cabdb3108f">Wait, how do these algorithms work?</a> was originally published in <a href="https://medium.com/cuinfoscience">CUInfoScience</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[First Rule of Fandom: Don’t Talk About Fandom]]></title>
            <link>https://medium.com/cuinfoscience/first-rule-of-fandom-dont-talk-about-fandom-5b8ce84006?source=rss----e0e808b9deda---4</link>
            <guid isPermaLink="false">https://medium.com/p/5b8ce84006</guid>
            <category><![CDATA[fandom]]></category>
            <category><![CDATA[privacy]]></category>
            <category><![CDATA[fanfiction]]></category>
            <dc:creator><![CDATA[Brianna Dym]]></dc:creator>
            <pubDate>Tue, 20 Oct 2020 22:11:32 GMT</pubDate>
            <atom:updated>2020-10-20T22:11:32.635Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*tgNnsgtdJWKvjQVlRPrZEg.jpeg" /></figure><p>The first rule of fandom is that we don’t talk about fandom, not to people who aren’t part of fandom at least. What does this rule mean, and why does it exist? Well, it means that people writing copious amounts of fanfiction probably don’t talk about it with their boss, their peers who aren’t in fandom, and definitely not their parents. It’d just be weird.</p><p>The <em>fandom</em> I’m talking about here is the broader online community known more specifically as <em>transformative fandom</em>. It’s a community spanning multiple social media platforms where people read and write fanfiction, create fan videos, draw fanart, and geek out about all sorts of media. For those of you who don’t know, fanfiction is a type of creative work that <em>transforms</em> a source material. Someone might write an alternative ending to the <em>Game of Thrones </em>television series or the offscreen adventures of Captain Kirk and his crew.</p><p>Since before it existed online, fandom has held many intricate social norms that keep the community safe, which makes it an excellent place to observe how social norms interact with online platforms.</p><p>Many of fandom’s norms are from before the community migrated online. Unspoken rules like “don’t talk about fandom” developed out of the community’s need for secrecy back in the 70s. Media property owners weren’t so thrilled about fans writing, illustrating, and distributing fanworks, especially fanworks featuring sexually explicit content. So even before the dawn of the internet, fans had learned that in order to keep fandom safe they had to be careful about who could know about it.</p><p>And after fandom went online, <a href="https://kotaku.com/it-used-to-be-perilous-to-write-fanfiction-1826083509">authors threatening to sue</a> fan writers for plagiarism among other privacy and safety issues ensured that fandom maintained those norms to this day. In an interview study, we spoke to 25 members of fandom to learn more about their unspoken rules, how they enforce them, and what goes wrong when people violate those rules.</p><p>When fans talked about unspoken rules in fandom, they referenced ways they saw people break those rules. For example, participants described the uncomfortable feeling of watching people get harassed for making a mistake in illustrating a character or being a fan of something contentious (for example, Kylo Ren and Rey from Star Wars as a romantic pairing). They explained that, instead, fans are supposed to not judge each other for the things they are fans of. This rule helps prevent harassment in online fan spaces that are generally open to the public.</p><p>For the past several years, fan communities have gathered on open platforms like Tumblr and Twitter, where there aren’t that many tools available to users to preserve their privacy and safety. Not only did participants express fears over being harassed, they worried that their online data would wander away without knowing it. Journalists can swoop in and gather up fan art for their <em>Top 10 Sexy Vampires From Your Favorite Shows</em> article without asking permission from fan artists. One participant explained that this can be pretty scary if a fan artist doesn’t want their art style to be recognized by friends and relatives. It’s one thing for your family to know you like to draw stuff, but it’s another if that includes shirtless glittery vampires. Specifically, it can be a problem for LGBTQ+ fans that are in the closet at home but illustrating explicitly gay fanart online.</p><p>The social norms that govern fandom are in place to protect community members from bad things happening, bad things that can’t always be prevented by the social media sites the community relies on. Our participants showed us that, even though people know these norms and how to follow them, there are vulnerabilities to these norms that can still lead to major privacy and safety problems. However, we can consider opportunities to design better platforms by paying attention to when social norms break down, where their vulnerabilities are, and how the community suffers for being unable to enforce those norms more effectively.</p><p>We provide specific examples of how social norm vulnerability can hurt an online community and propose ways to help that community from multiple different perspectives. You can read more about <a href="https://cmci.colorado.edu/~cafi5706/CSCW2020_NormVulnerability.pdf">Social Norm Vulnerability and its Consequences for Privacy and Safety</a> and consider ways to help online communities remain safe and positive spaces.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=5b8ce84006" width="1" height="1" alt=""><hr><p><a href="https://medium.com/cuinfoscience/first-rule-of-fandom-dont-talk-about-fandom-5b8ce84006">First Rule of Fandom: Don’t Talk About Fandom</a> was originally published in <a href="https://medium.com/cuinfoscience">CUInfoScience</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How We’ve Taught Algorithms to See Identity]]></title>
            <link>https://medium.com/cuinfoscience/how-weve-taught-algorithms-to-see-identity-a34e2b731dcc?source=rss----e0e808b9deda---4</link>
            <guid isPermaLink="false">https://medium.com/p/a34e2b731dcc</guid>
            <dc:creator><![CDATA[Morgan Klaus Scheuerman]]></dc:creator>
            <pubDate>Tue, 20 Oct 2020 21:50:12 GMT</pubDate>
            <atom:updated>2020-10-20T21:50:12.889Z</atom:updated>
            <content:encoded><![CDATA[<h4>Constructing Race and Gender in Image Databases for Facial Analysis</h4><figure><img alt="A portrait of a Black woman looking into the camera lens. An animated gif switches between the labels for “Gender” and “Race." src="https://cdn-images-1.medium.com/max/1024/1*FTK1gY9KhmWNkGK48CVgAw.gif" /><figcaption>GIF by Morgan Klaus Scheuerman. Original image by<a href="https://unsplash.com/photos/1XAjAKrqJXY"> Ricardo Velarde on Unsplash</a>.</figcaption></figure><p><em>This blog post summarizes the above paper about how race and gender are operationalized in image databases used for training and evaluating facial analysis systems. This paper will be presented at the </em><a href="https://cscw.acm.org/2020/"><em>23rd ACM Conference on Computer-Supported Cooperative Work and Social Computing</em></a><em>, a top venue for social computing scholarship. It will also be published in the journal Proceedings of the ACM (PACM). </em><a href="https://cmci.colorado.edu/idlab/assets/bibliography/pdf/Scheuerman2020-cscw-databaseidentity.pdf"><em>A free PDF version is located here</em></a><em>.</em></p><p>Race and gender are a part of our everyday reality. We all experience the visible external features associated with race and gender when we interact with the world. As a result of these interactions, we all also hold our own affinities with race, gender, and a multitude of other identities. Race and gender have also always been incorporated into technical systems. From the U.S. Census to Facebook, we see technical representations of these identities everywhere. They have also now become commonly embedded into databases as features used to train machine-learning based algorithms. Perhaps the most salient example of this is in <em>facial analysis technology</em>, <a href="https://dl.acm.org/doi/abs/10.1145/3359246">a computer vision method for analyzing information about human faces</a>.</p><p>Within the last decade, facial analysis technology has quickly transitioned from a theoretical research problem to commercial reality. Facial recognition is already widely used; for example, embedded in technologies (e.g., <a href="https://about.fb.com/news/2019/09/update-face-recognition/">social media</a>, <a href="https://www.xfinity.com/hub/mobile/facial-recognition-on-phone">device locking</a>) and <a href="https://www.vox.com/recode/2019/7/18/20698307/facial-recognition-technology-us-government-fight-for-the-future">employed by law enforcement</a>. Facial classification is used to target<a href="https://futurism.com/japanese-taxis-facial-recognition-target-ads-riders"> marketing campaigns at specific demographics </a>and to<a href="https://www.cnbc.com/2017/11/23/facial-recognition-is-tracking-customers-as-they-shop-in-stores-tech-company-says.html"> track physical consumer behavior inside stores</a>. Historically, formally classifying race and gender has also resulted in harm; for example, the <a href="https://pubmed.ncbi.nlm.nih.gov/22136012/">differential access to healthcare for marginalized genders</a> and the targeted <a href="https://www.cairn.info/revue-l-europe-en-formation-2010-3-page-91.htm">extermination of racial groups</a>.