"Facebook Promotes More Harassment": Social Media Ecosystem, Skill and Marginalized Hijra Identity in Bangladesh
Fayika Farhat Nova, Michael Ann Devito, Pratyasha Saha, Kazi Shohanur Rashid, Shashwata Roy Turzo, Sadia Afrin, Shion Guha
1157"Facebook Promotes
More
Harassment": Social MediaEcosystem, Skill and Marginalized Hijra Identity inBangladesh
FAYIKA FARHAT NOVA,
Marquette University
MICHAEL ANN DEVITO,
Northwestern University
PRATYASHA SAHA,
University of Dhaka
KAZI SHOHANUR RASHID,
University of Liberal Arts Bangladesh
SHASHWATA ROY TURZO,
East West University
SADIA AFRIN,
Daffodil International University
SHION GUHA,
Marquette UniversitySocial interaction across multiple online platforms is a challenge for gender and sexual minorities (GSM) dueto the stigmatization they face, which increases the complexity of their self-presentation decisions. Theseonline interactions and identity disclosures can be more complicated for GSM in non-Western contexts due toconsequentially different audiences and perceived affordances by the users, and limited baseline understandingof the conflation of these two with local norms and the opportunities they practically represent. Using focusgroup discussions and semi-structured interviews, we engaged with 61
Hijra individuals from Bangladesh, aseverely stigmatized GSM from south Asia, to understand their overall online participation and disclosurebehaviors through the lens of personal social media ecosystems. We find that along with platform audiences,affordances, and norms, participant skill/knowledge and cultural influences also impact navigation throughmultiple platforms, resulting in differential benefits from privacy features. This impacts how Hijra perceiveonline spaces, and shape their self-presentation and disclosure behaviors over time.
Content Warning:
This paper discusses graphic contents (e.g. rape and sexual harassment) related to Hijra.CCS Concepts: •
Human-centered computing → Empirical studies in collaborative and social com-puting .Additional Key Words and Phrases: Social media ecosystem; Hijra; LGBTQ+; Skill; Gender identity; Audiencemanagement; Self-presentation
ACM Reference Format:
Fayika Farhat Nova, Michael Ann DeVito, Pratyasha Saha, Kazi Shohanur Rashid, Shashwata Roy Turzo,Sadia Afrin, and Shion Guha. 2021. "Facebook Promotes
More
Harassment": Social Media Ecosystem, Skilland Marginalized Hijra Identity in Bangladesh.
Proc. ACM Hum.-Comput. Interact.
5, CSCW1, Article 157(April 2021), 35 pages. https://doi.org/10.1145/3449231
Authors’ addresses: Fayika Farhat Nova, Marquette University, [email protected]; Michael Ann DeVito,Northwestern University, [email protected]; Pratyasha Saha, University of Dhaka, [email protected]; Kazi Shohanur Rashid, University of Liberal Arts Bangladesh, [email protected]; Shashwata Roy Turzo,East West University, [email protected]; Sadia Afrin, Daffodil International University, [email protected];Shion Guha, Marquette University, [email protected] to make digital or hard copies of all or part of this work for personal or classroom use is granted without feeprovided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and thefull citation on the first page. Copyrights for components of this work owned by others than the author(s) must be honored.Abstracting with credit is permitted. To copy otherwise, or republish, to post on servers or to redistribute to lists, requiresprior specific permission and/or a fee. Request permissions from [email protected].© 2018 Copyright held by the owner/author(s). Publication rights licensed to ACM.2573-0142/2021/4-ART157 $15.00https://doi.org/10.1145/3449231Proc. ACM Hum.-Comput. Interact., Vol. 5, No. CSCW1, Article 157. Publication date: April 2021. a r X i v : . [ c s . H C ] F e b They raped me. . . multiple times. When I tried to file complaint against the rapistsin the police station, the police officers raped me again. They said they were doingme a favor, as I would never get a respected man because of my identity. I was stuckwithin those vicious circle of sexual harassment until finally, I decided to share theseexperiences on social media expecting to be heard and saved (P2, 20)This is the lived experience of P2, who is a Hijra from Bangladesh and a participant in this study.Hijra, who are widely referred to as "third gender" [69] individuals, are a group of people in southAsia who do not conform to binary notions of male or female gender but rather combine or movebetween them [69] . Hijra are stigmatized and excluded from the society because of their perceivedgender identities [62], and many Hijra, like P2, experience extreme social exclusion, discrimination,harassment, and violence, with little or no access to physical, mental or social support. Instead,Hijra turn to social media for self expression, and social support.Social media can play an amplified role for stigmatized populations, especially those with littleaccess to physical assistance, including LGBTQ+ communities [26, 56, 116]. For such communities,social media acts as a primary space for identity exploration and development [26, 116], a primarysource of social support and justice against harassment [12, 13, 93, 94], a resource for combatingstigmatization around mental health [22], and both a guide and public platform for experiences suchas gender transition [56]. All of these happen across a multi-platform social media ecosystem, withindividuals moving their content and attention between multiple platforms based on audiences,affordances, and their perceptions of the overall "spaces" available on a platform [41]. Prior work[41] has broken significant ground by exploring the importance of social media to members ofgender and sexual minorities (GSM) in a Western, mostly US context. However, in order to improveand include social media platforms for everyone, we must better understand the need for anduse of these platforms by stigmatized users in markedly different cultural contexts. Similarly, aswe continue to improve the field’s understanding and treatment of gender in platform design,it is essential to account for non-Western conceptions of gender and the needs and behavior ofnon-Western gender minority groups, such as Hijra, that may not necessarily align with the culturalor practical realities of Western GSM individuals.To address these concerns, we engaged with Hijra populations from Bangladesh and explored howGSM from non-Western contexts participate and self-present on different social media platforms,using DeVito et al.’s personal social media ecosystem framework for LGBTQ+ populations as atheoretical lens of inquiry [41]. We find: • Hijra primarily rely on social media platforms for three reasons: (a) communication withfamily, (b) Hijra community participation, and (c) sex work. Depending on each of thesepurposes, Hijra share content to targeted online audiences, as motivated by the platform’safforded levels of presentation flexibility and visibility control. • Technical knowledge and skill is a major factor in enabling Hijra to navigate social mediaplatforms, with widespread lack of skill negatively impacting the way Hijra perceive platformaffordances. Skill, when added to the input and influence of local authority figures (suchas Hijra matriarchal leaders known as Gurumas), also motivates shifts in content acrosspersonal ecosystems and/or limitation of social media use. • A reliance on Western cultural signifiers in designing platform features and navigation aidslessens the utility of social media for Hijra. Advanced and continually updating platform As all participants in this study referred to themselves as Hijra, we will use the term "Hijra" to denote individual(s) fromHijra community, and "hijra" to refer their gender.Proc. ACM Hum.-Comput. Interact., Vol. 5, No. CSCW1, Article 157. Publication date: April 2021. privacy features do not necessarily provide a sense of safety or practical benefit to Hijra ifthose features and their signifiers are not culturally familiar to them.Whereas in prior work, GSM communities like LGBTQ+ users’ online self-presentation and par-ticipation were analyzed through their audiences [41], platform affordances and the usability ofthe space, our paper adds to this conversation by finding that for GSM in non-Western contexts,this framework does not fully work. As the platforms’ intended affordances to its users are notalways aligned with Hijra’s understanding of the platforms, for reasons like limited platformknowledge/skill or less culturally appropriated platform design, existing framework is unable toaccurately explain how Hijra self-present themselves online with regards to their audiences andspaces. Previous literature has emphasized the importance of digital literacy within vulnerablecommunities in terms of their social media participation [96, 98, 103]; however, such understandingis absent in the case of GSM from non-Western contexts.Therefore, our work makes several contributions to the CSCW community: 1) We extend and im-prove the current social media ecosystem framework [41] by introducing and integrating technicalknowledge and skill set in the framework based on the observations from Hijra community, 2) Wecontextualize the presence of community and cultural influence within Hijra groups, which helpsus to better understand how GSM from non-Western contexts come to trust and prefer certainonline platforms for their self-presentations and participation, 3) We advocate for design practicesin HCI that integrate cultural context and marginalized views in the design phase to build moreaccurate, more inclusive social media environments for stigmatized GSM from non-Western context.Whereas existing work in ICTD discuss such inclusion and design practices from developing context[73, 83, 95, 113], our study contributes to the conversation by including GSM populations like Hijrain HCI and CSCW.
There is a growing scholarly recognition of the experience and diversity of sexual and genderorientations beyond binary gender and heterosexual identities [23]. Recent work in social computinghas explored the benefits, pitfalls, and design opportunities around social media for GSM identitiesin a mostly US context [6, 58, 76, 106]. Similarly, researchers have begun to seriously grapple withthe impacts of our concepts of gender on AI-based applications such as facial recognition in amostly-western context [67, 107]. However, while these studies move us forward significantly,they ultimately categorize gender and sexuality through a strictly Western lens [7, 23]. One sidedperspectives can cause discrimination by disregarding the individual and cultural differences thatthese GSM populations experience throughout their lives. Due to the stigmatization and exclusionHijra experience [105], their economic backwardness [60], their unique non-Western hijra identity,and their location in South Asia, Hijra are a crucial population to represent in order to broaden ourunderstanding of GSM social media use in a non-western context.As Hijra may be unfamiliar to many CSCW readers, we will first briefly illustrate the contextin which Hijra use and rely on social media. We will then review specific concerns around self-presentation and digital literacy, and the overall personal social media ecosystem lens.
Hijra is an institutionalized third gender role that is neither male nor female, but contains elementsof both [87]. Similar to Hijra, there are other gender minority identities that historically exist inmany cultural contexts, known as bakla in the Philippines, xaniths in Oman, serrers among thePokot people of Kenya, and
Hijra, jogappas, jogtas, or shiv-shaktis in South Asian countries suchas, India, Bangladesh and Pakistan [7, 69]. Importantly, hijra is not simply a South Asian versionof the Western concept of non-binary identity, a common misconception [7]. Hijra mostly live in Proc. ACM Hum.-Comput. Interact., Vol. 5, No. CSCW1, Article 157. Publication date: April 2021. segregated housing communes, where unwanted intersex or trans children are raised in a saferenvironment [7]. Hijra identity includes traditional procedures and distinct commitments uniqueto this form of gender minority identity, such as the time-honored ritual of leaving one’s home —or being forced out — and undergoing induction into a clan of Hijra led by an elder matriarchalindividual known as a "nayak" or
Guruma , at which point the new inductee is known as a "chela"or follower [20]. Gurumas and chelas have their own codes of conduct, and they often speak inFarsi-inflected variations of their local language known
Ulti to ensure secrecy [20].Government estimates say there are around 10,000 Hijra in Bangladesh, although the "BadhanHijra Songha", a transgender-Hijra rights group, states that the figure is actually around 100,000[117]. Even though Hijra are legally recognized in Bangladesh, they are still socially excluded [69],with the word "Hijra" commonly used to mockingly refer to undesirable digression from normativemasculinity [62]. Despite formal recognition, this lingering lack of acceptance for gender identitiesbeyond the binary results in limited employment opportunities for Hijra, many of whom turn tosex work or begging [71]. Cisgender members of society often act to enforce the lower social statusof Hijra, preventing them from accessing social institutions, resources, and services [62, 69]. Hijraalso face abnormally high rates of hate crimes involving rape, harassment and physical abuse [52].Due to Hijra’s vulnerable social status and exclusions from the society, several non-governmentalorganizations (NGOs) try to support these communities for issues like sexual health, educations,job opportunities and so on [15]. NGOs such as Badhan Hijra Sangha are working hard to raiseawareness of general and sexual well being among Hijra community; to create basic human, civic,and social rights for them and establish network with national, international GO/NGO /institutions[1]. They are also providing legal supports and skill development training for alternative livelihoodoptions for Hijra to reduce economical discrimination [1]. Additionally, Sachetan Somajsheba HijraSangha, Bondhu Social welfare society, Hasab, Shomporker Naya Shetu, iccddr,b etc. are someof the notable NGOs in Bangladesh who have taken many initiatives for Hijra communities toestablish group dynamics and advocacy with different level of stakeholders [102].Solidarity among Hijra is quite strong because they as a group are discriminated and excludedfrom the rest of society because of their specific cultural conventions and group norms that are notaccepted by the normative populations [69]. Even though dimensions of their social deprivationand harassment have never received attention in development sectors [69], it is time to changethat notion in the CSCW research community, as we aim to reflect on our practices and design tobe more inclusive and flexible towards all genders and sexualities.
