 Hi everyone, thank you so much for having me at the wiki workshop. Today I'm going to be presenting on a term we call pure produced friction specifically looking at the tradeoffs of page protection on Wikipedia. So just a little bit about who's involved on this project. So my name is Leah. I'm a second year PhD at the group lens lab over at the University of Minnesota. I'm advised by Stevie Chancellor who's also the supervising author on this work and then Nick Vincent notable wiki workshop alum is also on this project as well. So for today I'm going to be presenting on page protection research we've been doing for about the past 18 months. And I'd love to open up some dialogues about this work specifically about the Wikipedia ideologies that are underpinning page protection. Are there any interesting data sources that we're missing particularly in our proposed methods? And we have some design recommendations that we'll present on but I'm curious if there are other design recommendations specifically ones you see really aligning with the wikimedia foundation. So to go over some terms a friction is a design decision that makes an interaction more effortful. So we think about a user pathway from A to B a friction can be conceptualized as a hurdle. This can be something like requiring an account. So it's not blocking the user from doing anything but it does require more effort on their part. Now content moderation introduces barriers while users trying to create a lasting contribution to the platform. Some notable interventions are things like banning and de-platforming so they block a user from being able to get from point A to point B entirely. Some of these barriers are ineffective though so for example hashtag banning on Instagram communities figured out how to subvert it by using lexical variations. So even though it was intended as a barrier it became more of a friction. Now page protection on Wikipedia is specifically a friction because it requires users to gain certain credentials in order to be able to edit a page. In theory any user can gain these credentials with enough time and effort. Furthermore it represents an intentional trade-off between open participation and information quality to very big values in the Wikipedia community. This tension is made clear when we look at the Wikipedia page protection policy which talks about stopping disruption while still allowing productive editors to make changes. Now my work focuses on asking is page protection achieving its intention. Specifically for work that is under minor revisions at CSCW we looked at a uniquely participatory category of Wikipedia namely the internet culture category and we looked at 108 articles that are currently page protected. When we look at drop-off so the amount of users that edited before a page was protected but didn't edit after we find that there's about 75% of users who are dropping off 7, 14 and 30 days after a page is protected. To put this into context we created a control set using propensity score matching you can find that the drop-off in blue is significantly lower. Moreover the average number of anonymous users in our treatment set was about 24% of pre-intervention users so we can see that it's very clear more than just anonymous users who are typically the bad actors are dropping off or stopping editing after a page is protected. We found some other very interesting findings in this work for the sake of time I'm going to breeze past them but I'm happy to answer questions about them later. So a lot of the work we did for the CSCW submission was descriptive in nature. We're proposing work that's a bit more causal so specifically we're looking to use regression discontinuity design replicating the hill and shaw methods where they explored the hidden costs of requiring accounts however we plan on using page protection as the intervention rather than account requiring so this will allow us to make these quasi causal claims about what dynamics is page protection specifically causing. Now we want to explore this tension of friction under two main research questions first of all who does page protection affect namely looking at power users so just page protection call upon these experienced wikipedia users once a page is protected and have them stop the disruption. Now two is looking at minority editors so it's important to note here that historically wikipedia has a massive gender added gap. Now if page protection is exacerbating this gap or maybe even blocking these marginalized voices from contributing that is definitely something that we should be aware of so we can truly explore the trade-offs in a meaningful way. Thinking about trade-offs an important question to answer here is is page protection effective? And so we want to validate our intuitions about that looking at vandalism is page protection blocking vandalism but then furthermore is it promoting more substantive edits so we have a few metrics we want to look at there as well. Now thinking about the anticipated contribution of this work wikipedia is a free content that anyone can use edit and distribute what page protection does is it changes the nature of that statement it adds in effort for the sake of information quality. Now this is really valuable because it's pretty lightweight it's not requiring any sort of heavy detection models anything like that so we're wondering what page protection can serve as an example of a lightweight moderation that other platforms such as social media platforms can use. Furthermore it's interesting to think about how page protection is affecting pre-existing community barriers especially along the lines of gender when we think about what controversial articles are on wikipedia oftentimes they are about these gender-generized concepts or racialized concepts. Now furthermore we want to think about design interventions that can keep the good of page protection while mitigating unwanted adverse effects. Some suggestions we've had are making page protection more participatory so maybe including something along the lines of a digital jury or maybe incorporating a bit more standpoint theory so not only thinking about how can we do justice by the community how can we do justice by Wikipedia readership but also especially on articles that are about communities or human subjects how can we do justice by those human subjects as well. Beautiful so that's all I had for today again thank you so much for having me at the workshop and I'm happy to answer any questions at the end of this presentation.