</p><p>The outputs produced by facial analysis systems are premised on their training and evaluation data: the images used to “teach” the system what a subject looks like. In our study, we analyze race and gender in training databases from a critical discursive perspective. Specifically, we investigate how race and gender are codified into 92 image databases, many of which are publicly accessible and widely used in computer vision research.</p><p>To do this, we analyzed how race and gender are represented in image databases and how those representations are derived. What <em>categories </em>are employed to represent race and gender? What <em>sources </em>are being used to derive race and gender labels? How are database authors describing the<em> annotation procedures </em>for race and gender categories?</p><p>Facial analysis systems rely on the underlying utility, reliability, and accuracy of the ground truth data provided. We observed very few instances where database authors documented how they determined information about race and gender in facial images.</p><figure><img alt="A table showing the findings for both annotations and annotation descriptions for race and gender in our dataset." src="https://cdn-images-1.medium.com/max/1024/0*XPGRG4JERKNbn_Fm" /><figcaption>The above table shows a count of sources and annotations descriptions in explicitly annotated databases. Only 2 databases contained both sources and annotation explanations for both race and gender.</figcaption></figure><p>We saw vast inconsistencies across racial categories and methods for determining the race of a person in image data, with some databases employing three racial categories (e.g., MS-1-Celeb) and others employing seven (e.g., 10K US Adult Faces). Meanwhile, <a href="https://dl.acm.org/doi/abs/10.1145/3359246">as previously documented</a><a href="https://dl.acm.org/doi/10.1145/3274357"> in prior work</a>, we found consistently binary categories for gender.</p><p>At the same time, we found that the majority of image databases rarely contain underlying source material for how those identities are defined. Further, when images are annotated with race and gender information, database authors rarely describe the process of annotation, such as who is doing the annotating and what visual cues are leading them to make specific judgments. This made it difficult, if not impossible, for outsiders to understand how identity-based decisions are being made when constructing databases.</p><p>We posit that the lack of similar engagement with ground truth race and gender data undermined the very purpose of image databases to be usable, reliable, and accurate. Further, it ignores the sociohistorical reality of race and gender, including its weaponization in historical classification systems. We thus observed race and gender categories — that are socially and historically complex — portrayed as<em> obvious, static, and apolitical.</em></p><p>In our paper, we discuss opportunities for improving the current state of image database construction and documentation. By encouraging database authors to embrace a deeper engagement with their own positionality in constructing and annotating race and gender, we can make databases more useful and transparent to machine learning practitioners. By actively acknowledging the socio-historical context their databases stem from, we can attempt to avoid replicating the harms done by historical classification systems. By embracing more flexible forms of categorizing, like self-annotation of identity information, we can avoid overly simplistic binary categories. And by explicitly stating the limitations of use, we can help protect marginalized groups which might otherwise be harmed by the deployment of facial analysis technologies.</p><p>Morgan Klaus Scheuerman, Kandrea Wade, Caitlin Lustig, and Jed R. Brubaker. 2020. How We’ve Taught Algorithms to See Identity: Constructing Race and Gender in Image Databases for Facial Analysis. <em>Proc. ACM Hum.-Comput. Interact. </em>4, CSCW1, Article 58 (May 2020), 35 pages. <a href="https://doi.org/10.1145/3392866">https://doi.org/10.1145/3392866</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a34e2b731dcc" width="1" height="1" alt=""><hr><p><a href="https://medium.com/cuinfoscience/how-weve-taught-algorithms-to-see-identity-a34e2b731dcc">How We’ve Taught Algorithms to See Identity</a> was originally published in <a href="https://medium.com/cuinfoscience">CUInfoScience</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Entering Doors, Evading Traps]]></title>
            <link>https://medium.com/cuinfoscience/this-blog-post-summarizes-a-paper-about-how-transgender-and-non-binary-people-come-out-and-contend-bdfbcad39055?source=rss----e0e808b9deda---4</link>
            <guid isPermaLink="false">https://medium.com/p/bdfbcad39055</guid>
            <category><![CDATA[social-media]]></category>
            <category><![CDATA[coming-out]]></category>
            <category><![CDATA[transgender]]></category>
            <category><![CDATA[cscw2020]]></category>
            <category><![CDATA[research]]></category>
            <dc:creator><![CDATA[Anthony Pinter]]></dc:creator>
            <pubDate>Tue, 20 Oct 2020 21:35:15 GMT</pubDate>
            <atom:updated>2022-09-08T16:06:03.923Z</atom:updated>
            <content:encoded><![CDATA[<p><em>This blog post summarizes a paper about how transgender and non-binary people come out and contend with the visibility that accompanies those coming outs in online spaces. It will be presented at the</em><a href="https://cscw.acm.org/2020/"><em> 23rd ACM Conference on Computer-Supported Collaborative Work and Social Computing</em></a><em>.</em></p><figure><img alt="A person standing on a pink hill looks up towards a starry sky." src="https://cdn-images-1.medium.com/max/600/1*MvdQeq31coSv4rJXPKlLYw.jpeg" /><figcaption>Illustration credit: <a href="https://www.bostonglobe.com/ideas/2019/07/05/trump-wants-deny-transgender-rights-religious-grounds-but-what-gender-identity-itself-matter-soul/A7VFzaO6rUCn3krlLum4tO/story.html">S. I. Rosenbaum (Boston Globe)</a></figcaption></figure><p>LGBTQ+ people face tremendous challenges when looking for supportive and safe communities offline. Often, these spaces might be difficult to access because of the constraints of one’s physical location. For example, rural communities might not have the same supportive LGBTQ+ groups that a more urban or populous environment does. Because of constraints like this, LGBTQ+ people often turn to online spaces as avenues for support. Online spaces, like social media sites, remove barriers that might otherwise prevent access to crucial resources as one is coming to their identity.</p><p>Of course, identities and beliefs change just as readily online as they do offline. So while social media sites can be wonderful sources of resources and support, they can also present their own challenges when spaces and the people within those spaces change. Research has often positioned online spaces as places to find and be one’s authentic self without fear of stigma or persecution.</p><blockquote><strong>However, what happens when the online spaces one finds themself in no longer supports their authentic, lived self?</strong></blockquote><p>We conducted an interview study with trans people where we heard about their experiences with coming out online and contending with being visibly trans in online spaces. Participants told us about the ways they used platforms in support of their goals around finding information and support, decisions they made about who and where to come out to, their experiences with the visibility that resulted from those disclosures, and steps they took to proactively manage their audiences to protect themselves from negative experiences.</p><p>To interpret and understand our participants’ experiences, we adopt and extend a conceptual lens adopted from the work of transgender theorists Gossett, Stanley, and Burton, who were concerned with the cultural production of transness. Gossett et al. offer the conceptual metaphor of a door as cultural moments during which transgender people can access “visibility and the resources, recognition, and understanding that comes with visibility.” In online spaces, doors can be seen as the moments in which people make known some aspect of their identity to some group of others, whether it is through system features like profile bios, through what they post (e.g., pictures or status updates), or groups they join within the platform (e.g., Facebook Groups).</p><p>However, doors are simultaneously traps for some,</p><blockquote>“accommodating trans bodies, histories, and culture only insofar as they can be forced to hew to hegemonic modalities.”</blockquote><p>For Gossett et al., these cultural moments that lead to visibility benefit some at the expense of others. They acknowledge that the visibility and resultant benefits realized from doors can be good, but only as long as it is accompanied by an understanding of how that door will trap some people in societal expectations that do not mesh with their unique subjectivities (for example, being “trans” is often portrayed as being white, with the resources to undergo transition in a timely fashion; this is not all trans people’s reality).