Impression management involves the processes by which people control how they are perceivedby others [75]. Social media plays an important role in how this impression or self-presentation isconstructed for most users [35]. However, particularly for GSM, this management of impressionbecomes more critical, as it provides them opportunities to experiment with their self presentationand identity to the the world [43, 74]. Facebook and other social networking sites have beena major area of research, particularly to understand what practices and behaviors users adoptduring their self-presentation online [28, 66, 112]. Previous research has highlighted how socialmedia can facilitate and assist LGBTQ+ users from Western context during self-disclosure throughconstructing, managing, and expressing identity projections [18]. [34] investigates the use ofFacebook by LGBTQ+ users to understand how identity construction, management and negotiation,and activism take place within these communities in a rural American social setting. The existingwork on LGBTQ+ users identifies a variety of identity management strategies that these populationsadopt on social media including monitoring their online self-expression, using privacy and securitycontrols, strategically managing their audiences and so on in Western setting [34, 43, 82].
Proc. ACM Hum.-Comput. Interact., Vol. 5, No. CSCW1, Article 157. Publication date: April 2021.
Fig. 1. Tin Tali, traditional clap to demand attention or recognition by Hijra (Online photo)
This process of constructing self-presentation decisions on distinct social media platformsbecomes complicated due to LGBTQ+ users’ complex and stigmatized identity [40]. Accordingto [34], LGBTQ+ users struggle with their use of Facebook, as many of them are not open abouttheir sexuality and gender identity and prefer to keep it as a secret on the online platforms byconstructing a different, more acceptable identity than their actual self. While the vast majorityof work on LGBTQ+ has been done from Western perspectives, there is some existing work thattry to explore such GSM communities from a global South context. Studies like [16, 47] havefocused on queer and Hijra population from India to explore their realities in terms of social,economic, political, emotional, psychological, and legal issues. Indian LGBTQ+’s adoption of emaillists, message boards, and weblogs to communicate with each other online have also been exploredby researchers [72, 85, 86]. However, few studies seem to have focused upon the possible use ofpopular social networking platforms from Global South. Literature like [42] have added to thatconversation by studying how Indian LGBTQ+ individuals create multiple identity on distinctsocial media, such as Facebook, to protect their gender minor identities from unwanted audiences.However, such exploration of GSM communities from developing context is understudied and candiffer from how Hijra communities adapt to those online practices [82].Self-presentation and impression management is a crucial part of life for Hijra due to theirstigmatized identity [108]. In Bangladesh, where often talk on sex education is considered taboo[37], figuring out one’s own gender and sexuality becomes a struggle, especially for the low-literatecommunity such as Hijra [10]. Some Hijra outwardly acknowledge that they are women throughsexual relationships, gender roles, and clothing, but are aware that they are separate and distinctfrom cisgender women [63]. Different expectations from the gender roles, biological sex, and sexualorientation within Hijra can make it harder for them to cultivate their identity in the society [49].As individuals shape their behavior for certain audiences in specific contexts [53], many Hijrasimilarly continue to intentionally and surgically construct their self-presentation decisions fortargeted audiences. Prior work on Hijra identity suggests that it is useful to view Hijra impressionmanagement through Goffman’s dramaturgical model of self-presentation [31, 42, 53]. Goffman’sdramaturgical model posits that often individuals prefer to represent themselves differently to safe"backstage" audience regions and potentially-threatening "frontstage" audience regions [31, 53]. It ispurely contextual and up to the individual’s need to decide which behavior or identity becomes the
Proc. ACM Hum.-Comput. Interact., Vol. 5, No. CSCW1, Article 157. Publication date: April 2021. front stage. Many Hijra are forced to put on aggressive and hostile front on the streets as a defensemechanism to fight street harassment [60, 80, 108] even though in personal lives within theirown community, they may have completely different personalities. Hijras can also be aggressive,especially when not handed money as they wend their way through traffic, begging as a primarymeans to earn money [81] in a discriminatory society. Often these rude remarks are followed by tin tali (a specific way of clapping by Hijra demanding for recognition of their existence [20]
Fig.1 ) and lifting up sarees showing their castrated genitals or breasts in public [69]. This rude andsexualized presentation of Hijra in frontstage regions is a product of continuous repression andcriticism they face in daily lives and can be very disconnected from their presentation in private,backstage regions. In direct contrast to presentation of Hijra in frontstage regions, their presentationin safe "backstage" regions is generally nicer, calmer and very friendly [108]. Previous researchhas discussed this notion of front and backstage by exploring how individuals with stereotypedidentities actively shape their online social identity through different self-presentation or disclosurestrategies [55]. They often apply approaches of strategic outness, first assessing a specific socialsituation before determining whether to disclose their gender or sexuality online [43]. As disclosureor exploration of stigmatized identity on social media platforms can often lead to negative outcomessuch as hate speech, name calling, and unsolicited sexual photos and messages [26, 46, 54, 59], itcan effectively force users to restrict their online participation [93, 114] or employ heavily tailoredprivacy settings to manage visibility of their gender and sexuality related content within a singleplatform [41].
One potentially major complicating factor around self-presentation behavior on social mediaplatforms is lack of digital literacy or skill, as it enables users to choose effective technologypractices to meet their goals [96]. As the self-presentation goals of Hijra can be complex and mayrequire use of advanced privacy features, lack of skill and knowledge of the platform features canbe a major challenge for such stigmatized users [61]. According to previous research, privacy andsecurity settings of social media platforms can often be difficult to navigate and imperfect in termsof user’s requirement of information control [82]. For instance [17] discusses how privacy literacymay change online behavior and perceived online safety within users; they define online privacyliteracy as users’ knowledge about technical aspects of online data protection, and ability to applythose strategies for own privacy regulation. Another previous study [98] has explored how differentlevels of privacy knowledge and skills among African American younger adults can effect theironline practices. [98] has identified how populations with different level of skill and particular areconsistently left out from benefits of technology because they cannot access the full potentials ofthe technology.As an explanation for certain communities’ limited skill to navigate online platforms, somestudies suggested a combination of imperfect interface design and a lack of Internet literacy [17, 78].[96] also shows low online privacy literacy within the participants in terms of technical familiarityand policy knowledge. Technical skill and knowledge are often barriers for older populations wholack the understanding of rigorous privacy settings and features provided by the platform [115].It has been also found that lower internet skill within women has consequently increased thegender gap on online spaces [61].Whereas these studies looked at the impact of technical skill andknowledge on populations from Western context, several studies found that it is more commonwithin non-white users to fall behind in online privacy control behavior due to their limited skillset [97] and knowledge [96]. Even though previous research on digital literacy have not exploredHijra or other GSM’s skill and knowledge of using online social media platforms from global Southcontext, [103] talks about how less technical skill within women in south Asia impact and limit
Proc. ACM Hum.-Comput. Interact., Vol. 5, No. CSCW1, Article 157. Publication date: April 2021. their technology use. Such exploration of different communities around the world directs ourattention towards understanding the concept of digital literacy and knowledge within GSMs fromnon-Western context, such as Hijra. Due to Hijra’s stigmatized identity, it is absolutely essential forthem to utilize different platform affordances to ensure privacy and control over self-presentationsettings, and thus, this study tries to fill that void of knowledge by exploring Hijra communitiesfrom Bangladesh.
In prior work researchers have conceptualized how GSM self-presentation, existing in across multi-platform ecosystem, has allowed differential presentation across different audiences and sites [41].Specifically, Devito et al. [41] posited three specific elements of social media ecosystems that driveself-presentation behavior (and content) to appropriate outlets: audiences, affordances, and theconflation of the two with local norms, which they call "spaces".
Audience . DeVito et al. [41] found that different audience compositions per platform and, im-portantly, user perceptions of these audiences were a key motivator for personal social mediaecosystem use and movement of content across said ecosystems [41]. LGBTQ+ users generallyconceptualized their audiences as either abstract (relatively unknown) or targeted (specific peoplewho are the potential connections users may have and want to share contents with through theirsocial media platforms). Users imagine their audience based on factors ranging from goals andindividual psychological expectations from others, allowing them to act self-protectively despiterarely having access to precise audience composition information.
Affordances . According to DeVito et al. [41], user perception of a platform and its appropriate placein one’s personal social media ecosystem is heavily affected by the affordances, or possibilities foraction, each platform offers to users . Stigmatized users explore and look beyond single platforms,considering the range of affordances available across their personal ecosystem when makingself-presentation decisions.
Space . DeVito et al. [41] use the term "space" to describe the conflation of platform and audienceby users making self-presentation decisions. Through the lens of this conflation, users form aconcept of what "type" of platforms are available, what they are for, and who is welcome there.By examining the conflated spaces, we are able to look not just at the social context intended byplatform designers, but rather what a platform represents to the users themselves in comparison toother platforms within one’s personal social media ecosystem.DeVito et. al’s framework of personal social media ecosystem [41] provides a solid theoreticalfoundation for research into Hijra self-presentation and social media platform use. We draw fromthis theoretical perspective to frame our research questions and also as the qualitative lens for ouranalysis. However, although the existing framework provides valuable perspectives on stigmatizedGSM populations in online spaces, it was developed entirely in a Western context and may needextension to apply to Hijra and other stigmatize populations from developing, non-Western contexts.To better account for Hijra and other non-Western gender minorities in the design of the socialmedia platforms they rely on for crucial services and communication, it is imperative to extendthis lens beyond a Western context. As such, using DeVito et al.’s personal social media ecosystemsas a guide, we ask:
RQ1.
What are the audience related concerns that Hijra have in their social media ecosystem?
RQ2.
How do they manage their audiences through the affordances of different social mediaplatforms?
RQ3.
How do audience and afforances influence Hijra to move around in different social mediaplatforms?
Proc. ACM Hum.-Comput. Interact., Vol. 5, No. CSCW1, Article 157. Publication date: April 2021.
