{"id":339,"date":"2026-04-17T14:26:07","date_gmt":"2026-04-17T18:26:07","guid":{"rendered":"https:\/\/sites.comminfo.rutgers.edu\/tammari\/?page_id=339"},"modified":"2026-04-18T22:19:07","modified_gmt":"2026-04-19T02:19:07","slug":"research","status":"publish","type":"page","link":"https:\/\/sites.comminfo.rutgers.edu\/tammari\/research\/","title":{"rendered":"Research"},"content":{"rendered":"
\n

How do vulnerable and marginalized populations negotiate agency, identity, and recovery through digital technology? My work sits at the intersection of HCI, computational social science, and STS — organized around three streams derived from SAMHSA’s trauma-informed framework<\/a>: (1) empowerment principle; (2) Transparency principle; and (3) the Cultural context principle. <\/p>\n<\/div>\n

\n\n\n
01<\/span><\/td>\n\n
Empowerment principle<\/div>\n
User Empowerment: Crisis Informatics, Peer Support & Community Resilience<\/div>\n<\/td>\n<\/tr>\n<\/table>\n
\n

Agency over one’s social environment is a prerequisite for safe engagement — a condition that crises systematically destroy. This stream examines how online communities enable real-time coordination and peer support during acute crises (COVID-19, natural disasters) and chronic ones (contested illness, stigmatized health conditions, bureaucratic systems), and how platform design mediates whether this coordination empowers or further marginalizes users.<\/p>\n\n\n\n\n
\n
Long COVID epistemic community<\/div>\n

Analysis of 2.8M tweets tracing how patients collectively built the medical case for long COVID before institutional recognition, forming advocacy networks around knowledge-sharing.<\/p>\n<\/td>\n

\n
Peer roles in stigmatized health communities<\/div>\n

Studies of r\/maleinfertility and South Asian women-only Facebook groups revealing how moderators evolve into social activists and cross-posters function as boundary-spanning advocates.<\/p>\n<\/td>\n<\/tr>\n

\n
Grief bubbles & role transportability<\/div>\n

How recommendation systems trap users in recursive loops around traumatic identity — and a design principle for routing users to empowering roles rather than curated suffering. Yet platform design profoundly mediated whether this helped or harmed: algorithmic recommendations pulled some users into misinformation loops, insufficient moderation obscured the line between experiential and evidence-based knowledge, and information overload produced learned helplessness where parents disengaged rather than finding the resources they needed about COVID uncertainty.<\/p>\n<\/td>\n

\n
Disaster phase community discourse<\/div>\n

Crises vary in duration and form. Acute crises like COVID-19 and Hurricane Ida triggered real-time logistical coordination, locating scarce supplies, sharing resources among foster families, faster than formal institutions like the CDC. COVID discourse became increasingly polarized over time; earlier CDC statements were weaponized as receipts to counter new guidance, introducing data friction at critical flashpoints (e.g., new vaccine introduction). Extending this work internationally, a small set of interpretable Twitter topic features provides a strong signal explaining village-level wealth variation in Zambia, with poorer communities focused on immediate survival and wealthier ones on abstract policy.Zambian discourse itself shifted from ephemeral COVID concerns (lockdowns, testing) to durable infrastructure debates (energy plant construction, environmental effects).<\/p>\n<\/td>\n<\/tr>\n

\n
WhatsApp crisis informatics (ongoing)<\/div>\n

GTFR-funded project extending crisis informatics to end-to-end encrypted platforms used by immigrant and lower-income communities in New Jersey. Working with Lazos America Unida, the project uses participatory action research to map how marginalized Latino communities share resources and authenticate information during crises, then moves into co-design workshops where community members build system requirements from their own situated knowledge upward.<\/p>\n<\/td>\n<\/tr>\n<\/table>\n