</p><p>From our participants, we heard about a variety of ways that doors can appear and turn into traps after being passed through. Doors manifest as people join new social media platforms, or create new accounts within a site that they have previously been on. For example, participants told us about creating specific accounts within sites like Instagram, where they curated their audiences to allow them to post more freely — creating a door that for them was welcoming to their identity.</p><p>If doors could be created to specifically support one’s identity, then traps manifested when participants came out in a space that previously was not centered on their trans identity. For example, coming out on Facebook created issues because of the number of connections that were often present on our participants’ Facebooks — these connections had expectations about who that person was and what they might post, and a coming out post could disrupt those expectations, in turn leading to negative reactions.</p><p>So what then happens if one is stuck in a trap? If their transness does not conform to the expectations of what trans looks like in a given moment? Gossett et al. consider this problem, and offer trapdoors as a means of escaping a trap. They define trapdoors as,</p><blockquote>“those clever contraptions that are not entrances or exits but secret passageways that take you someplace else, often someplace yet unknown.”</blockquote><p>Trapdoors are doors, certainly; they offer the opportunity for visibility, resources, and support in the face of a trap. But all doors are traps to some, meaning that a trapdoor that offers a better alternative to one person might not offer that same benefit to another. Conched in the experiences of our participants, we heard conflicting explanations of what spaces are “good” and “bad” — for some, Reddit was a “cesspool” that they wanted nothing to do with, while for others Tumblr was that “cesspool”. The tension here illustrates that for some, Reddit or Tumblr was a door that simultaneously was a trap for others.</p><p>In analyzing the experiences of our participants through this conceptual lens, we extend the theory of doors, traps, and trapdoors to social media spaces. Using our participants’ experiences to illustrate our claims, we argue that social media sites offer a myriad of doors that people might enter and become visible. In doing so, they might find that door to in fact be a trap, causing them to search for alternatives that might better support their identities — a trapdoor.</p><p>From the experiences of our participants, and our analysis of these experiences through the lens of doors, traps, and trapdoors, we make several suggestions for future work that might benefit from adopting this lens as well as considerations for designers.</p><ul><li>Doors are often opaque, not allowing one to see what the space looks like beyond until they are passed through. Designers have sought to address this by allowing people to see sample posts or group rules on Facebook before joining a group. However, the ease with which one can join many groups (and are often encouraged to join many groups through curation and recommendation systems) presents challenges when we think about how to prevent people from falling into traps. A possible solution might be to not make suggestions so readily or encourage people to join groups within the News Feed interface, instead leaving it incumbent on users to search for new groups themselves.</li><li>We see traps manifesting at both a platform (i.e., Reddit, Tumblr, Facebook) level as well as a space level (i.e., specific to an account within a platform or within a specific group on a platform). Thus, designers might consider directing users towards more supportive spaces given the information they provide about themselves via recommendations, or conversely directing them away from spaces that are less supportive of them via warning alerts (like those already in place for violent content on Facebook and Twitter).</li><li>In the wake of falling into a trap, one might look for new spaces to join (a trapdoor). Designers could consider leveraging available data about member migration from group to group to help direct an individual to a new door and subsequent space that is not a trap.</li></ul><p>Ultimately, our participants’ experiences and the analysis we provide through this lens highlights that there is not a single solution, a single door that isn’t a trap. Designers are often focused on limiting negative experiences.</p><blockquote>Our work here demonstrates that an alternative might to instead acknowledge it is impossible to design something that doesn’t negatively impact someone, and instead consider building and designing systems that enable recovery and resilience when one inevitably stumbles into a trap… or in other words, provide individuals the passageways to someplace new, someplace that might be more welcoming of their unique identity and experiences.</blockquote><p>Citation:</p><p>Anthony T. Pinter, Morgan Klaus Scheuerman, Jed R. Brubaker. 2020. Entering Doors, Evading Traps: Benefits and Risks of Visibility During Transgender Coming Outs. <em>Proc. ACM Hum.-Comput. Interact. TBD, </em>CSCW,<em> </em>Article TBD (October 2020), 27 pages.</p><p><a href="https://cmci.colorado.edu/idlab/assets/bibliography/pdf/pinter2020-cscw.pdf">Link to the full paper here</a></p><p><em>If you have questions or comments about this research, please email Anthony Pinter at </em><a href="/service/mailto:anthony.pinter@colorado.edu"><em>anthony.pinter@colorado.edu</em></a><em>.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=bdfbcad39055" width="1" height="1" alt=""><hr><p><a href="https://medium.com/cuinfoscience/this-blog-post-summarizes-a-paper-about-how-transgender-and-non-binary-people-come-out-and-contend-bdfbcad39055">Entering Doors, Evading Traps</a> was originally published in <a href="https://medium.com/cuinfoscience">CUInfoScience</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Using public data for research? Consider the ethical implications with these insights from fandom]]></title>
            <link>https://medium.com/cuinfoscience/using-public-data-for-research-consider-the-ethical-implications-with-these-insights-from-fandom-4d6777a1c438?source=rss----e0e808b9deda---4</link>
            <guid isPermaLink="false">https://medium.com/p/4d6777a1c438</guid>
            <category><![CDATA[public-data]]></category>
            <category><![CDATA[research-ethics]]></category>
            <category><![CDATA[tech-ethics]]></category>
            <category><![CDATA[fandom]]></category>
            <dc:creator><![CDATA[Brianna Dym]]></dc:creator>
            <pubDate>Fri, 26 Jun 2020 15:54:19 GMT</pubDate>
            <atom:updated>2020-06-26T15:54:19.732Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="A series of ones and zeros cascade down a blanks screen. They are colored in the pattern of a rainbow." src="https://cdn-images-1.medium.com/max/1024/1*zZJBx0oexKLkBB5TVc0pFA.jpeg" /><figcaption>Image credit: Discovod @ Dreamstime.com</figcaption></figure><p>What’s the worst that could happen if your online personal data were shared without your permission elsewhere? <a href="https://www.nytimes.com/2015/02/15/magazine/how-one-stupid-tweet-ruined-justine-saccos-life.html">For people making wildly inappropriate statements</a>, we’ve seen that they might lose their jobs. But what if you’re sharing something seemingly harmless? What’s the worst that could happen if that fanart of Kirk and Spock holding hands or statement about how much you love the new She Ra series wound up in a <em>Buzzfeed</em> article — or even in an academic article? <strong>And what ethical considerations should journalists and researchers using public data from online spaces like Twitter or Tumblr be taking into account?</strong></p><p>For most people the outcome of amplifying their online content is mild embarrassment at worst, but for many, sharing even seemingly harmless content carries a certain measure of risk. In an interview study, we explored <em>transformative fandom</em> as a case study for understanding the concerns of vulnerable communities online more broadly. This type of fandom is a corner of the internet where people gather to geek out over their favorite shows, movies, or books while sharing fanart and fanfiction about those stories within their community.</p><p>Fandom provides an interesting case study for considering ethical use of public data because the type of content <em>seems</em> harmless. But the community itself has faced stigma and marginalization. <a href="https://cmci.colorado.edu/~cafi5706/CSCW2020_MovingAcrossLands.pdf">A recent survey study we conducted</a> tracks to known demographics of transformative fandom — mostly women and non-binary participants, with 80% identifying as LGBTQ. Moreover, <a href="https://medium.com/acm-cscw/transformative-spaces-how-fandom-creates-communities-of-support-for-lgbtq-people-4123744c49cd">our other recent work on fandom</a> as an LGBTQ support community revealed that many participants aren’t out in their offline lives. Fandom also just has strong, longstanding privacy norms for a variety of reasons. In other words, though sharing artwork online might seems like a fairly low stakes way to be involved in an online community, many fandom participants are in vulnerable positions and have reason to fear for their privacy.