Fig. 2. Field Areas for Data Collection in Dhaka
To answer these questions, we conducted a six-month-long qualitative study of Hijra in Dhaka,Bangladesh from March-August 2019. The study was conducted in 6 neighborhoods of Dhaka (seeFig. 2) : Lalmatia, Kakrail, Mugdapara, Gulistan, Manda, and Kamalapur. During this period, webuilt trust and encouraged our participants to participate in the study by spending ample time intheir slums to meet with participants and built rapport. We employed multiple elicitation methods,including semi structured interviews, Focus Group Discussions (FGDs) [90] and unstructured onlineobservations [91] of public and private (with consent) content related to Hijra. To ensure a robustunderstanding of how Hijra interact with social media, we triangulated these multiple data sourcesin our analysis. All study procedures were approved by the Institutional Review Board at the leadauthor’s academic institution, a private research university in the midwestern, United States. Participant recruitment was performed through a snowball-style iterative process of networkingand trust building in Hijra communities. The third author, leading efforts on the ground, firstconducted 3 preliminary semi-structured interviews with acquaintances who belong to Hijracommunity. From there, we were approved to conduct our first FGD consisting of 6 Hijra in a dormat Lalmatia, where Hijras from different districts were gathered for a cultural event. Our secondFGD also consisted of 6 participants, all Hijra sex workers who were visiting Dhaka for a trainingprogram arranged by a local NGO. We arranged a meeting with them with the help of
SachetanSomajsheba Hijra Sangha , a non-profit organization working for the welfare of Hijra community.The remaining 3 FGDs were conducted with local Hijra in a place they considered both convenientand safe.Though the in-person activities in this study were conducted in Dhaka, we ensured representa-tion from participants all over the country. We talked to Hijra communities from 8 divisions forthe FGDs and individual interviews: Barisal(3 participants), Chittagong(4), Dhaka(37), Khulna(3),Mymensingh(3), Rajshahi(5), Rangpur(3), and Sylhet(3). In total we had 61 participants (45 from Here, slum refers to the colloquial reference in South Asian English to lower quality of informal housing without anyderogatory connotation.Proc. ACM Hum.-Comput. Interact., Vol. 5, No. CSCW1, Article 157. Publication date: April 2021.
Demographics Percentage
Sex By Birth Male 93%Intersex 7%Sex Assigned at Birth Male 100%Gender hijra 100%Preferred Hijra Identity Hijra 47.6%Third Gender 21.3%Women 31.1%Sexuality Gay 100%Age Range 18-25 yrs 65%26-33 yrs 22%yrs > 33 13%Highest Level of Education No Education 38%Primary Education 24%Some High School 17%High School Diploma 13%Bachelors Degree 8%Location From Dhaka 87.43%
Table 1. Demographics of the Participants
FGDs, 16 from one-one interviews).
Table 1 shows for additional demographic information. Almostall participants were employed doing "Hijragiri" (traditionally, the ritual of badhai , or blessingsconferred on a newborn through dancing and singing) [102], the collection of cholla (tolls fromjurisdictions), training to become skilled in the Ulti language [60], and sex work. Out of 61 par-ticipants, 19 were directly involved with sex work, and only 3 participants were involved withwhite-collar jobs. The socio-economic status of all the participants was relatively low, as Hijracommunity in Bangladesh conventionally consists of those from lower economic levels [64].
Data collection employed 5 Focus Group Discussions (FGDs), 16 one-on-one interviews, and onlineobservations on participants’ social media participation with their prior consent. Interviews andFGD data were collected in the form of field notes and audio recordings. We collected photos,screenshots, videos from our online observations. All the interviews and FGDs were conducted inBengali, and each participant in the focus group and interview was compensated with BDT 400Tk,which is roughly around $5 and more than the minimum daily wage (BDT 50TK) in Bangladesh[2]. In all data collection activities, we focused on the social media practices of Hijra communities,with specific attention paid to audience management strategies and social media participation.Our initial data collection strategy centered around the 5 FGDs, each of which was 2.5 hours long.However, we noticed that often the most senior members of the group (such as Guruma) respondedto our questions or took control of the room, as opposed to all the members being involved in thediscussion. To balance this, we decided to conduct the one-on-one interviews with follower Hijrawho live under Gurumas. The 16 one-to-one interviews were conducted with the help of a Gurumawho is also a Hijra activist, and helped us reach more people to talk to individually. The timeand date for these interviews were chosen according to the participants’ preference. Interviewsaveraged 40 minutes each.Our interview and FGD protocols consisted of 24 sets of questions, with multiple sub-questionsunder each of them. Answering every question was not compulsory, and the participants could
Proc. ACM Hum.-Comput. Interact., Vol. 5, No. CSCW1, Article 157. Publication date: April 2021. skip questions if they wished. Later, the interviews were translated and transcribed by the 4th, 5thand 6th authors of this study, whose first language is Bengali.The online observations were collected through exploring different social media platforms thatHijra mentioned in FGDs and interviews, such as Facebook and Bigo Live. While looking at thesocial media information provided by our participants, we also used different keywords suggestedby our participants, such as
Hijra , Third Gender etc., to find online groups and communities thatpromote anti- or supportive posts related to Hijra. Keeping ethical implications in mind, we onlycollected information on groups and pages through key words that were already explicitly publicfor everyone on social media platforms [30, 44] and were not only accessible to only a certaincommunity of users. As suggested by [122], it is unethical for researchers to use any personalinformation from social media if the data or information is restricted to a certain group of peopleor communities. Hence, we ensured the Facebook groups/pages we shared image from are publicand not restricted to certain communities or populations. [100] also adds to this conversation ofethical implications by emphasizing that more than expected information should not be revealedthrough images’ background or through the combination of visual and textual elements. To ensurethat, we intentionally refrained ourselves from publishing or revealing any post or personal profilethat were shared in those groups even though they were public.The FGDs and interviews were conducted in a secured setting, giving priority to participants’preference of time and place. All data were recorded only after the consent of participants, andnames used in the writing of this paper are pseudonyms. Observational data from the participantsocial media account were only recorded with proper consent, and care was taken not to recordparticipant names, profiles, or any kind of identifiable information.
Working with the Bangladeshi Hijra community was a complexprocess. Even though this study includes a high number of participants to understand Hijra’s socialmedia ecosystem, there were few challenges that our authors faced during data collection. Thefirst challenge was to gain access to Hijra participants. Being largely disconnected from the civilsociety, Hijra are very protective towards their own community and hesitate to give outsidersaccess. It was even challenging to even commute to these communities for our author, as rickshawpullers did not want to provide a trip to such locations, and uncooperative locals sometimes passedharassing comments to the author upon seeing her entering their places alone. Wishing to avoidthe known problem of fake street Hijra [68], we initially intended to work through communityorganizations or NGOs to find participants. However, these organizations often did not ultimatelycooperate due to privacy and funding issues. The second challenge was in actual data collection.Even though our author contacted Hijra through different organizations or NGOs, not everyonewere interested in participating in the study. However, this number of non-participation was nothigh. Only 3 out of 64 Hijra we initially contacted declined to participate in the study becauseof their discomfort in sharing personal information on their gender, sexuality and social mediaparticipation. We respected their decision and only collected data from the rest 61 participantswho were interested in the study. Additionally, 3 out of those 61 participants had some workconflicts where they received calls from their clients during FGDs and had to leave. In such cases,the data collection from those specific participants was stopped by the author with permission.We still included them in our dataset due to the depth and value of the data they provided andtheir sincere interest to be parts of the study. However, Hijra were not the primary complicationwe had to face during data collection; the middlemen who worked with us in the field were. Aswe had to go through NGOs to connect with the Hijra communities, sometimes they providedmiddle-men from their organization to coordinate and manage Hijra in the field (such as callingthem one by one for the interview). Even though these individuals were sent as volunteers to
Proc. ACM Hum.-Comput. Interact., Vol. 5, No. CSCW1, Article 157. Publication date: April 2021. assist our ethnographer in the field, without NGOs knowledge, these middlemen kept asking for alarge amount of money as a bribe from our author with subtle threats of discontinuing the studyotherwise. In a developing country like Bangladesh, corruption in the form of bribes is a hugeconcern [38]. Despite such obstacles, we gave our best effort to continue the research and collecteddata from a large number of samples as majority of our participants were comfortable with thestudy and did not mind sharing their data. Finally, this project involved close observation of Hijra’sday-to-day life while staying with them more than 6 hours a day. Due to the disturbing natureof some aspects of the day-to-day treatment of Hijra by society, team members found some partsof data collection personally upsetting, requiring later treatment by a therapist. We enumeratethese issues in this paper as a guide for other researchers seeking to conduct similar research inchallenging settings.
After data collection, we anonymized the data before analysis. To anonymize, the first and thirdauthor renamed all the FGD participants as P1, P2... etc., and for interviews, as X1, X2... etc. As ourdata was mostly qualitative, we used a grounded, thematic approach [120] on our collected data.For each source of data, the first author created codebooks following an open-coding approach toallow flexibility for new themes to emerge. We wanted to understand the rationale behind Hijraperceptions on gender identity, and online participation. The codebook was created through severaliterative rounds of coding until we reached theoretical saturation [51]. The first author shared thecodebook with other members of the team in each iteration and upon thorough discussion, theteam reached agreement on the generated codes. The categories formed from the codes were latergrouped into different themes which helped us construct our findings from this study. Althoughthe FGDs provided us initial themes and observations on Hijra’s group dynamics and overallexperiences in their day-to-day and online lives, the interviews provided us more specific andnuanced information on Hijra. From the interviews, we drew deeper detail on their struggles,confusion, and frustration both in offline and online world, both confirming and expanding ourinitial FGD-based themes. The online observations helped us to provide visual references to thereaders and connect the experiences of Hijra (that they shared through FGDs and interviews) withpractical instances.
Our study was a team effort, with varying specialties and expertise within the research team. Eventhough we set out with the intention to include a Hijra representative in our research team, therewas a lack of interest from the Hijra community. Due to Hijra’s stigmatized social status, oftenthese communities are supervised by different NGOs who regulate Hijra’s well-being and socialparticipation. This results in a de facto degree of social control between NGO and Hijra, whichaffects our access to these participants. Without NGO’s permission, no Hijra can directly participatein any study or research with any outside researchers due to possible conflicts or complexities thatmay arise, both political and social, hampering Hijra’s safety. Therefore, even though the activistGuruma assisted us in the paper, to ensure safety and anonymity, that Guruma did not want to bean author in our paper. In an effort to ensure a balanced expertise in the research team, our authorgroup consists of cis and trans authors who work with GSM participants from both Global Northand South. Three authors have previously worked with stigmatized and vulnerable populations,including Hijra, children and women, through prior engagement with different NGOs, one authoris a transgender member-researcher for Western LGBTQ+ populations, and the remaining threeteam members have formal training on transcribing transcripts for academic research. Except thesecond author of this paper, who is American-born, the authors have native proficiency in speaking,
Proc. ACM Hum.-Comput. Interact., Vol. 5, No. CSCW1, Article 157. Publication date: April 2021.
Fig. 3. Hijra gender and identity classifications writing, and understanding the Bengali language that Hijra used in this study; grew up in SouthAsia and are familiar with local cultural contexts. The third author of this study was responsiblefor collecting data and has been familiar with the relevant neighborhoods for more than 10 years.