Selected publications<\/div>\n\n\n\n\n\n\n\n\n
\n
<\/div>\n<\/td>\n
Ammari, T., Khondoker, Z., Wang, Y., & Roda, N. (2026). Beyond the Silence: How Men Navigate Infertility Through Digital Communities and Data Sharing.<\/a> CHI ’26.<\/em> CHI<\/span><\/td>\n<\/tr>\n
\n
<\/div>\n<\/td>\n
Ammari, T., Nofal, M., Naseem, M., & Mustafa, M. (2022). Moderation as Empowerment: Creating and Managing Women-Only Digital Safe Spaces.<\/a> Proc. ACM Hum.-Comput. Interact. 6, CSCW2.<\/em> CSCW<\/span><\/td>\n<\/tr>\n
\n
<\/div>\n<\/td>\n
Ammari, T. (2026). Patient-Made Knowledge Networks: Long COVID Discourse, Epistemic Injustice, and Online Community Formation.<\/a> arXiv preprint.<\/em> arXiv<\/span><\/td>\n<\/tr>\n
\n
<\/div>\n<\/td>\n
Ammari, T. (2025). Crisis Messaging Journeys: Epistemic Struggles over CDC Guidance During COVID-19.<\/a> arXiv preprint.<\/em> arXiv<\/span><\/td>\n<\/tr>\n
\n
<\/div>\n<\/td>\n
Ammari, T., Ahn, E., Lakhankar, A., & Lee, J. Y. (2025). Finding Understanding and Support: Navigating Online Communities to Share and Connect at the Intersection of Abuse and Foster Care Experiences.<\/a> Proc. ACM Hum.-Comput. Interact. 9, CSCW.<\/em> CSCW<\/span><\/td>\n<\/tr>\n
\n
<\/div>\n<\/td>\n
Randazzo, C., Kim, M., Kwestel, M., Doerfel, M. L., & Ammari, T. (2025). ‘We’re losing our neighborhoods. We’re losing our community’: A Comparative Analysis of Community Discourse in Online and Offline Public Spheres across Disaster Phases.<\/a> Proc. ACM Hum.-Comput. Interact. 9, CSCW.<\/em> CSCW<\/span><\/td>\n<\/tr>\n
\n
<\/div>\n<\/td>\n
Jung, W., Chear, C., Kim, A. H., Shah, V., & Ammari, T. (2026). Spatiotemporal Change-Points in Development Discourse: Insights from Social Media in Low-Resource Contexts.<\/a> arXiv preprint.<\/em> arXiv<\/span><\/td>\n<\/tr>\n<\/table><\/div>\n<\/div>\n
\n
\n\n\n
02<\/span><\/td>\n\n
Transparency principle<\/div>\n
Trust & Transparency: Opacity, Agency & Safety-by-Design<\/div>\n<\/td>\n<\/tr>\n<\/table>\n
\n

Unpredictable platform behavior reproduces the conditions of traumatic environments. This stream examines how opacity in algorithmic and AI systems — from content moderation to voice assistants, autonomous vehicles, and health platforms — erodes trust, strips users of informed agency, and generates manufactured consent. It also proposes transparency as a structural corrective.<\/p>\n\n\n\n\n
\n
Harassment via safety tools<\/div>\n

How RedditCares, designed to flag users in distress, was systematically weaponized as a mass-flagging harassment vector against the populations it was meant to protect.<\/p>\n<\/td>\n

\n
Older adults & AI companions<\/div>\n

Analysis of 600K+ voice assistant commands from 456 older adults reveals relational and emotional use of Alexa — and the transparency failures that result when users treat AI as a confidant.<\/p>\n<\/td>\n<\/tr>\n

\n
AI breakdowns & repair literacy<\/div>\n

Naturalistic ChatGPT logs show how students develop “use genres” and how failure — when communicated transparently — cultivates critical AI competence rather than learned helplessness.<\/p>\n<\/td>\n

\n
Autonomous vehicles & manufactured consent<\/div>\n

How opacity in AV systems normalizes data harvesting and how users construct inaccurate folk theories of system behavior when system communication is opaque — with real safety costs.<\/p>\n<\/td>\n<\/tr>\n

\n
ANCHOR project (ongoing)<\/div>\n

AI-Driven Navigation for Community Health — with Google & RWJ University Hospital — applying TID transparency requirements to clinical AI serving uninsured rural residents in NJ.<\/p>\n<\/td>\n<\/tr>\n<\/table>\n

Selected publications<\/div>\n\n\n\n\n\n
\n
<\/div>\n<\/td>\n
Perry, Y. & Ammari, T. (2026). Navigating Algorithmic Opacity: Folk Theories and User Agency in Semi-Autonomous Vehicles.<\/a> arXiv preprint.<\/em> arXiv<\/span><\/td>\n<\/tr>\n
\n
<\/div>\n<\/td>\n
Oewel, B., Ammari, T., & Brewer, R. N. (2023). Voice Assistant Use in Long-Term Care.<\/a> CUI ’23.<\/em> CUI<\/span> ⭐ Best Paper<\/span><\/td>\n<\/tr>\n
\n
<\/div>\n<\/td>\n
Ammari, T., Chen, M., Zaman, S. M. M., & Garimella, K. (2026). Learning to Live with AI: How Students Develop AI Literacy Through Naturalistic ChatGPT Interaction.<\/a> arXiv preprint.<\/em> arXiv<\/span><\/td>\n<\/tr>\n
\n
<\/div>\n<\/td>\n
Randazzo, C. & Ammari, T. (2023). “If Someone Downvoted My Posts—That’d Be the End of the World”: Designing Safer Online Spaces for Trauma Survivors.<\/a> CHI ’23.<\/em> CHI<\/span><\/td>\n<\/tr>\n<\/table><\/div>\n<\/div>\n
\n
\n\n\n
03<\/span><\/td>\n\n
Cultural context principle<\/div>\n
Social & Cultural Context: Race, Immigration & the Limits of Design<\/div>\n<\/td>\n<\/tr>\n<\/table>\n
\n