</p><p>Like many online communities, the interactions between folks in fandom happen on publicly viewable sites like Tumblr (or they did <a href="https://slate.com/technology/2018/12/tumblr-fandom-adult-content-ban-livejournal.html">pre-Tumblr-forbidden-nipple-algorithm days</a>). People might write that fanfiction or rant about how they love x, y, and z about the latest season of <em>The Clone Wars</em>, but they’re expecting that content to remain disconnected from their offline lives. People share this content within fandom with the expectation that it will stay there while still taking precautions to preserve privacy, such as using pseudonyms. This expectation is hardly unreasonable, considering that <a href="https://journals.sagepub.com/doi/full/10.1177/2056305118763366">most social media users are unaware data like their tweets is of interest to researchers</a>.</p><p>What’s the worst that could happen if your content showed up somewhere else without your permission? Participants described fears related to losing jobs or accidentally being outed as LGBTQ to homophobic and transphobic family members. People shared data in fandom “for fandom” and were frustrated they didn’t have any way to protect it from secondary use short of choosing not to participate in these communities, which can further isolate marginalized people from <a href="https://medium.com/acm-cscw/transformative-spaces-how-fandom-creates-communities-of-support-for-lgbtq-people-4123744c49cd">important communities of support</a>.</p><p>The concerns our participants expressed are not unique to fandom. However, thanks to a long history of navigating the social stigma attached to fandom as well as dealing with <a href="https://cmci.colorado.edu/~cafi5706/group2020_fiesler.pdf">a wraught relationship with copyright battles</a>, community members of fandom are aware of the risks associated with sharing their data online, and were all too happy to share their thoughts on how they feel outsiders to the community ought to use their data. <strong>These insights inform </strong><a href="https://howwegettonext.com/scientists-like-me-are-studying-your-tweets-are-you-ok-with-that-c2cfdfebf135"><strong>how professionals like researchers and journalists can more ethically interact with public data online</strong></a><strong>, particularly when it comes from vulnerable or marginalized communities, even if the data itself does not seem sensitive.</strong></p><p>So what can researchers (and other folks making secondary use of public data, like journalists) do to help vulnerable people online stay safe?</p><ol><li><strong>Obtain permission</strong> — It doesn’t really matter if the data is public. If you’re seeking to minimize the harm you might cause, obtaining permission from folks before amplifying their content is the #1 way you can assess the risks involved for this person. Not everyone wants to be featured in the next “45 wholesome tweets to brighten your day” list, and for legitimate privacy concerns. And in cases where content might be quoted verbatim, e.g., in a research paper or in a news article, it may not be laborious to obtain permission <em>only</em> from those whose content will be amplified, even if not for everything you analyzed. (A caution: It is possible that <a href="https://cmci.colorado.edu/~cafi5706/UnexpectedExpectationsNMSPreprint.pdf">people may react poorly to knowing that they were part of a research study</a>, so keep in mind possible harms when navigating this.)</li><li><strong>Obscure data</strong> — If you can’t obtain permission or if you just want to reduce traceability to cut down on unintended harms, consider obscuring the data (through techniques like paraphrasing or other <a href="https://www.tandfonline.com/doi/abs/10.1080/1369118X.2011.641993">ethical fabrication</a>) you quote to limit how easily people can trace the content to its source. Even if your participants might not think the data you collect will harm them, you never know when your data set could suddenly <a href="https://cmci.colorado.edu/~cafi5706/group2020_designfiction_fiesler.pdf">turn into a source of targets for trolls and other bad actors</a>.</li><li><strong>Spend time with the community</strong> — Most importantly, get to know the space that you’re taking information from. This important practice applies <em>even if</em> you are only analyzing data in aggregate and there is no risk of amplification or other privacy violations. The better you know a community, the better you can understand the norms and potential harms in researching a community.</li></ol><p>When you share something in an online community, odds are you aren’t thinking about that data being used in a research context, picked up by a journalist, or weaponized by trolls. Usually, people are sharing to connect with a community and be part of a conversation. When we as researchers scoop up that data and share it in a different context, we run the risk of amplifying it to the wrong audience as well as exposing someone to risks they perhaps didn’t consider before making that post, whatever seemingly harmless content it might contain. By taking just a few precautions into consideration, we can help reduce the unintended harms our work might otherwise cause.</p><p>For more detail, check out our paper recently published in <em>Transformative Works and Cultures</em>:</p><p>Dym, Brianna, and Casey Fiesler. “<a href="https://journal.transformativeworks.org/index.php/twc/article/view/1733/2445">Ethical and privacy considerations for research using online fandom data.</a>” <em>Transformative Works and Cultures</em> 33 (2020).</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=4d6777a1c438" width="1" height="1" alt=""><hr><p><a href="https://medium.com/cuinfoscience/using-public-data-for-research-consider-the-ethical-implications-with-these-insights-from-fandom-4d6777a1c438">Using public data for research? Consider the ethical implications with these insights from fandom</a> was originally published in <a href="https://medium.com/cuinfoscience">CUInfoScience</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Day Oil Prices Went Negative — Experiments in Sonification]]></title>
            <link>https://medium.com/cuinfoscience/the-day-oil-prices-went-negative-experiments-in-sonification-380da9953f2b?source=rss----e0e808b9deda---4</link>
            <guid isPermaLink="false">https://medium.com/p/380da9953f2b</guid>
            <category><![CDATA[sonification]]></category>
            <category><![CDATA[design]]></category>
            <category><![CDATA[crude-oil-price]]></category>
            <category><![CDATA[data]]></category>
            <category><![CDATA[sound]]></category>
            <dc:creator><![CDATA[Jordan Wirfs-Brock]]></dc:creator>
            <pubDate>Tue, 19 May 2020 21:24:13 GMT</pubDate>
            <atom:updated>2020-05-19T21:24:12.926Z</atom:updated>
            <cc:license>http://creativecommons.org/licenses/by/4.0/</cc:license>
            <content:encoded><![CDATA[<h3>The Day Oil Prices Went Negative — Experiments in Sonification</h3><p>This is a story about the day oil prices went negative, on April 20, 2020. To be more specific, <a href="https://www.eia.gov/todayinenergy/detail.php?id=43495">West Texas Intermediate crude oil futures contracts for May 2020 plunged to -$40 a barrel</a> (just typing that felt awkward). I first began paying attention when this thread began unfolding in a Slack channel with former energy reporter colleagues:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/843/1*qQCCp70h33b1kBzgkB5XUA.gif" /></figure><p>Then I hopped on Twitter, where people were wondering…could they, would they, go <strong><em>negative</em></strong> ? A question that was answered soon: Yes, yes they absolutely can.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/510/1*EiDPNTOXvolgINKimvBuvA.png" /><figcaption>“Breaking: WTI crude futures trade at negative price for first time” — <a href="https://twitter.com/business/status/1252298699792883713">Tweet via Bloomberg</a></figcaption></figure><p>Oil prices eventually stabilized back in positive values the next day. And as I write this, in mid-May, WTI crude futures contracts are around $26/barrel. (Shifting baselines are a curious thing…when I was an energy journalist in 2014, it was major news when oil went from over $100/barrel to around $50/barrel over the course of 5 months.)</p><p>But on the day oil prices went negative, people were reacting with shock and confusion. With oil prices dancing below zero, we collectively wondered: What do negative oil prices even mean? As economists and <a href="https://www.marketplace.org/2020/04/21/covid-19-negative-oil-prices-us-russia-opec/">journalists explained this phenomenon</a>, my data-nerd/audiophile mind turned to a different question: <em>How do you sonify negative values?</em></p><h4><strong>What do negative values sound like?</strong></h4><p>Sonificiation is a technique for representing data with sounds — graphs for your ears. (For more on sonification generally, check out <a href="https://sonification.de/handbook/">The Sonification Handbook</a>, <a href="https://icad.org/">ICAD</a>, or <a href="https://medium.com/@jordanwb/sounds-of-a-volatile-stock-market-39ba135faa63?sk=c6411e2591dfa628fd1c13bcc8916ec7">my last blog post</a>.) Sonifying a negative value presents a mental pickle: How can you have a <em>negative </em>sound? But when I started to think about the equivalent problem in visualization, I realized that visualizing negative values is similarly bizarre. Heck, take a step back and the whole concept of <em>negative numbers</em> is counterintuitive. But because I’ve been thinking in negative numbers and reading them off charts for more than 30 years, I don’t remember what it was like to use negative numbers for the first time. It must have been confusing, even disorienting. (If you recently had to teach a child this concept, I would love to hear your take.) We’ve developed standard conventions for representing negative values visually — in spreadsheets: negative signs, red text, or parentheses; in charts, lines or bars dipping below the x-axis and color cues.