We report our results by identifying what audience related concerns Hijra have on social media,explaining their connections with platform affordances and skill, and describing the contexts inwhich they shift across social media ecosystem. Exploring these results will provide us betterunderstanding on Hijra’s online self-presentation and participation.Our results indicate that Hijra have complex gender and identity constructions. All of ourparticipants mentioned their gender to be "hijra" in both interviews and FGDs. However, suchconstruction of gender gets complicated when some of our participants mentioned classifying hijrafurther into
Meye hijra (for intersex) and
Chele hijra (for trans females), which is an internal genderclassification some Hijra follow and not officially addressed by any legislation. Even though wedid not have any participants who were proponents of this classification, it exists in some Hijracommunities. Our data also suggests that despite mentioning gender to be "hijra", not all of ourparticipants were happy with this legal term to explain their Hijra identity. As the term symbolizeshistorical exclusion of Hijra from the mainstream society, 13 out of 61 Hijra (around 21%) participantsin our study preferred to identify themselves as "Third Gender" than "hijra". Additionally, 19 out of61 participants (around 31%) mentioned their desire to be perceived as women although they do notfall under the traditional definition of cis-gender female. They experimented with their feminineidentity through cross-dressing as traditional Bengali women while growing up (wearing saree,make up etc.). For a better understanding of our readers, we have created a chart (see
Fig. 3 ) thatrepresents how Hijra classify their identity in Bangladesh.Due to such complicated identity constructions, Hijra extensively employ their personal socialmedia ecosystems to meet a number of different goals, as motivated by perceptions of platformaudience, affordances, and spaces. However, our results also strongly indicate that local culturaland community influences, such as authority figures or group dynamics, as well as technicalskill/knowledge also motivate the movement of activity and content across one’s ecosystem, evensometimes overriding the importance of affordances. In this section, we will specifically focuson the intersection between Hijra’s audience concerns and perceived affordances. Then, we willlook into online participation shifting behaviors within Hijra to paint a holistic picture of genderminority self-presentation within one’s personal social media ecosystem in a non-Western context.
Proc. ACM Hum.-Comput. Interact., Vol. 5, No. CSCW1, Article 157. Publication date: April 2021.
RQ1 asked what audience-related concerns Hijra consider when using social media, each reflectinga core reason for using social media. Our data suggest that Hijra in Bangladesh have three typesof primary audiences on social media, with conflicting sets of disclosure concerns: family, otherHijra, and cisgender men. Uses ranged from daily communication to online sex work. Each of theseaudiences not only helped our participants form their personal social media ecosystem, but alsoshaped their online behaviors and concerns that help them to construct their identity online.
Our data indicates that for 37 out of 61 participants (around60%), one of the primary audiences of interest are their existing family and friends. While familyand friends are a common audience for social media content generally, the requirement that Hijralive away from their families in separate communities [62], as well as persistent stigmatizationby larger society [15], heightens the importance of social media for reaching this audience. Forexample, interviewee X4(Age 18) mentioned:I live away from my family...we are not accepted in the society...and they (mass popu-lation) even leave the place if we sit beside them in public transportation...(which iswhy) it is easier for me to be connected with my family and friends in FacebookSocial media platforms - and for Hijra, especially Facebook - have made it easier and more convenientfor our participants remain connected with those they are close to while also respecting thestructures of being a Hijra and avoiding public humiliation in the physical realm. However, thisconvenience does not come without complications. Despite the obvious utility of Facebook forHijra, it can also create new anxieties related to heightened concerns over managing and selectivelydisclosing gender identity online. The need to be connected with family via Facebook can collidewith the discomfort or serious disclosure concerns of the many Hijra, who then have to hide aportion of their online participation from their families. As interviewee X11(30) mentioned:One of the biggest things in my life is that my family doesn’t know that I am Hijra...I have to do everything, specially in Facebook, by hiding my own identity...often itbecomes very hardSimilarly, participants like X8 and X10 also mentioned putting in extra effort to keep the platformthey use for family connections walled off from the rest of their online life, such as strictly neverusing their meye nam (female names as Hijra adopt as hijra) or not adding anyone from the Hijracommunity. As X8(25) mentioned mentioned,I use male as my gender (on social media)...(although) I love to think myself as a female.I have my family in my Facebook profile and they don’t know about my identity...I liketo keep it that wayWhile such strategic outness online can maintain audience-related boundaries for Hijra, andtherefore safeguard Hijra identity, participants still describe this as a "struggling" or "uncomfortable"position. For some Hijra, this concern over audience extends to a hard choice to completely hidetheir Hijra identity from their families as well as the rest of the world on platforms such asFacebook. Viewing Hijra identity as one of the major obstacles to use social media platformswithout judgement, some Hijra prefer to suppress their own identity in order to enable theirfamilies to still be part of their online audience. This suppression of online identity, however,directly conflicts with crucial benefits of social media platforms. For example, it prevents receivingsocial and instrumental support that is tailored to one’s actual, and not sanitized, identity and lifeexperience, which past work with sexual and gender minority communities suggests is crucial tomaking said support effective and worthwhile [29, 119].
Proc. ACM Hum.-Comput. Interact., Vol. 5, No. CSCW1, Article 157. Publication date: April 2021.
Hijra also use social media to seek guidance and suggestions fromtheir fellow Hijra. 22 out of 61 (around 36%) participants in our study said that connecting withother Hijra in this way, which enables finding appropriate communities in the physical world, is acrucial function of social media, as participant P38(40) noted in FGD4:Now-a-days, we get connected with Hijra from the online communities we have inFacebook...In early days they physically had to find out the house of other Hijra...nowdue to the easy access to internet and Facebook pages, they can find us easilyAs Hijra are often forced to leave their home due to their identity, online communities such as"Badhan Hijra Sangha", "12 Vaja", and "Bangladesh Hijra Kalyan Foundation" (see
Fig. 4 ) serve asa crucial tool for easing Hijras’ transition from their home to Hijra communities. These onlinegroups function as one of the primary tools for finding Hijra followers, Gurumas, and NGOs in theoffline world.This ability to connect via social media groups plays an outsized role in social and especiallypsychological support around the trauma many Hijra experience in their daily lives. Our participantswere clear that their experiences include being regularly sexually harassed and potentially evenraped entirely due to being perceived as vulnerable identity and lower status by the mainstreamnormative population, and there may be little or no support for combating or even just recoveringfrom these incidents in their own home communities. This vulnerability and the lack of supportaround it make online platforms crucial spaces for sharing extreme harassment experiences whichoriginate offline or even in other online social spaces, as sharing these experiences can be crucialto dealing with the psychological fallout and reducing dangerous feelings of isolation. Additionally,online spaces for Hijra are often the ideal starting point for receiving tangible, instrumental support,as while they provide Hijra in distress a chance to be heard by other Hijra, they also connect Hijradirectly to the Gurumas who run these online spaces, and may have resources to share. ParticipantP14(38), who is a Guruma, from FGD3 mentioned,I took in Ruma (pseudonym) at the age of 7. Ruma was being sexually assaulted bymore than 30 men from Ruma’s local area just because Ruma was a hijra...I got to knowabout Ruma from Facebook as Ruma shared those experiences in our online communitylooking for help...I took help of an organization and rescued Ruma to a safe placeFor many like Ruma, groups have been an escape and a crucial resource for many Hijra whowere rescued by Gurumas or NGOs who work with sexually abused Hijra. The groups provide abounded space and a distinct audience for Hijra, and especially provide an opportunity to be openabout traumatic events without being entirely public. However, notably, the need for these kinds ofsupport, services, and connections sometimes drives those that are not in groups, or who fail to findthe support they need in groups, to post their experiences more broadly Instead of sharing in anygroup or community pages, they share those experiences in their personal social media profiles,hoping to be seen and heard by other Hijra. For example, during FGD3, participant P26(20) said:Well, we found her posting on being sexually abused that she shared in her profile...itcame to our attention through multiple sharing...and our Guru Ma wanted to save herand now she’s here with us (in the hijra community)Pursuing the specific audience of Hijra, and finding the appropriate platform within their personalecosystem in which to find Hijra who might be in need, allowed P26 to find and help Hijra that areso isolated as to never join an online Hijra group.Through social media groups, and individual connections, Hijra employ their personal socialmedia ecosystem to find the right audiences from which to draw support, so they no longer have tosuffer alone through traumatic, abusive experiences. For most, social media groups provide a place
Proc. ACM Hum.-Comput. Interact., Vol. 5, No. CSCW1, Article 157. Publication date: April 2021.
Fig. 4. Different Facebook Online Communities of Hijra in Bangladesh to share and be supported around experiences that they could not share physically or virtuallythrough a sanitized profile or a general-audience social media group/page. Moreover, by allowingconnection and solidarity between Hijra, platforms with Hijra audiences can allow Hijra a space tonot hide their online identity, but rather receive direct, psychologically-beneficial support fromtheir peers by fully expressing it - both the good and the bad.
The final, and most potentially fraught, audience Hijra pursueis cisgender men. 36 out of 61 participants (around 59%) mentioned being purposefully connectedwith cisgender men (individuals who identify as men and were assigned male at birth) throughdifferent social media platforms with an expectation of developing a romantic relationship withthem. In FGD3, participant P18(18) said:Hijra people mainly starve for a guy’s company...We add them (online)...we hang outand have fun online all the timeHijra largely turn to social media, as it provides them a convenient platform to access cisgendermen. Building relationships with Hijra is prohibited under sociocultural, religious and political rulesand customs [69], which potentially exclude them from having any sort of traditional relationshipswith men. Hence, social media becomes a crucial channel - potentially, the only viable channel formany Hijra, as it allows them an opportunity to have a romantic or even just flirty relationshipwith men, and to explore basic relationship possibilities.Interestingly, this audience is paired with an audience many Hijra explicitly avoid online: cisgen-der women. As interviewee X4(18) said:We add [men] online who are really handsome...we love talking to them...If we findthem not pretty enough we remove them... But we don’t add any women [cis-females].We envy them. We have a fear men will stop hanging out with us, if we add themonlineOur participants say they often feel overwhelmed by having cisgender women (individuals whoidentify as women and were assigned female at birth) on the same social media platforms theyfrequently use. Cis-women are seen as something of a threat to Hijra when men are also on theplatform, as in that context they are viewed as stiff competition for the attention of men. Asinterviewee X4(18) said:
Proc. ACM Hum.-Comput. Interact., Vol. 5, No. CSCW1, Article 157. Publication date: April 2021.
Fig. 5. Types of Audiences and Relationships of Hijra in Distinct Social Media Platforms
Here’s the difference between a Hijra and a woman...A girl can hardly be with a manin this society but as a Hijra I can sleep with ten other men at the same time. It doesn’tbother me neither the society...Sometimes we feel so happy that we think a womandoesn’t even get the pleasure of being (sexually) with a man like us...But then again, weget sad thinking no man we meet online will ever marry us just because we are HijraDue to their uncomplicated gender identity and sexuality, cis-women serve as a constant reminderto Hijra about their comparatively socially excluded and disadvantageous status. In this way, thepresence of cis-women as an audience on some platforms complicates audience managementshould desirable men be on the same platforms, especially in situations where Hijra are essentiallyperceived as sex objects. Being desperate to be accepted by normative cisgender men on onlineplatforms, where women are avoidable, influences many Hijra to exclude cisgender women asaudiences in their social media ecosystem.Hijra also connect with cisgender men on social media platforms to find opportunities to generateincome via sex work. As it is hard for Hijra to land traditional jobs due to their gender nonconformityand overall social status [15], earning money through sex work often becomes a primary livelihood.As participant P19(40) said during FGD3, social media is often the most accessible way to set upthis sex work:Here (in Hijra community) people can be illiterate, but they surely can use Bigo Live,Facebook and other audio/video applications like IMO, Messengers... Because, they canearn money without any toilThis ability to find sex work is a primary motivation for many Hijra for both using social media andexpanding their networks to include cis men. Some participants even mentioned being connectedwith clients from otherwise-inaccessible foreign countries (such as Pakistan, India, Saudi Arabiaetc.) for sex work through these platforms. Interviewee X5(20) mentioned:Three of my clients from Bigo Live and Facebook actually came to Bangladesh fromabroad to meet me...One of them offered me to go with him and live a married life inabroadHaving potential clients as part of their social media audience makes it easier for many Hijra tonot only make a living via sex work, but also expand their opportunities via gaining popularitybeyond physical borders. While these goals and uses do not align with the core aims of platformsthemselves such as Facebook or Bigo Live, they nonetheless remain crucial for many Hijra. However,
Proc. ACM Hum.-Comput. Interact., Vol. 5, No. CSCW1, Article 157. Publication date: April 2021.