Platform designers must be sensitive to the ways race, immigration status, language, and historical marginalization shape who can safely engage. This stream examines how algorithmic systems and platform architectures reproduce structural inequities — and how marginalized communities develop distinctive adaptive practices to navigate and resist them.<\/p>\n\n\n\n\n
\n
Race, pseudonymity & online discourse<\/div>\n

Analysis of Reddit discussions of Watchmen<\/em> and Lovecraft Country<\/em> showing how pseudonymity enables both anti-racist counterpublics and colorblind masked racism with deniability.<\/p>\n<\/td>\n

\n
Korean immigrant AI adoption<\/div>\n

Two adaptive practices — pragmatic disengagement and interdependent navigation — reveal AI literacy as relational, culturally situated, and co-constructed across generations.<\/p>\n<\/td>\n<\/tr>\n

\n
Infrastructural ventriloquism<\/div>\n

A South Korean cooperative mobility platform that invoked traditional mutual-aid values while admin-only logs and opaque algorithms prevented genuine community co-ownership.<\/p>\n<\/td>\n

\n
Computational poverty mapping<\/div>\n

Twitter topic features accounting for 60%+ of village-level wealth variation in Zambia; multimodal pipelines for aid allocation in Brazzaville, extended to NJ census tracts.<\/p>\n<\/td>\n<\/tr>\n

\n
Racial bias in clinical NLP (ongoing)<\/div>\n

NLP analysis of nursing notes from a Rio de Janeiro hospital identifying linguistic indicators of racial bias and building fairness-aware algorithms for equitable clinical documentation.<\/p>\n<\/td>\n<\/tr>\n<\/table>\n

Selected publications<\/div>\n\n\n\n\n\n\n
\n
<\/div>\n<\/td>\n
Mason, S. & Ammari, T. (2026). Racism, resistance, and Reddit: how popular culture sparks online reckonings.<\/a> Information, Communication & Society.<\/em> IC&S<\/span><\/td>\n<\/tr>\n
\n
<\/div>\n<\/td>\n
Walters, A., Ammari, T., Garimella, K., & Jhaver, S. (2025). Online knowledge production in polarized political memes: The case of critical race theory.<\/a> New Media & Society 27, 9.<\/em> NMS<\/span><\/td>\n<\/tr>\n
\n
<\/div>\n<\/td>\n
Seo, J. & Ammari, T. (2026). Interdependent Navigation and Pragmatic Disengagement: How Older Korean Immigrants Selectively Engage with Digital Technologies.<\/a> arXiv preprint.<\/em> arXiv<\/span><\/td>\n<\/tr>\n
\n
<\/div>\n<\/td>\n
Seo, J. & Ammari, T. (2026). Governance, technology, and the limits of digital solidarity economies: A South Korean case study.<\/a> Internet Policy Review 15, 1.<\/em> IPR<\/span><\/td>\n<\/tr>\n
\n
<\/div>\n<\/td>\n
Jung, W., Kim, A. H., Hung, Y., Chear, C., Shah, V., & Ammari, T. (2026). Digital pulse of development: Leveraging social media discourse for poverty analysis.<\/a> Information Processing & Management 63, 7.<\/em> IP&M<\/span> 🌟 Best Paper HM<\/span><\/td>\n<\/tr>\n<\/table><\/div>\n<\/div>\n","protected":false},"excerpt":{"rendered":"

How do vulnerable and marginalized populations negotiate agency, identity, and recovery through digital technology? My work sits at the intersection of HCI, computational social science, and STS — organized around […]<\/p>\n","protected":false},"author":545,"featured_media":0,"parent":0,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"_et_pb_use_builder":"off","_et_pb_old_content":"","_et_gb_content_width":"","footnotes":""},"class_list":["post-339","page","type-page","status-publish","hentry"],"_links":{"self":[{"href":"https:\/\/sites.comminfo.rutgers.edu\/tammari\/wp-json\/wp\/v2\/pages\/339","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/sites.comminfo.rutgers.edu\/tammari\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/sites.comminfo.rutgers.edu\/tammari\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/sites.comminfo.rutgers.edu\/tammari\/wp-json\/wp\/v2\/users\/545"}],"replies":[{"embeddable":true,"href":"https:\/\/sites.comminfo.rutgers.edu\/tammari\/wp-json\/wp\/v2\/comments?post=339"}],"version-history":[{"count":19,"href":"https:\/\/sites.comminfo.rutgers.edu\/tammari\/wp-json\/wp\/v2\/pages\/339\/revisions"}],"predecessor-version":[{"id":396,"href":"https:\/\/sites.comminfo.rutgers.edu\/tammari\/wp-json\/wp\/v2\/pages\/339\/revisions\/396"}],"wp:attachment":[{"href":"https:\/\/sites.comminfo.rutgers.edu\/tammari\/wp-json\/wp\/v2\/media?parent=339"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}