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*R3Rtp8BTL1q5EoX9hWuBCg.png" /><figcaption>On the left, a spreadsheet with accounting notation uses parenthesis to indicate negative values. On the right, Yahoo Finance leans heavily on a red/green color system. Neither one is necessarily an obvious solution.</figcaption></figure><p>But we don’t have standard conventions for representing negative numbers through sound. Or, if they do exist, most of us haven’t learned how to listen for them. So here are some experiments exploring how we might sonify negative values, using crude oil prices as a test case.</p><h4><strong>But first, what does the data look like?</strong></h4><p><em>** Data munging diversion ** </em>Turns out, it’s hard to get <em>intraday</em> prices (minute by minute change) of a commodity, especially after the fact. Some financial intelligence services have this data, but it’s often hidden behind a paywall. However, I was able to read the data for 30-minute intervals off of an <a href="https://tradingeconomics.com/commodity/crude-oil">interactive graph from Trading Economics</a>. Some values were missing, which I either imputed them (using the mean of the values on either side) or estimated them based on a <a href="https://www.eia.gov/todayinenergy/detail.php?id=43495">graph from the Energy Information Administration</a>. I decided to leave the gaps where trading was closed. Below are graphs of the data from the EIA, and the data I was able to reconstruct:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*fYIXfzcqEjJ5X3NkZ1ZRUA.png" /></figure><p>The graph I created on the left uses 30-minute intervals, whereas the EIA one has even finer time-series data, thus there are some subtle differences . However, for the purposes of this sonification exercise, where my goal is to convey the <strong><em>strangeness</em></strong> of negative values, the 30-minute intervals will do the trick.</p><h4><strong>Finally…the sonification part!</strong></h4><p>First, to establish a baseline for our listening, let’s try a <strong>straight-up vanilla sonification</strong>: a simple audio graph of crude oil prices, with price mapped onto pitch. I created this using <a href="https://app.twotone.io/">TwoTone.io</a>, a free web-based sonification tool. The 48-hour period of April 20 and 21 is mapped onto 24 seconds of sonification time. Remember, we have trading data from every 30-minutes, so that means you’ll hear roughly 2 hours (4 tones) compressed into each second of the sonification. I’ve chosen a marimba and a 3-octave range to capture the $52-spread from the maximum ($16.72) to the minimum (-$39.55).</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fw.soundcloud.com%2Fplayer%2F%3Furl%3Dhttps%253A%252F%252Fapi.soundcloud.com%252Ftracks%252F821339911%26show_artwork%3Dtrue%26in%3Dj_wirfs-brock%252Fsets%252Fsonifying-negative-crude-oil&amp;display_name=SoundCloud&amp;url=https%3A%2F%2Fsoundcloud.com%2Fj_wirfs-brock%2Foil-prices-translated-into-marimba-notes%3Fin%3Dj_wirfs-brock%2Fsets%2Fsonifying-negative-crude-oil&amp;image=https%3A%2F%2Fi1.sndcdn.com%2Fartworks-gstldoyCMCcMcrz0-npgK2g-t500x500.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=soundcloud" width="800" height="166" frameborder="0" scrolling="no"><a href="https://medium.com/media/72f47cd1b0807f10a195bdc5e2163b71/href">https://medium.com/media/72f47cd1b0807f10a195bdc5e2163b71/href</a></iframe><p>You can hear how the price plunges and then recovers, but there’s no indication of when the price crosses over into negative values, or when it nears zero. <em>So how can we make the negative values — which are the critical element of the dataset we want to emphasize — more noticeable?</em></p><p>First, let’s try <strong>changing the <em>timber</em></strong> — a musical term for the quality or texture of the sound — of the notes by changing the instrument. When the prices go negative, the notes will <strong>switch from a marimba to a trumpet</strong>. Why those two instruments? For the very practical reason that they have overlapping octave ranges, which means the frequency/pitch of the notes can change seamlessly.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fw.soundcloud.com%2Fplayer%2F%3Furl%3Dhttps%253A%252F%252Fapi.soundcloud.com%252Ftracks%252F821339902%26show_artwork%3Dtrue%26in%3Dj_wirfs-brock%252Fsets%252Fsonifying-negative-crude-oil&amp;display_name=SoundCloud&amp;url=https%3A%2F%2Fsoundcloud.com%2Fj_wirfs-brock%2Foil-prices-marimba-changes-to-trumpet-for-negative-values%3Fin%3Dj_wirfs-brock%2Fsets%2Fsonifying-negative-crude-oil&amp;image=https%3A%2F%2Fi1.sndcdn.com%2Fartworks-gstldoyCMCcMcrz0-npgK2g-t500x500.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=soundcloud" width="800" height="166" frameborder="0" scrolling="no"><a href="https://medium.com/media/3979365d93d0a4ff97f7a3f5024c40b8/href">https://medium.com/media/3979365d93d0a4ff97f7a3f5024c40b8/href</a></iframe><p>Ok, this gets a bit closer to capturing the unsettling nature of the data. But we can do better: Let’s try <strong>changing the key from <em>major</em> to <em>minor</em></strong> when the price goes negative. I got this idea from <a href="https://medium.com/@_stc/emotion-sonification-aacd304b4757">Agoston Nagy, who used major and minor keys to sonify positive and negative emotions of Tweets</a>. Now, I’m not a musician (in fact, I’m embarrassed by how little music theory I retained from my years of violin lessons), thus I wasn’t sure what keys to use. So I turned to this <a href="https://www.wmich.edu/mus-theo/courses/keys.html">delightful translation of Schubart’s musical key characteristics</a> from 1806. The description of F# Minor stood out to me:</p><blockquote><em>“A gloomy key: it tugs at passion as a dog biting a dress. Resentment and discontent are its language.”</em></blockquote><p>For going from positive to negative values, we’ll use a type of key modulation called <em>relative modulation</em>, which takes us from A Major (<em>“This key includes declarations of innocent love, satisfaction with one’s state of affairs; hope of seeing one’s beloved again when parting; youthful cheerfulness and trust in God.”</em>) to F# Minor. And for going from negative back to positive, into a world that is not quite the same as before, we’ll use parallel modulation from F# Minor to F# Major (<em>“Triumph over difficulty, free sigh of relief uttered when hurdles are surmounted; echo of a soul which has fiercely struggled and finally conquered lies in all uses of this key.”</em>). Fun fact: <a href="https://www.musicnotes.com/now/tips/a-complete-guide-to-musical-modulation/">This modulation that is used in the song “Happy Together.”</a> Here’s what it sounds like:</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fw.soundcloud.com%2Fplayer%2F%3Furl%3Dhttps%253A%252F%252Fapi.soundcloud.com%252Ftracks%252F821339881%26show_artwork%3Dtrue%26in%3Dj_wirfs-brock%252Fsets%252Fsonifying-negative-crude-oil&amp;display_name=SoundCloud&amp;url=https%3A%2F%2Fsoundcloud.com%2Fj_wirfs-brock%2Foil-prices-key-change-modulation-for-negative-values%3Fin%3Dj_wirfs-brock%2Fsets%2Fsonifying-negative-crude-oil&amp;image=https%3A%2F%2Fi1.sndcdn.com%2Fartworks-gstldoyCMCcMcrz0-npgK2g-t500x500.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=soundcloud" width="800" height="166" frameborder="0" scrolling="no"><a href="https://medium.com/media/38a971db2a22c7b7a9a09eaffd2ee7d7/href">https://medium.com/media/38a971db2a22c7b7a9a09eaffd2ee7d7/href</a></iframe><p>I don’t notice these changes as much; that could simply be my untrained musical ear, or the fact that I am not modulating on the right chords. Perhaps this could be more clear by including the essential chord of each key in the background behind the marimba and trumpet melodies. Musicians out there, please chime in if you have ideas on how to improve this!</p><p>Let’s add a couple of final elements to really make the switch from positive to negative values noticeable: <strong>a <em>hissing</em> sound in the negative range</strong> (inspired by <a href="http://faculty.salisbury.edu/~smhetzler/Minicourse2008/">Stephen Hetzler’s sonifications of mathematical functions</a>) and a <strong><em>record scratch</em> when values cross from positive to negative</strong> or vice versa. Here’s the result:</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fw.soundcloud.com%2Fplayer%2F%3Furl%3Dhttps%253A%252F%252Fapi.soundcloud.com%252Ftracks%252F821339806%26show_artwork%3Dtrue%26in%3Dj_wirfs-brock%252Fsets%252Fsonifying-negative-crude-oil&amp;display_name=SoundCloud&amp;url=https%3A%2F%2Fsoundcloud.com%2Fj_wirfs-brock%2Foil-prices-hiss-for-negative-values-and-record-scratch-for-sign-change%3Fin%3Dj_wirfs-brock%2Fsets%2Fsonifying-negative-crude-oil&amp;image=https%3A%2F%2Fi1.sndcdn.com%2Fartworks-gstldoyCMCcMcrz0-npgK2g-t500x500.