Fig. 6. Different Social Media Platforms Used by Hijra (images are collected and shared with participants’consent) such functionality of using social media platforms to earn money through sex work did not comewithout any consequences for many Hijra. While disclosing their identity online provides Hijraopportunity to earn money, it also puts them at risk of compromising their privacy and hijra identityonline. Despite of being aware of such disadvantages, many Hijra are forced to decide betweentheir privacy and online identity disclosure to ensure their livelihood through sex work.Audiences are an integral part of Hijras social media ecosystem, playing a large motivating role interms of both identity/disclosure management across platforms as well as achieving particular goalsof social meia use. This then plays a large role in setting Hijra’s overall expectations from onlineparticipation (
Fig. 5 ). Whereas certain audiences are expected and welcomed in the ecosystem,same and other audiences can cause discomfort and unwanted existential crisis for some Hijra - aproblem often resolved by spreading audiences across one’s personal social media ecosystem.
As we have demonstrated above, Hijra must balance multiple audiences with very different orienta-tions towards Hijra identity, and therefore conflicting disclosure requirements, in order to deriveboth social and instrumental benefits from social media use. As prior work has shown, availableaffordances, especially for control over audiences and visibility of content, have a large impacton both decision-making around self-presentation and identity disclosure [39] as well as platformchoice within a social media ecosystem [41]. This holds true for Hijra, who must carefully considertheir identity management options and the tools available to them.Affordances (possibilities for action/features that each platform offers) to users play a big role inhow users perceive and experience distinct social media platforms, often for distinct audiences.Our data suggest that, as Hijras social media ecosystems consist of different platforms (see Fig.6) , afffordances offered by each platform have an influence on their preferred online participation.Distinct social media platforms offer distinct means to control audiences online, a primary concernfor Hijra. According to our data, the most popular social media platforms for Hijra were Bigo Live,Facebook, Messenger, IMO, and WhatsApp. Many participants used these platforms almost daily tomaintain their communication, audiences and sex work.
Proc. ACM Hum.-Comput. Interact., Vol. 5, No. CSCW1, Article 157. Publication date: April 2021.
Our participants recounted how certain suites of affordances helpfulfill their individual audience management needs. For example, Hijra primarily maintain commu-nication with their families and close friends using Facebook. Our participants pointed out severalaffordances that make Facebook easy to use when it comes to communicating with family on adaily basis. While our participants recounted using multiple platforms with family, Facebook wasseen as the preferred platform overall due to the affordances provided. For example, compared toother platforms, Facebook affords more of what DeVito et al. call "presentation flexibility" [39] orthe ability to use multiple formats and styles to present oneself to others. For example, intervieweeX15(20) said,I post my photo, chat with friends and families in there (Facebook)...Facebook alsohelps me not only to share my inner feelings with them through sharing posts but alsolet me do check-ins to let them know where I am or where I am going...it is so easier toupdate them about my life...I can also do audio or video call in Facebook MessengerOur participants expressed a desire to use robust media types, not just text and photos, in main-taining these family relationships, and also noted the importance of having all these media typesavailable on a single platform so as to not fragment the family audience.However, for some Hijra (13 out of 61, around 21%), Facebook has not always been an idealplatform to utilize this presentation flexibility. While multi-factor media sharing within Facebookhas provided Hijra robust ways to express themselves to their family connections, it has alsopotentially restricted their flexibility to disclose their true identity in front of these audiences.Even though disclosure of hijra identity was a sensitive information to our participants (suchas X8, X10), some Hijra preferred to share this information on social media willingly with theirfamily connections as a way to express their true self online; however, they were unable to do so.According to interviewee X13(24),No, I did not give hijra as my gender online. . . Even though my family knows about myhijra identity and I am openly Hijra in Facebook, I have put female as my gender as analternative option. . . I know I am not a female but actually in our country there are onlytwo options of gender in Facebook- male and female. . . no where Hijra option is givenHijra’s presentation flexibility on social media is significantly effected by the affordances Facebookhas provided to its users. Despite Facebook’s continuous effort to provide its users flexibilityexpressing their preferred identity online, it particularly fails to assist Hijra- who wish to sharetheir hijra identity with their audiences online, including family connections. Facebook’s "custom"option to specify own gender, apart from male/female, by the users themselves does not work forHijra, as (our participants mentioned) the feature is confusing and unfamiliar to them. For example,from FGD5 participant P43(35) mentioned,We can choose hijra (as gender)? How? I can only see male and female in the op-tions. . . There was a third option probably ("custom"), but I had no idea what that meant.Does that mean hijra?Term such as "custom" is unfamiliar within Hijra, and it provides no information to them regardingtheir flexibility to choose their own gender in their preferred online platforms. As Hijra is notexplicitly included in the gender spectrum of social media platforms, it potentially restricts andforces the members of Hijra community to construct their identity online within the dichotomy ofmale and female.As our data suggests, while Facebook is a big part of Hijra connection with their families, byway of contrast, in flirty communications or sex work situations, it is more useful for Hijra tohave access to a platform with far more limited presentation flexibility, so as to bound the possible
Proc. ACM Hum.-Comput. Interact., Vol. 5, No. CSCW1, Article 157. Publication date: April 2021.
Fig. 7. Leveling up in Bigo Live interactions they can be expected to have with potential audiences or clients. For example, 28 outof 61 Hijra participants (around 46%) preferred to use IMO while building romantic rapport withcis-males due to its primary focus on chat and limited features on presentations formats. FromFGD1, participant P1(18) said,Through IMO, I talk to the men who are nice and interesting...I usually do audio callsthrough IMO and use my female voice to present myself as woman...they never realizethat I am a hijraIMO’s core functionality of regulated one-to-one or group conversations through only audio, video,and written chat (unlike Facebook that also includes other features within the platform) helps Hijrato represent themselves in a way that benefits their goal of building a romantic rapport. Some Hijraalso mentioned using IMO and its limited features and affordances to engage in sex work that iscarefully curated only for their clients. Regarding this, interviewee X10(35) mentioned,I am engaged in sex work through IMO...I like to use it...it is simple and easy and doesthe work for me...I do video chat there and collect money afterwardsAs chat, and not additional functionality such as games or item listings, takes center stage on IMO,the tightly-specified form of limited presentation flexibility afforded on this platform makes iteasier for Hijra keep interactions bounded to sex work and nothing else.
Privacy concerns are, of course, paramount for Hijra when makingdecisions as to how to fulfill their needs via social media. One of these concerns relates to whatDeVito et al. call "audience transparency," or the afforded ability to be aware of who is in one’saudience. Affording increased awareness of not just who is in one’s audience, but also the rele-vant characteristics of those audience members, has proved useful to Hijra pursuing sex work inparticular, as interviewee X11(30) said,Actually I don’t face that problem (privacy issue) in IMO... By making a call throughIMO, I can understand how old are they... In IMO, who have my phone number onlythose people can contact with me. That’s why I like to use IMO more than Facebook
Proc. ACM Hum.-Comput. Interact., Vol. 5, No. CSCW1, Article 157. Publication date: April 2021.
Being engaged in a profession like sex work, concerns regarding who they connect with becomesan important aspect for many Hijra online who wish to stay away from being harassed or bulliedfrom unwanted prospective clients or others. As it is relatively easy to find someone on Facebookeven without knowing their full name or information, it becomes important for many Hijra tohide their personal profile or identity from certain audiences, who can be their potential harassersin future. As many Hijra’s livelihood is dependent on online sex work, they prefer to keep it assafe as possible using platforms like IMO, where clients or audiences with Hijra’s personal phonenumber can only reach to them. However, exposure to a broader but targeted sex work client basehas also been an important part of Hijra sex work online that conforms audience awareness. SomeHijra (15 out of 61, around 24%) preferred to use Bigo Live for its feature of level that helps users tobroadcast their live video to a wider audience, such as foreign sex clients, with similar interests.For example interviewee X2(35) mentioned,Having higher levels in Bigo Live helped me to expand my fan base and to connect morewith foreigners...the higher the level, the more possibilities that people and foreignerswill see my broadcasts and videos...for the kind of work I do (sex work), I prefer to beconnected with clients who are foreigners...They not only pay more but most them arealso nicer and more politeBigo Live allows users to live-stream their favorite moments, and make friends from all aroundthe world through live video/audio/text chat [3]. It also offers its users to boost/promote theircontents to compatible users through increasing their level within the app. One can level up inBigo Live by logging in every day, sending virtual gifts to other users and so on ( see Fig. 7 ). Tomake it simple, the higher the level, the more popular one is and the more easier it is to earn money.Through engaging more on the platform and increasing the level in Bigo Live, some Hijra expectto be seen and discovered by foreign users who are interested in sex work and will provide bigamount of money for their work. Distinct feature offered by Bigo Live has allowed Hijra to expandtheir popularity beyond the border and created an opportunity to be connected with expectedaudiences. Even though for some such exposure may be a privacy issue, for many Hijra it offers astrategic process of selecting clients or audiences to improve their professional lives.Such preference of relationships and audiences in distinct social media platforms provide ussome significant insights on the perceptions of Hijra in Bangladesh who become a part of distinctsocial media ecosystems.
Although distinct social media platforms haveprovided Hijra ways to communicate or earn money through sex work, they also have broughtadditional harassment, making the platform’s afforded level of visibility control, which we refer aspresenting themselves with selective visibility, essential concern when countering harassers. Forexample, many Hijra take advantage of the block feature in Facebook when the level of derogatorylanguage used against them on their own social media platforms becomes intolerable. During FGD1,participant P1(18) mentioned:For example, in Facebook or Bigo Live, when I upload a picture or video, people makecomments like “Hijra” “hot/sexy”, “show me your naked body” etc. I instantly blockthem from thereThe blocking features on Facebook and Bigo Live help Hijra maintain their social media profilesby preventing unwanted harassment. However, in some cases, the blocking feature alone is notenough to afford enough visibility control to Hijra to adequately combat harassment and bullying.Specifically, our participants report that this is a problem with specific harassers who create multipleaccount to circumvent blocks. On this issue, interviewee X2(35) said,
Proc. ACM Hum.-Comput. Interact., Vol. 5, No. CSCW1, Article 157. Publication date: April 2021.
Fig. 8. Anti-Hijra Groups in Facebook ...some people just keep calling and harass me online... when I cannot tolerate anymore, I block them...The irony is some of them open new accounts and add again...oneday I may find out that is the same guy that I blockedHere, we see a conflict between a platform’s afforded high visibility control (individual, fine-grained block tools) and afforded low identity persistence (easier alternatives to the tools) interms of audience management for Hijra. While the platforms are, indeed, trying to afford bettervisibility control to ensure safety, there they are sometimes simultaneously providing ways tocreate new/multiple accounts online for its users that bypasses the usability of the blocking featurein the first place for Hijra.Inability to control who gets access to sensitive information and how has been a big issue forHijra as it often turns into a matter of serious privacy and security concerns. Leaking personalphotos/videos publicly, using personal information to create fake profiles and seek money fromothers and spreading rumors have been common events for Hijra who were either unable to restrictthe harassers or their contents. On this topic, interviewee X6(25) said,...he was sharing personal photos of me and was asking for money from others inFacebook...I could guess who it was but was not sure...one of my colleagues suggestedme to disable my id and report the culprit to prevent him from accessing my personalinformation...but the problem was I had no idea how to disable my ID or to report toFacebook about this event and the culpritMany Hijra lack the necessary training and technical skills to effectively employ platform tools,even to the extent that it seriously impacts their security online. While a majority of the users maybe familiar with features like blocking someone or deactivating personal id to ensure visibilitycontrol online, for some Hijra, these privacy tools may not seem easily accessible, and visible,but rather excessively intricate and complicated. Being unable to utilize such complicated privacycontrols has also been a concern for our participants while reporting different anti-Hijra onlinecommunities that promote hate speech against Hijra (see Fig. 8) . Even though platforms likeFacebook provide users security tools to report online hate speech and harassment, Hijra areseldom made aware of those tools to protect their content or shape their audiences, as participantP5(20) demonstrated in FGD1:
Proc. ACM Hum.-Comput. Interact., Vol. 5, No. CSCW1, Article 157. Publication date: April 2021.
Can we report these online communities in Facebook? Really? How?... What is a report option? I am sorry...We are not good with technology... we don’t know how to use thisoption onlineTo our participants, tools to report online community pages are often unfamiliar, as there isno clear instruction or indication provided to them regarding their options to file complaintsagainst harassers. Even though both block and report are privacy tools to control unwanted eventsand individuals online, Hijra are not aware of all these options due to their limited knowledgeon the platform’s affordances. Not being familiar with options like report, and not seeing anyvisible action against such harassing pages, individuals or contents have forced many Hijra toperceive platforms like Facebook as more hostile towards them. Such limited technical skill-setand knowledge/awareness at controlling their contents, events and individuals intensify Hijra’svulnerability online and make their social media ecosystem more complicated.
Unexpected negative events and harassment can trigger Hijra to shift, limit or stop their onlineparticipation through different social media platforms. As we explore RQ3, our data suggest thatHijra adopt strategic decisions to shape their participation online that are often motivated bythe platforms’ affordances, Hijra skill-set and community influences conforming those negativeexperiences.
Hijra’s experiences of beingseverely bullied and harassed online give Hijra good reason to identify which online spaces, whichDeVito et al. define as the user’s own conflation of platform (including affordances) and audience[41], are less likely to be unsafe. Even though Hijra face extreme harassment online, instead ofdeleting their profiles from distinct social media platforms, 11 out of 61 Hijra (around 18%) preferto shift from one platform to another in a hope to search for a space that will provide betterexperiences in terms of their identity. As interviewee X9(18) mentioned:What I feel best about Bigo Live is it has less harassment...Facebook promotes moreharassment...which is why I left Facebook and moved to Bigo Live...I still have myaccount in Facebook but I don’t use it anymoreAs mentioned in the previous section, Facebook is perceived more hostile towards Hijra, Hijralooks for spaces that are supportive towards them and their identity. However, it is not uncommonfor them to keep their old profiles open. Such a decision to migrate from Facebook to other socialmedia platforms comes with it own consequences. As Facebook is the primary media for manyHijra to be connected with their loved ones, including friends and families, such migration puts adent in their virtual social lives forcing them to compromise their participation online. Whereason one side, Facebook is working as an alternate for social interactions for many Hijra, stressorslike additional negative experiences and privacy concerns force them to transfer or migrate theirparticipation elsewhere.Apart from migrating their participation, many of the Hijra (29 out of 61, around 47%) mentionedeither limiting or withdrawing their participation from distinct social media platforms to protect orsave their identity as Hijra online. For example, interviewee X13(24) mentioned,There are many boys who come in live in Bigo Live, disguise themselves as girls andmake vulgar and defaming videos...There is no way to distinguish the fake Hijra fromus in Bigo Live...For them other people blame and shame Hijra like us...For these reason,I don’t go in live or use Bigo Live much now
Proc. ACM Hum.-Comput. Interact., Vol. 5, No. CSCW1, Article 157. Publication date: April 2021.
Social media platforms like Bigo Live has no tool to identify fake Hijra or fake profiles of Hijra thatoften try to defame the whole Hijra community online. For such cases, it becomes hard for manyHijra to participate online freely, which often leads to limited participation in the space or platform.Apart from this limited participation, our data also suggested social media non-use within ourparticipants. Some Hijra mentioned withdrawing their participation from distinct social mediaplatforms because of getting hacked and being unable to retrieve the profile. As X12(29) said:My Facebook Messenger got hacked...the person who hacked it asked for money toothers and leaked some of my personal photos...I did not know what to do or how tostop it, so I stopped using it [Facebook Messenger]Limited skills to handle concerning situations like profile being hacked has also forced Hijra towithdraw their participation online. Even though our participants were familiar with the use ofsocial media platforms like Facebook Messenger, they were not familiar enough to handle situationslike these where a little bit more knowledge or skill on using the platforms was required. Suchexperiences were concerning for Hijra, which lead them to stop and leave the platform for goodcosting their proper interaction their families or clients through those platforms.
The decision to shift to another platform in one’s ecosystem alsosometimes depends on the group dynamics Hijra value within their communities. As Hijra havestrong bonds within their community, their decisions to move platforms or use certain platforms incertain ways often get influenced by what others from the community suggest, or the informationpeers or authority figures provide. Participant X6(25) (a Guruma from Comilla) mentioned:I heard if you do drugs or something bad, your online ID will be hacked. But BIGOLive is safe though...I also suggest my fellow Hijra to be safe when using these onlineplatformsOften, along with fellow Hijra, authoritative figures such as the Gurumas have power over whata follower should know about or use for their online participation. Even though the informationprovided by the Gurumas are not entirely correct or true every time, due to the influence theyhave over their followers, it effects the way other Hijra shape their participation in the ecosystem.The strong bonds within Hijra communities, which are crucial to internal solidarity and mutualprotection, provide reasons for Hijra to shift their participation from one platform to another basedentirely on this type of personal say-so. This collective mentality is also sometimes visible withinHijra online in a larger sense. Some of our participants also mentioned a tendency to adopt orreject distinct social media platforms based on their collective experiences or own internal groupdynamics. Related to such group dynamics, from FGD1 participant P1(18) said,Well, we use smartphones. We eventually know about different social media plat-forms...If someone in our community uses or prefer any specific one [social mediaplatform] for you know...different reasons, we all get to know about it and try it outWhen a member of the community introduces a new app or platform that is beneficial (in terms ofprivacy, better communication, clients etc.) for the whole group, it often influenced our participantsto shift their social media participation to that new platform. As solidarity within Hijra communitiesis strong, it guides Hijra to adopt a new technology or social media platform.While existence on social media becomes an issue of safety and privacy for Hijra, restrictingparticipation online seemed more feasible for many of our participants. By limiting, withdrawingand shifting participation from one social media to other, Hijra strategically try to control suchinstances as much as possible with a cost of their smooth participation online.
Proc. ACM Hum.-Comput. Interact., Vol. 5, No. CSCW1, Article 157. Publication date: April 2021.
Our study highlights important aspects of social media ecosystem of Hijra, a GSM populationwho have their own specific audience concerns and perceptions of affordances on different socialmedia platforms. While RQ1 and RQ2 specifically inquire into Hijra’s social media ecosystem tounderstand with whom, how and where they build their connections and perceived affordances, RQ3digs further to see where the system fails and thus force Hijra to migrate or shift their participationon online social media.
As our results suggest, Hijra break through traditional gender boundaries and cannot be reducedto merely metonymic, Western figures for an analysis of gender fluidity [70]. During the study, itwas interesting to explore participants’ choice of terminologies to define their identity that do notfit under the existing English language terms for other GSM communities. It is possible that manyamong the LGBTQ+ community from Western context are unaware of most of these local termsused to define GSM identities. In Western contexts, many LGBTQ+ communities use the term queeras an "umbrella" term that encompasses all who identify as lesbian, gay, bisexual, or transgender[123]. However, due to Hijra’s unique multi-classifications of gender perceptions (that may notconform to the normative social expectations and definition of gender minority and sexuality),it is unfair to try and explain Hijra identity using English LGBTQ+ terminology [7]. Whereasnon-Western mainstream societies fail to articulate the concept of "Meye" (for intersex) and "Chele"(for transgender) within Hijra communities by forcing "hijra" as general gender term for all Hijra,there it is even potentially offensive to use Westernized gender and sexuality categories for them.Even in the Western world, the categories that are used to define GSM are not self-evident, andraise the need of asking localized questions on what these categories mean to the people in aspecific country [79]. Patil [99] describes this notion as a heterosexual matrix within predominantwestern epistemic frame that defines gender intelligibility from male-female dichotomy and posesa challenges for the people who can not be classified into a normative or even expanded westernschema. As a consequence, this study adds to the conversation on GSM that exists in non-Westerncontext by exploring Hijra’s identity perceptions, of which some are impossible to be translated orto fit into the Western models of gender and sexual identities [84].As a stigmatized gender minority group, Hijra lack social contact with others and therefore,online spaces like social media platforms hold particular value for this community people. Whileprevious work on GSM [26, 54] discuss the struggles of LGBTQ+ populations being a part of theonline spaces, our paper extends the definition by accounting Hijra from non-Western contextsinto the conversations. The study shows that Hijra use different social media platforms to carryout certain sets of actions and expectations that are carefully distributed throughout their onlineparticipation. Whereas, the dominant mainstream population often considers the decision of beinga part of an online space as a choice rather than a necessity for many stigmatized populations,Hijra’s dependencies on social media platforms to achieve primary livelihood through online sexwork or social life (that they are deprived of in the physical world) reveal their vulnerabilitiesand necessary dependencies on online spaces. Previous studies [5, 11, 32] have explored suchengagement of Hijra in the field of health or social science, but it is absent in the field of CSCWand social computing. Due to their complex identity, Hijra fall into the conundrum of choosingbetween online exposure and identity protection that in turn motivates them to look beyond onlyaudience management objectives while navigating through different online platforms. This studyhighlights those complexities by engaging in deeper explorations on GSM social media ecosystemsand evaluating the framework more intensely from non-Western context.
Proc. ACM Hum.-Comput. Interact., Vol. 5, No. CSCW1, Article 157. Publication date: April 2021.