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=soundcloud" width="800" height="166" frameborder="0" scrolling="no"><a href="https://medium.com/media/0ecefb239784c4c8566096a024511f49/href">https://medium.com/media/0ecefb239784c4c8566096a024511f49/href</a></iframe><p>Which of these techniques, if any, did you find effective? How might you improve on any of these concepts? For me, this is a good start, but it still doesn’t fully capture the shock and confusion that the data evoked.</p><h4><strong>Let’s get a little weird with it.</strong></h4><p>This time, we’re going to throw mapping prices onto pitch out the window. Instead, we’ll <strong>use a familiar song clip and <em>modulate the tempo</em> as the price gets higher or lower</strong>. This idea is inspired by <a href="https://www.pri.org/stories/2015-08-14/how-many-floods-will-these-american-cities-have-2030-2045">Kuang Keng Kuek Ser’s project for PRI using “Ice Ice Baby” to convey projected sea level rise</a>. No shade to Vanilla Ice, but we’re going to use much better artist: The Talking Heads. Specifically, the song <a href="https://open.spotify.com/track/4oEKgLZ8uTSioNKuO8Kydu?si=M7RlURubQveJPrEY-YhU_g">“Nothing But Flowers,”</a> because it’s about ambivalence toward a post-apocalyptic world. Also, it has a line that goes:</p><blockquote><em>From the age of the dinosaurs / Cars have run on gasoline.</em></blockquote><p>The tempo of the song will correspond to the magnitude of the price, scaled up or down from $16/barrel. That is, as values get close to $0, the song is going to get very, very slow. If the price were to get up to $32/barrel, it would be double-speed. Instead of just speeding up and slowing down a song, we’re also going to <em>reverse</em> it. So all of the negative values will be played backwards. Negative values close to $0, like -$5/barrel, will be slow and backwards; negative values far away from $0, like -$39/barrel, will be fast and backwards.</p><p>We’re going to focus on a 16-second clip, looped over and over and distorted according to the data, so that you can hone in on the data-driven changes without being distracted by the many tempo and key changes throughout the rest of the song.</p><p>Ready? Here’s a version with me narrating what’s happening (I recommend listening to this first…it’s also my favorite sonification experiment):</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fw.soundcloud.com%2Fplayer%2F%3Furl%3Dhttps%253A%252F%252Fapi.soundcloud.com%252Ftracks%252F821339962%26show_artwork%3Dtrue%26in%3Dj_wirfs-brock%252Fsets%252Fsonifying-negative-crude-oil&amp;display_name=SoundCloud&amp;url=https%3A%2F%2Fsoundcloud.com%2Fj_wirfs-brock%2Foil-prices-interpreted-1%3Fin%3Dj_wirfs-brock%2Fsets%2Fsonifying-negative-crude-oil&amp;image=https%3A%2F%2Fi1.sndcdn.com%2Fartworks-gstldoyCMCcMcrz0-npgK2g-t500x500.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=soundcloud" width="800" height="166" frameborder="0" scrolling="no"><a href="https://medium.com/media/a01b89617c4392deb8962f4bc2d5f5c7/href">https://medium.com/media/a01b89617c4392deb8962f4bc2d5f5c7/href</a></iframe><p>And for the extra brave souls among you, here’s a non-narrated version:</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fw.soundcloud.com%2Fplayer%2F%3Furl%3Dhttps%253A%252F%252Fapi.soundcloud.com%252Ftracks%252F821339932%26show_artwork%3Dtrue%26in%3Dj_wirfs-brock%252Fsets%252Fsonifying-negative-crude-oil&amp;display_name=SoundCloud&amp;url=https%3A%2F%2Fsoundcloud.com%2Fj_wirfs-brock%2Foil-prices-interpreted-through%3Fin%3Dj_wirfs-brock%2Fsets%2Fsonifying-negative-crude-oil&amp;image=https%3A%2F%2Fi1.sndcdn.com%2Fartworks-gstldoyCMCcMcrz0-npgK2g-t500x500.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=soundcloud" width="800" height="166" frameborder="0" scrolling="no"><a href="https://medium.com/media/18cf9aaf6f35032c5c229e11ce8d9db5/href">https://medium.com/media/18cf9aaf6f35032c5c229e11ce8d9db5/href</a></iframe><p>Weird, huh? I find it both discombobulating and mesmerizing — which sums up that mood of that day pretty well.</p><h4><strong>One final experiment.</strong></h4><p>Let’s take a moment and focus on what <em>negative values</em> truly mean. In the context of crude oil prices, it meant that there was not enough room to store the oil being produced, because the demand for oil had dropped. Thus, people with oil were willing to pay others to take it off their hands. It was a switch in polarity, a change in direction — things flowed one way, and then they flowed the other way.</p><p>What might that mean in the context of sound? Perhaps when prices are positive, sounds are played <em>for you</em>. But then, when they become negative, <em>you, the listener, now have to create the sounds</em>. <strong>It’s a switch in the direction of flow of sound production. </strong>This funky concept is not really meant to be a representation of the data that you might actually use to read off a number or make a calculation, but rather an experience to make you stop and think about what the data actually mean.</p><p>Because this type of sonification requires <em>listener participation</em>, I’ve written it in the form of a score:</p><blockquote>Play the video clip below. When values are positive, listen to the notes. When values are negative, perform the notes. Lower negative values correspond to lower tones.</blockquote><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FkeLDhs9l078%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DkeLDhs9l078&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FkeLDhs9l078%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/b8908d38fbea80e9d346af33d99afaf8/href">https://medium.com/media/b8908d38fbea80e9d346af33d99afaf8/href</a></iframe><p>Here’s my attempt at performing the score:</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fw.soundcloud.com%2Fplayer%2F%3Furl%3Dhttps%253A%252F%252Fapi.soundcloud.com%252Ftracks%252F821339791%26show_artwork%3Dtrue%26in%3Dj_wirfs-brock%252Fsets%252Fsonifying-negative-crude-oil&amp;display_name=SoundCloud&amp;url=https%3A%2F%2Fsoundcloud.com%2Fj_wirfs-brock%2Foil-prices-performing-the-negative-values%3Fin%3Dj_wirfs-brock%2Fsets%2Fsonifying-negative-crude-oil&amp;image=https%3A%2F%2Fi1.sndcdn.com%2Fartworks-gstldoyCMCcMcrz0-npgK2g-t500x500.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=soundcloud" width="800" height="166" frameborder="0" scrolling="no"><a href="https://medium.com/media/2e083fd10e02b95d372bccfc99d8358d/href">https://medium.com/media/2e083fd10e02b95d372bccfc99d8358d/href</a></iframe><p>As I was doing this, I realized that I was relying on my visual senses to see the data and then translate it, in as near to real time as I could, into sound — something that was extremely difficult, and that I’m sure I butchered. Intuitively, I wanted to make the sounds <em>louder</em>, instead of lower, as the values got more negative. But the quality of the sounds I made wasn’t really the point — <strong>the point was the experience of having to produce them</strong>.</p><h4>Where else can these experiments take us?</h4><p>How else might we convey negative values with sound? Perhaps by using sounds that have some conceptual metaphorical connection with polarity, like the sounds of breathing in and breathing out, stacking up Jenga blocks and then removing them, or writing on a white board and then erasing it. My favorite part about doing these sonifications is that I get to hear your reactions and ideas for taking these experiments further. So what do <em>you</em> think negative values should sound like?</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=380da9953f2b" width="1" height="1" alt=""><hr><p><a href="https://medium.com/cuinfoscience/the-day-oil-prices-went-negative-experiments-in-sonification-380da9953f2b">The Day Oil Prices Went Negative — Experiments in Sonification</a> was originally published in <a href="https://medium.com/cuinfoscience">CUInfoScience</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Risk Communication is a Risky Business: The Importance of Communicating Timelines during Periods…]]></title>
            <link>https://medium.com/cuinfoscience/risk-communication-is-a-risky-business-the-importance-of-communicating-timelines-during-periods-53ef6a4f01b?source=rss----e0e808b9deda---4</link>
            <guid isPermaLink="false">https://medium.com/p/53ef6a4f01b</guid>
            <category><![CDATA[covid19]]></category>
            <category><![CDATA[risk-communications]]></category>
            <category><![CDATA[crisis-informatics]]></category>
            <dc:creator><![CDATA[Leysia Palen]]></dc:creator>
            <pubDate>Thu, 30 Apr 2020 23:22:08 GMT</pubDate>
            <atom:updated>2020-04-30T23:22:08.268Z</atom:updated>