Similar to the existing framework of social media ecosystem [41], our participants described relyingon perceived audiences who are an integral part of their personal social media ecosystem. Audiencessuch as "targeted imagined audience” based on communal ties [77] (close families and friends aswell as other Hijra members of the community), or “outright” targeted audience (such as cis-males)[41] play an important role in Hijra’s online ecosystem that is constructed via interplay betweenspaces and affordances. Due to the stigmatization Hijra face for their identity, they are in constantsearch of audience awareness, controlled exposure and inclusiveness by segregating identity relatedcontents in distinct spaces. For example, even though participant X11 strategically restricts sharingany hijra related content in Facebook due to having social and cultural obstacles (social imagetowards family and friends), X11 does not hold back disclosing those contents in Bigo Live orIMO because of the platforms’ afforded safety and controlled exposure. For many Hijra, havingtightly-specified form of communication space with easier, less complicated features heightens theirpresentation flexibility towards their audiences and builds a sense of control on who they connectwith and how (such as participant X2). Often, affordances in Hijra’s social media ecosystem arenot enough to address their audience related concerns on self-presentations and visibility control.Even though the current lens of social media ecosystem focuses on the combination of audienceand affordances in audience management, for Hijra, such interplay does not always work due totheir lack of understanding of the platform’s affordances as well as required technical skill-set andknowledge to manage audiences.According to DeVito et. al. [41], GSM populations from Western context tailor their onlineLGBTQ+ presentation via affordances they see as helpful; however for Hijra, same affordancesbecome challenging and limited due to the perceived low identity persistence of the spaces andtheir limited technical skill set and knowledge (such as participant X6). Unable to control theirself-presentation using provided privacy tools, many Hijra face unavoidable harassing experiencesonline, which in turn, impact the way Hijra perceive their audience management strategies. Whiletotal around 65% of our participants strategically migrated or limited their online participation fromone platform to another due to these negative experiences, they kept their personal profiles openand purposefully not managed in terms of audiences. Even though Western GSM populations havea tendency to follow rigorous processes in terms of sharing content with specific audiences online[56], Hijra (e.g. participant X9) tend to less so while shifting participation for audience specificreasons without erasing or fully closing previous profile at all, which directs our observations tothe likelihood of their reversion [19]. Due to Hijra’s professional and personal objectives, they oftenneed to ensure maximum exposure online; this may influence them to revert back to online spacesthey left earlier. Current models of social media ecosystem needs to address such dynamic shiftingof participation by GSM users who purposefully keep their audience management strategies relaxedwhile leaving certain social media platform.
While our data strongly suggest that Hijra are at many times aligning their personal social mediaecosystems through Personal Social Media Ecosystem framework [41], it also clearly reveals areaswhere this lens must be extended to better account for non-Western contexts. Our findings onHijra’s struggle with self-presentation on social media extends the existing lens of social mediaecosystem by considering technical skill and knowledge as a fourth element in the framework.These findings align with the current literature on digital literacy that recognize user skill as anvital part of privacy control and must include GSM communities from non-Western contexts.
Proc. ACM Hum.-Comput. Interact., Vol. 5, No. CSCW1, Article 157. Publication date: April 2021.
In our study, we have observed a strong influence of technicalknowledge/literacy or skill on Hijra’s way of addressing their audience concerns and spaces.This skill/knowledge represents additional factor that potentially and significantly shapes howstigmatized GSM populations outside the US construct their concept of audience privacy andspaces and direct their participation online accordingly [92]. Past literature like [39, 41] haveexplored LGBTQ+ and their audience concerns through the online spaces and affordances but didnot account the concept of skill that can highly impact the spectrum of social media ecosystem forHijra. By zooming out from the current social media ecosystem framework and looking througha macroscopic lens, we can explore the comprehensibility of the affordances and users’ specificskill sets that are required by the platforms to fully utilize the affordances. A prior study finds thattechnology design for marginalized people in mind are supposed to have low barriers of access[48]. Having tons of privacy tools for the users is not enough if they are not accessible, especiallyfor the marginalized GSM populations like Hijra. Here accessibility does not mean access to thetechnology, but rather denotes the concept of accessing the knowledge that is required to reap fullbenefits from the platforms’ affordances. Our data suggests that this struggle with technologicalunderstanding and skill is not an individual problem, but rather a common experience within manyHijra communities. Sambasivan et al. [104] have mentioned how in developing countries this typeof struggle is not uncommon, as many people from low-income communities lack textual anddigital literacy that effect their technology- operation skills. Hence, while we try to explore howstigmatized users from Western context interact with social media to construct their identity by onlyconsidering provided platforms’ affordances, we may potentially exclude other stigmatized GSMcommunities from non-Western context whose perceived platform affordances are significantlyimpacted by their technical knowledge and skill-set.Research on postcolonial computing shows that mainstream computing knowledge is oftenignorant towards local understanding of technologies, which creates a space of marginalization andfailure against local communities [8, 111]. Localized knowledge and understanding of technologieshave been historically marginalized, suppressed and neglected [109, 111]. This type of mindsetrepresents a heightened danger for stigmatized individuals like Hijra. Even though Hijra are familiarwith the basic utilization of the platforms, in many cases, they have no knowledge of the vast controlsettings that the platforms offer to them. For example, even though the usability of "disabling id" or“reporting” feature may seem straightforward to most users in Western context, due to the limitedknowledge that is accessible to Hijra, participant X6 expressed frustration of not being able toidentify and utilize the features during privacy concerns. Devito et. al’s framework on social mediaecosystem defines the online behavioral norms of LGBTQ+ communities from Western context bysituating focus on their reliability on multiple spaces to share distinct content as a privacy controlmeasure [41]. Even though it touches based on how privacy controls on some individual platformsseemed to be irrelevant in the face of contextually- dependent behavior for their users, it did notfactor the skill and knowledge that potentially dictate the relevancy of the privacy controls of aplatform. Users can become unable to personalize or control the data they share online [96], as thetask of exploring and mastering this protective technology is often left entirely to the adoptersthemselves [9, 14]. For instance, for participants like P5, Facebook (with its huge investments intoprivacy tools) can be perceived as less safer and inclusive for presenting Hijra identity as opposed toother platforms due to P5’s limited skill and knowledge about the features provided by the platformto report offensive anti-Hijra communities online. As Facebook relied on the assumption that usersare skilled enough to adopt platform affordances, it potentially ignore marginalized populationslike Hijra, who may not have the same access to the knowledge and skill. Being unable to utilizethe privacy controls, many marginalized users face harassment online that effectively force them torestrict their participation on social media platforms [93, 114] or employ heavily tailored privacy
Proc. ACM Hum.-Comput. Interact., Vol. 5, No. CSCW1, Article 157. Publication date: April 2021. settings to manage visibility within a single platform [41]. For Hijra, they blame the space andredirect their participation elsewhere instead of utilizing the platform affordances. By accountingskill, we are able to directly interrogate the impact of the affordances on users’ perceptions ofonline spaces, self-presentation and content sharing strategies. Therefore, this study suggests toinclude skill as a fourth element in the social media ecosystem framework that, beyond just Hijraor other GSM populations from certain contexts, potentially controls how users perceive onlinespaces and interact.
While we establish technical skill orknowledge as a primary element within social media ecosystem for Hijra in non-Western context,we also observe community and cultural influence on these low skilled users that potentiallyconstruct their online practices as well. Hijra, with limited knowledge on platforms features, areoften dependent on the information flow passed to them from their closely knitted community.Previous literature on information seeking states that users with limited resources seek and makesense of any information they receive when they have high-level self-presentation goals, butmay not know exactly how to achieve them [40, 50, 101]. Connections like friends and familiesstep in such cases and inform the users of their social media platforms’ affordances, includinglax privacy settings [40].However, for Hijra, such exogenous information come from other Hijramembers of the community or sometimes from authoritative figures like Guruma, who often alsohave limited access to information on affordances. Their influences as sources of informationon platform affordances impact the way Hijra’s ecosystem are built or perceived. For example,participants like X6 (who is a Guruma) and P1 (fellow Hijra) are potentially motivated to directsocial media participation based on the (limited) knowledge on platforms’ affordances they have,whilst influencing other Hijra with limited skill set within the community to construct similarperception.As many Hijra get engaged with unsolicited work to earn money through distinct social mediaecosystem, to protect themselves from unwanted harassment and invasion of personal privacy,Hijra are at constant look for adopting new social media platforms that will bring the more clientsand better privacy. This adoption of new platforms often happens earlier with Gurumas who wishto ensure their followers safety online before they engage in it. It is fascinating to see how Hijraare not only influenced by their peers but also by their Gurumas who not only stipulate theirphysical lifestyle, but also dictate how their online participation should look like (often for theirown safety). Taking influence of authoritative figure and community influence on limited skill-setinto consideration help us to explore Hijra’s perception, or in a broader sense GSM perception, onbuilding their social media ecosystem more accurately.Apart from community influence, cultural influence on skill-set has also been an importantfactor for Hijra in Bangladesh. While many social media platforms offer its users extensive optionsof privacy settings, they are not always culturally appropriate and signified and thus, becomeunnecessary and useless. For example, even though participant P5 was struggling through intenseonline harassment for being Hijra online, it never occurred to them to "report" against someoneor pages/groups, as the feature is not culturally signified to this low-skilled population. Whileprevious research states that social media can reinforce personalized content curating behaviorswithin users that range from hiding posts to blocking other users [121], there some Hijra primarilyonline rely on the feature "block" than any other censorship tool to restrict harassment or unwantedconnections. However, such feature of blocking users became impractical by the existence offake profiles online. As Hijra often add unknown people to their existing online profiles to buildconnections, experiences of being deceived by the fake profiles make it harder for Hijra to trustthe space. Trust and risk have been theorized as the most influential factors affecting individual
Proc. ACM Hum.-Comput. Interact., Vol. 5, No. CSCW1, Article 157. Publication date: April 2021. adopting or rejecting behavior toward social media platforms [118]. Whereas Hijra acknowledgeharassing contents created against them are done by homophobic users in Bangladesh, they alsocount the space responsible for hosting such negative contents. Such distrust on a specific spaceinfluence Hijra to leave or shift their participation to another platform that is less harassing towardsthem or needs less technological understanding to control. This indicates an inadequacy within theplatform design system that fails to serve GSM populations from a culturally different context.
Our findings add to the broader conversation on building more inclusive technologies for GSMpopulations by identifying specific design practices that need to be adopted by the designersduring platform development. Be it for gender construction or privacy controls during online self-presentation, GSM users like Hijra from non-Western contexts struggle with platform designers’current design practices that are mostly informed by the Western notions of affordances [8, 88, 103].Thus, to ensure more comprehensive online platforms for participants like Hijra, this study arguesfor incorporating non-Western marginalized views in design practices by the platform designers.This aligns with the recent body of work that signify the importance of developing appropriatetechnologies for non-marginalized as well as marginalized populations in both global north andsouth contexts [8, 17, 41, 56, 89]. Our study highlights two of the most crucial implications fordesign practices in the context of non-Western GSM users that may help designers to develop moreinclusive social media platforms.
First, due to the unfamiliarity with custom affordancesthat further requires users to fill out their own gender on social media, participants like X13, P43in our study mentioned being unable to utilize platform’s current mechanism to select genderthat potentially forces them to compromise their identity construction. As the current design ofmost of the online platforms incorporate less granular gender choices for their users that onlyinclude male, female and custom/others options, this extra step to customize gender while nothaving hijra as a direct option similar to male/female categorization confuse Hijra, making theiridentity constructions on social media further challenging. Regarding this classification of genderprovided by online platforms like Facebook, previous study mentioned that such classifications existssomewhere in-between a rigid binary and fluid spectrum [24]. Even after Facebook has incorporatednew ways to increase gender flexibility for its users, in their structural level, it still has continued to fitnon-binary genders into binary classifications to serve stakeholders while also shaping the perceivedneeds and preferences of both users and advertisement clients [24, 25, 107]. Similarly, Haimson andHoffman have also questioned Facebook’s underlying mechanisms used to impose legitimacy on itswebsite, disproportionately affecting trans individuals and drag performers [59, 107]. These priorstudies back our arguments on the existing shortcomings of gender classification design practicesof platforms like Facebook that need to be reevaluated and restructured from a non-Westernperspective for communities such as Hijra. While Facebook prides to support GSM populations byproviding 56 gender options (under custom) in the interface that pop up when a user attempts totype in their preferred gender term [4, 24], our study shows that for Hijra, it fails to support insimilar manner. A user needs to type in the whole term "hijra"/"Hijra"/"Meye Hijra"/"Chele Hijra"to select their gender, whereas for other GSM users, such as trans populations, typing in only aletter brings up possible suggestions related to that gender (see Fig. 9) . This confirms how platformdesigners’ understanding of gender is limited within a pre-set Westernized gender classificationsand force GSM users from non-Western context to follow extra complicated steps to construct theiridentity online. This limitation within design practices brings forth a classic tension in informationstudies around classification and categorization where classifications provide orders but may miss
Proc. ACM Hum.-Comput. Interact., Vol. 5, No. CSCW1, Article 157. Publication date: April 2021.