            <content:encoded><![CDATA[<h3>Risk Communication is a Risky Business: The Importance of Communicating Timelines during Periods of Suspended Animation</h3><p>I conduct research in the area of <em>crisis informatics</em>. Public health pandemics are different enough from my primary domain for research — the <em>informatics of natural hazards </em>— that I’ve been getting a sense of the situation by putting my own household in order while figuring how I might formally research this new crisis that has descended on us all. Why do I make this distinction between public health pandemics and other kinds of crises? It is because we know that <em>online information behaviors</em> are different across different kinds of crisis events. <a href="https://www.youtube.com/watch?v=Nw0LofRol40">People behave quite differently online in the face of a hurricane versus a school shooter</a>, for example, because the problems to be solved by those affected are different. Pandemics are going to be different yet again in terms of online information behaviors, and so there is a lot for me, even as a 15 year veteran of crisis informatics research, to learn. What we know about the digital experience in other kinds of crises might not fully generalize to a pandemic.</p><p>One of the subareas for which I conduct research is <em>risk communication</em>, or how official risk messages are issued to the public, and how the public responds to them. This is an already <a href="https://scholar.google.com/scholar?hl=en&amp;as_sdt=0%2C6&amp;q=risk+communication&amp;btnG=">deep literature</a> and, most recently, my colleagues and I have been adding to that scientific knowledge by examining how graphical risk communication diffuses quantitatively and qualitatively across social media <a href="https://www.melissabica.com/assets/files/Bica-CHI2019-CommunicatingHurricaneRisks.pdf">during hurricane events</a>. This pandemic is teaching us yet more, and it is on this topic that I would like to offer my first reflections on this crisis.</p><p>Of the many things we know about risk communication, I’ll share a few points. First, we know that risk communication is very hard to do. This is because to convey risk means to convey uncertainty, and to convey uncertainty means to engage in probabilities. Probabilities are hard to explain succinctly to a wide audience who are being reached by broadcast media of various kinds, all of whom might have a different relationship to the risk. Another important point is that risk communication needs to<em> avoid foreclosing uncertainty</em>; it also <em>needs to avoid keeping uncertain that which can be closed</em>. This means that messaging cannot be unequivocal unless uncertainty has lifted. But succinctness, which is also valued in risk communication, favors unequivocal statements. This is why I like to say that <em>risk communication is a risky business</em>.</p><p>In the case of the COVID-19 threat, social distancing and sequestering — which is not far from the sheltering-in-place practice that is employed in some natural hazards — is an activity we engage in even though we cannot <em>see</em> the hazard first-hand, and that is what makes us more uncertain. The end point for sheltering-in-place for natural hazards might not always be entirely clear — but we are pretty sure that we’ll know it when we see it. For this pandemic, the threat is invisible, and so there is no way to then know on our own when it is “safe to come out.”</p><p>Social distancing is also upsetting because it breaks normal routines, even under the best of conditions (eg., when one’s livelihood is not under direct threat). I am seeing my children, spouse, neighbors, students, and myself feel anxious about having to make sudden changes in our plans. But the worst of it is that we don’t know <em>when</em> it will end.</p><p>At this stage of the pandemic, no one does. However, one of the problems with risk communication as I see it in this case — and in the US at least — is insufficient <em>externalization of assumptions about how instructions map to timelines</em>. For example, rather than just closing schools and universities as has happened rather suddenly this week, we could say with a bit more elaboration:</p><p><em>“Our university is moving to all on-line remote teaching for the rest of the semester. The assumption we are working with at the time of this writing suggests that 5 weeks is sufficient. However, with only 2 more weeks of the semester beyond that, we want to support students and instructors, and limit switching back to in-person settings once this new commitment to teaching pedagogy is made. We want the transition to summer session to be as smooth as possible, given this disruption today.”</em></p><p>And for the more general case:</p><p><em>“We are now working with the assumption that X weeks of social distancing could significantly reduce the rate of spread. However, we expect that in one week of regionally-managed social distancing, we will have more empirical knowledge about the rate of contagion, and will update this estimate then.”</em></p><p>Risk communication to socially distance without an end in sight is individually distressing and societally devastating. Some temporal bounds, even if tenuous, with clear explanation for why those bounds make sense at the moment of communication, could be deeply reassuring, and improve mental health, market health, and compliance with social distancing practices. Suspended animation might be more readily recast as being efficient, less scary, and less risky to other parts of our lives such that we might be less reticent to participate in this critical collective effort.</p><p><em>Leysia Palen has been conducting crisis informatics research since 2004 as a Professor of Information Science and Computer Science at the University of Colorado Boulder. She and her research can be found at </em><a href="https://cmci.colorado.edu/~palen/"><em>https://cmci.colorado.edu/~palen/</em></a><em>.</em></p><p><em>This is the first article in a </em><a href="https://medium.com/@palen"><em>series</em></a><em> on research and reflection during the novel coronavirus pandemic.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=53ef6a4f01b" width="1" height="1" alt=""><hr><p><a href="https://medium.com/cuinfoscience/risk-communication-is-a-risky-business-the-importance-of-communicating-timelines-during-periods-53ef6a4f01b">Risk Communication is a Risky Business: The Importance of Communicating Timelines during Periods…</a> was originally published in <a href="https://medium.com/cuinfoscience">CUInfoScience</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Liminal Experience of Crisis: Facing Suspended Animation in the Era of COVID-19]]></title>
            <link>https://medium.com/cuinfoscience/the-liminal-experience-of-crisis-facing-suspended-animation-in-the-era-of-covid-19-b9f1cdb21f26?source=rss----e0e808b9deda---4</link>
            <guid isPermaLink="false">https://medium.com/p/b9f1cdb21f26</guid>
            <category><![CDATA[coronavirus]]></category>
            <category><![CDATA[liminality]]></category>
            <category><![CDATA[covid19]]></category>
            <category><![CDATA[crisis-informatics]]></category>
            <dc:creator><![CDATA[Leysia Palen]]></dc:creator>
            <pubDate>Thu, 30 Apr 2020 23:22:00 GMT</pubDate>
            <atom:updated>2020-04-30T23:22:00.148Z</atom:updated>
            <content:encoded><![CDATA[<h4>To be in crisis is to be betwixt and between.</h4><p>In times of crisis, we often feel betwixt and between. Anthropologists have talked about these moments as rites of passage, and are associated with planned events in life (like graduation, moving house, changing jobs) (Turner, 1987). To be in transition — to be neither here nor there — is to experience <em>liminality (</em>Boland, 2013). Unexpected events are perhaps the most acute in giving rise to <em>liminal spaces</em>, because endings are sudden, and the plans for what comes next are unknown. For unexpected events that happen on a regional — or even a global — scale, it is no longer simply the ritual of “graduation” that marks transition from one stage of human development to another — it is now a graduation ceremony <em>that might not happen at all</em>. This loss of experience that comes from the suspended animation of sequestering is <em>liminality in the large</em>, and though we are experiencing it together, which helps, we are also experiencing it alone, which is as painful as it has ever been.</p><p>These ideas come from in-progress work with my colleagues <a href="https://www.melissabica.com/">Melissa Bica</a>, <a href="https://cires.colorado.edu/researcher/jen-henderson">Jen Henderson</a>, <a href="https://twitter.com/jspinney19?lang=en">Jen Spinney</a>, and <a href="https://scholar.google.com/citations?user=NC29BNoAAAAJ&amp;hl=en">Erik Nielsen</a>, in which we have been examining how and why people post online during a multi-hazard event, which in turn stems from Erik’s atmospheric science work on the conceptualization of “TORFFs” — tornados and flash flooding—that can co-occur. TORFFs are an extreme (but not rare) case to examine: What does one do in the face of conflicting risk communication that instructs to “get to higher ground” because of flash flooding, but to “lower ground” for tornados? What is one to do when choices are hardly choices at all, but instead a weighing of unknown risk?</p><p>Furthermore, what is the effect of maintaining a digital presence online during periods of liminality? What we are finding is that, like our studies before, <strong><em>people who are affected by a crisis post online to seek help and advice in a variety of ways</em></strong><em>. </em>But often, once help and advice are offered, the discursive acts take on other meaning, particularly when the solution to the trouble is unknown and needs waiting out. Preliminary findings from our current research and prior research that have bearing on the COVID19 pandemic include the following points:</p><p><strong><em>A. People may find that documenting the experience online is helpful, particularly as it is felt personally and locally.<br></em></strong>Documenting experiences makes them more real; the act of documenting and sharing the documentation <em>structures</em> the liminal space, and gives it purpose that one can control. In the early days of the COVID-19 event, we are seeing people post pictures of empty market shelves. Some posts point at the perceived problem of so-called “panic shopping,” but we might take the total of it to communicate how strange and unexpected the world is right now at this very moment.</p><p>It might also be similar to sending photos from the warfront:<em> This is really happening.</em> And in anthropologist’s terms, it might be a way to confirm the beginning of a state of transition for oneself, and to prepare others who might also be watching.It might also be a way to say now, “<strong>I am here</strong> (in this space of uncertainty),” and, for afterward, “<strong>I <em>was</em> here</strong>,” not unlike the narrative trope of a person stranded on a desert island who marks the days by etching lines in rock.</p><p>For more geographically contained disasters like TORFFS, public posting might be a way of shouting out, “I don’t want to be alone.” For the pandemic, where we all face the same threat, it might more hopefully be, “You are not alone.”</p><p><strong><em>B. People seek to create community.<br></em></strong>One way to manage the anxiety of uncertainty is to help others with the resources or expertise that oneself possesses. We are seeing self-organized community movements of gathering supplies and helping those who cannot leave their homes during the COVID-19 crisis, actions that we have seen flourish since the Haiti earthquake in 2010 in an unimaginably large set of creative ways (and is worthy of a recap in a future post to properly describe and credit those activists who have organized such activities, as well as those researchers who have carefully documented it).</p><p>Another way in which this is manifesting is through people in other regions with advanced COVID-19 diffusion advising others about what their futures may be like if they don’t take precautions. For example, residents in Italy and Korea are issuing warnings to residents of other countries about the consequences of failing to take early action. <a href="https://twitter.com/velofemme?ref_src=twsrc%5Egoogle%7Ctwcamp%5Eserp%7Ctwgr%5Eauthor">Sarah Vieweg</a> and I also saw this behavior across two university shootings in 2007 and 2008, <a href="https://dl.acm.org/doi/10.1145/1460563.1460583">where students from the first university advised students from the second about how public or private they should choose to be on what was then the new world of Facebook</a>.</p><p><strong><em>C. People may continue to seek normalcy, but then must also negotiate power differentials.<br></em></strong>In seeking normalcy for oneself in times of uncertainty, we might expect services to be running as normal. The problem of course is that those people and groups who offer those services might be beholden to those who demand them. Supplies might be short. Deliveries might be delayed. Young children at home might put a productive work day out of reach. Or, alternatively, some services might be running as normal (construction, local food markets), and might need extra financial attention for those with the means to assist. Crises are the time to make room for reasonableness and generosity.</p><p>So what might these findings from prior experiences of other liminal experiences inform what we might do during this one? Here’s how I am interpreting these findings for myself in these early days of COVID-19, and turning them into action.</p><p><strong><em>1.</em> <em>I am maintaining daily routines.<br></em></strong>For me, this begins by doing the same mundane things I do every morning, now more than ever: wake up at the same time; open the window shades; put in a load of laundry; start the coffee; unload the dishwasher; and then guiltlessly sit down to read or watch something not news related that I enjoy. Then, and only then, do I turn to news and email.</p><p><strong><em>2.</em> <em>I am recognizing that things are different, and that by doing something new, even if small, I am marking this period of change.<br></em></strong>For me, I am reading news for much longer than my typical daily routine allows. I therefore need to stop from deep-ending on it with a brief walk outside specifically to see how the plants are doing. This is not time I would have normally taken, and so this step is meditative, and a small new ritual that marks time.</p><p>I also find myself returning to Facebook, which I had previously given up based on my concerns about the platform’s lack of action to misinformation spread. I still feel strongly about this, but I also need to connect with others about this pandemic experience in a semi-public way. My unwillingness to give up the familiarity of Facebook right now leads me to my next observation…</p><p><strong><em>3.</em> <em>I am acknowledging that loss cannot be given up in total and at once.</em></strong><em><br></em>For me, this means I am not quite ready to admit that my son’s graduation ceremony will be cancelled. And I am not sure he is either, so until we have confirmation of that, we’ll keep it open as a possibility without getting our hearts set. Similarly, the high school prom is likely not going to happen, but my daughter and I kept our plans to look for a dress. We ordered one on-line. We don’t know if it will be worn any time soon, but we like seeing the pretty frock on the closet door. The experience of finding it together is one the pandemic can’t take away.</p><p><strong><em>4.</em> <em>I am checking myself to make sure that I am plugging into power differentials in ways that benefit others.</em></strong><em><br></em>Some of us can work from home; many cannot. I am trying to be thoughtful about how I can keep some plans that help businesses and others’ incomes in ways that do not further cause them stress. For example, we had existing plans to re-roof our house this month. This is a service that might be able to continue safely and benefit a local business and their employees, but I will be sure to ask if that is true. In addition, I am thinking about moving up a landscaping job I had planned in about year for summer 2021 to this spring, if that helps another small business. The money returned to me from a cancelled trip is now available to use in a different way. I would schedule entirely for their optimization, because they, too, have child care challenges. I hope to encounter other ideas as time goes on.</p><p><strong><em>5.</em> <em>I am acknowledging my need for seeking and providing information.<br></em></strong>This is what you are witnessing here, in my effort to blog and to tweet, which is something I have found hard to do. It is difficult because it requires distilling months-long research into something brief, which feels like I am short-changing the investigation and cutting corners on the expertise others have to give. I also have a hard time keeping up with respect to demands closer to home and work. I think my solution will be to include this into my daily routine (routines!) that productively bounds the time a little.</p><p><strong><em>6.</em> <em>I am acknowledging my need for optimism.</em></strong><em><br>“</em>Today” is managed by routines. Two weeks from now is utterly unknown. Six months from now is as unknown as life usually is — at least I think — but the rituals I love are pretty likely. My kids and I decided to make candles during their now extended spring break that we will give as presents for the fall and winter holidays. It is an activity that is based on a forecast for a new-normal future. That it is attached to shared rituals also helps us imagine being on the other side of the liminal experience.</p><p>We are all in a state of liminality, and perhaps in suspended animation. Perhaps the best thing we can do is to try to find a way to <em>unsuspend</em>, and take action within our new household orders, while also treating oneself and others with care as we navigate this massive transition to an uncharted destination.</p><p><em>Leysia Palen has been conducting crisis informatics research since 2004 as a Professor of Information Science and Computer Science at the University of Colorado Boulder. She and her research can be found at </em><a href="https://cmci.colorado.edu/~palen/"><em>https://cmci.colorado.edu/~palen/</em></a><em>.</em></p><p><em>This is the second article in a </em><a href="https://medium.com/@palen"><em>series</em></a><em> on research and reflection during the novel coronavirus pandemic.</em></p><p>Victor W. Turner. 1987. Betwixt &amp; between: The liminal period in rites of passage. In <em>Betwixt &amp; Between: Patterns of Masculine and Feminine Initiation</em>, Louise Carus Mahdi, Steven Foster and Meredith Little (eds.). 3–19.</p><p>Tom Boland. 2013. Towards an anthropology of critique: The modern experience of liminality and crisis. <em>Anthropological Theory </em>13, 3: 222–239. <a href="https://doi.org/10.1177/1463499613496731">https://doi.org/10.1177/1463499613496731</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b9f1cdb21f26" width="1" height="1" alt=""><hr><p><a href="https://medium.com/cuinfoscience/the-liminal-experience-of-crisis-facing-suspended-animation-in-the-era-of-covid-19-b9f1cdb21f26">The Liminal Experience of Crisis: Facing Suspended Animation in the Era of COVID-19</a> was originally published in <a href="https://medium.com/cuinfoscience">CUInfoScience</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>