Fig. 9. Classifications of Gender in Facebook out certain categories [27, 57]. While research like [56] focused on similar constructions of genderthrough problematic design implications in Facebook, such research is predominantly based onWestern trans populations who may have better understandings on platform affordances than Hijrafrom non-Western context. Thus, to set the focus more on non-Western GSM context, our studyjoins the discussion and encourages designers to think holistically about how these marginalizedusers prefer to define their gender online and provide more accessible and culturally appropriatedoptions to them.
Second, our participants like P5 and P1 discuss Hijra’s online struggleof using perceived platform affordances to combat community based online harassment that theyface frequently due to their identity as Hijra. Even though platform designers include many privacycontrols setting for their users while developing platforms, such as features like reporting orblockng individuals/disabling personal id, it was evident within our findings that such mechanismoften do not work for Hijra from individual level. Our study finds, Hijra have strong communityaspect within themselves; as such, their perceived platform affordances is significantly influencedby their group dynamics and largely dependent on the sharing of information within communities.Thus, incorporating and introducing group level privacy tools, such as collective block list, maybenefit Hijra, as it can facilitate individual effort to combat online harassment with group supportas well as can create a more manageable and user-friendly experience for them. Conceptualizationsof online privacy remain mostly at the individual level in Western context [21, 33, 45], and whileresearch like [36, 65, 110] validates group privacy concerns as parts of design practices from aWestern non-marginalized non-GSM context, our study strengthens these design implications bysituating them in non-Western contexts for GSM. Of course, this must be weighed against technicalunderstanding of the GSM users, as these design practices will end up adding more complexities tothe existing privacy mechanism if the designers incorporate them within design practices withoutproviding proper guidelines to their users who may lack access to the needed knowledge of usingthese tools. Therefore, platform designers need to think more expansively about how they canaddress this inaccessibility of knowledge and skill by the GSM users while also adding privacyflexibility to them from non-Western context, as it significantly impact their self-presentationonline.
This study has some distinct limitations. Even though we wanted to ensure author accountability forthe findings reported in this study by member checking our results with participants, we could notfulfill our objective. Due to our participants’ extremely busy schedule and being severely affectedby COVID-19 in Bangladesh, we could not meet our participants to discuss the reported results.Even though we accept it as an unfortunate limitation to our study design, we have tried our bestto do extensive background research on Hijra community and carefully checked all the claims we
Proc. ACM Hum.-Comput. Interact., Vol. 5, No. CSCW1, Article 157. Publication date: April 2021. made in the paper to ensure accountability. Additionally, while this study primarily focused onHijra’s self-presentation and online practices through social media ecosystem framework, it wasoutside the scope of this study to examine external circumstances such as local laws and changingpolitical circumstances which can also effect Hijra’s social status both offline and online. Futurework could explore these structural factors while also investigating the possible intracommunitydifferences between trans and intersex Hijra individuals through an ecological lens [41], potentiallyyielding a more complete picture of Hijra, or in general GSM self-presentation in non-Westerncontexts.
Social interaction across multiple online platforms is a challenging issue for members of GSM dueto the stigmatizations they face in daily basis for their identity, which increases the complexityof their self-presentation decisions. In this paper, through investigating personal social mediaecosystem model, we have explored how GSM from non-western context, such as Hijra, constructtheir online participation and self-presentation focusing audiences, affordances and spaces in mind.This paper has also extended the model by incorporating user skill and knowledge as a fourthcrucial element in the ecosystem that significantly impacts Hijra’s perceptions on online spaceand online practices, such as audience management or platform migrations. Our contributionson community and cultural influence on Hijra’s online participation have inspired us towardsimplications for design practices that take account of more accessible and culturally appropriatedgender categorizations for Hijra as well as group level privacy controls to facilitate these populationsduring online harassment. We believe these suggestions on design practices can provide guidanceto the researchers and designers in further efforts to understand and support GSM in achievingtheir self-presentation goals, and lay the groundwork for future in-depth work on these populationsin an increasingly diversifying social media space.
We would like to acknowledge
Sachetan Somajsheba Hijra Sangha (a non-profit organization) and itsassociated people for their enormous support in helping us being connected with Hijra communitiesfor this project. Without their support, it would have been hard to conduct this study. We wouldalso like to thank the guruma (requested to be unnamed) who has helped us during the datacollection process and provided advice throughout the step. Furthermore, we would like to extendour gratitude to all of our reviewers who have provided us detailed and important suggestionsduring the review process that has helped us to build a much stronger submission.
REFERENCES
BMC international health and human rights
12, 1 (2012), 32.[6] Alex A Ahmed. 2018. Trans Competent Interaction Design: A Qualitative Study on Voice, Identity, and Technology.
Interacting with Computers
30, 1 (2018), 53–71.[7] Ibtisam Ahmed. 2017. The Hijra community and the complex path to decolonising gender in Bangladesh. https://bit.ly/33ygAwX[8] Syed Ishtiaque Ahmed, Md Romael Haque, Jay Chen, and Nicola Dell. 2017. Digital privacy challenges with sharedmobile phone use in bangladesh.
Proceedings of the ACM on Human-Computer Interaction
1, CSCW (2017), 1–20.Proc. ACM Hum.-Comput. Interact., Vol. 5, No. CSCW1, Article 157. Publication date: April 2021. [9] Syed Ishtiaque Ahmed, Steven J Jackson, Maruf Zaber, Mehrab Bin Morshed, Md Habibullah Bin Ismail, and SharminAfrose. 2013. Ecologies of use and design: individual and social practices of mobile phone use within low-literaterickshawpuller communities in urban Bangladesh. In
Proceedings of the 4th Annual Symposium on Computing forDevelopment . 1–10.[10] Sigma Ainul, Ashish Bajracharya, Laura Reichenbach, and Kate Gilles. 2017. Adolescents in Bangladesh: A situationanalysis of programmatic approaches to sexual and reproductive health education and services. (2017).[11] Arshad Altaf, Aysha Zahidie, and Ajmal Agha. 2012. Comparing risk factors of HIV among hijra sex workers inLarkana and other cities of Pakistan: an analytical cross sectional study.
BMC public health
12, 1 (2012), 279.[12] Nazanin Andalibi, Oliver L Haimson, Munmun De Choudhury, and Andrea Forte. 2018. Social support, reciprocity,and anonymity in responses to sexual abuse disclosures on social media.
ACM Transactions on Computer-HumanInteraction (TOCHI)
25, 5 (2018), 1–35.[13] Nazanin Andalibi, Pinar Ozturk, and Andrea Forte. 2017. Sensitive Self-disclosures, Responses, and Social Support onInstagram: the case of
Proceedings of the 2017 ACM conference on computer supported cooperative workand social computing . 1485–1500.[14] Keith Atkinson, Jaclyn Barnes, Judith Albee, Peter Anttila, Judith Haataja, Kanak Nanavati, Kelly Steelman, andCharles Wallace. 2016. Breaking barriers to digital literacy: An intergenerational social-cognitive approach. In
Proceedings of the 18th International ACM SIGACCESS Conference on Computers and Accessibility . 239–244.[15] Abdul Aziz and Sameena Azhar. 2019. Social exclusion and official recognition of hijra in Bangladesh. (2019).[16] Paola Bacchetta. 1999. When the (Hindu) nation exiles its queers.
Social Text
61 (1999), 141–166.[17] Miriam Bartsch and Tobias Dienlin. 2016. Control your Facebook: An analysis of online privacy literacy.
Computersin Human Behavior
56 (2016), 147–154.[18] Adam Bates, Trish Hobman, and Beth T Bell. 2020. “Let Me Do What I Please With It... Don’t Decide My Identity ForMe”: LGBTQ+ Youth Experiences of Social Media in Narrative Identity Development.
Journal of Adolescent Research
35, 1 (2020), 51–83.[19] Eric PS Baumer, Shion Guha, Emily Quan, David Mimno, and Geri K Gay. 2015. Missing photos, suffering withdrawal,or finding freedom? How experiences of social media non-use influence the likelihood of reversion.
Social Media+Society
1, 2 (2015), 2056305115614851.[20] Max Bearak. 2016. Why terms like ’transgender’ don’t work for India’s ’third-gender’ communities. https://wapo.st/2yXVvOj[21] France Bélanger and Robert E Crossler. 2011. Privacy in the digital age: a review of information privacy research ininformation systems.
MIS quarterly (2011), 1017–1041.[22] Victoria Betton, Rohan Borschmann, Mary Docherty, Stephen Coleman, Mark Brown, and Claire Henderson. 2015.The role of social media in reducing stigma and discrimination.
The British Journal of Psychiatry
New directions for student services
NewMedia & Society
19, 6 (2017), 880–898.[25] Rena Bivens and Oliver L Haimson. 2016. Baking gender into social media design: How platforms shape categoriesfor users and advertisers.
Social Media+ Society
2, 4 (2016), 2056305116672486.[26] Lindsay Blackwell, Jean Hardy, Tawfiq Ammari, Tiffany Veinot, Cliff Lampe, and Sarita Schoenebeck. 2016. LGBTparents and social media: Advocacy, privacy, and disclosure during shifting social movements. In
Proceedings of the2016 CHI conference on human factors in computing systems . ACM, 610–622.[27] C Bowker Geoffrey and Susan Leigh Star. 1999. Sorting Things out: classification and its consequences.[28] D Boyd. 2008. Why youth (heart) social network sites: The changing place of digital media in teenage social life.
Youth, identity, and digital media (2008), 119–142.[29] Melanie E Brewster, Bonnie Moradi, Cirleen DeBlaere, and Brandon L Velez. 2013. Navigating the borderlands: Theroles of minority stressors, bicultural self-efficacy, and cognitive flexibility in the mental health of bisexual individuals.
Journal of Counseling Psychology
60, 4 (2013), 543.[30] Janne CH Bromseth. 2002. Public places–public activities.
Methodological approaches and ethical dilemmas in researchon computer-mediated communication contexts. SKIKT’Researchers–Researching IT in context, University of Oslo (2002),33–61.[31] Liam Bullingham and Ana C Vasconcelos. 2013. ‘The presentation of self in the online world’: Goffman and the studyof online identities.
Journal of information science
39, 1 (2013), 101–112.[32] Venkatesan Chakrapani, Priya Babu, and Timothy Ebenezer. 2004. Hijras in sex work face discrimination in theIndian health-care system.
Research for sex work [33] Yunan Chen and Heng Xu. 2013. Privacy management in dynamic groups: understanding information privacy inmedical practices. In
Proceedings of the 2013 conference on Computer supported cooperative work . 541–552.[34] Margaret Cooper and Kristina Dzara. 2010. The Facebook revolution: LGBT identity and activism. In
LGBT identityand online new media . Routledge, 114–126.[35] Katie Davis and Emily Weinstein. 2017. Identity development in the digital age: An Eriksonian perspective. In
Identity,sexuality, and relationships among emerging adults in the digital age . IGI Global, 1–17.[36] Ralf De Wolf, Koen Willaert, and Jo Pierson. 2014. Managing privacy boundaries together: Exploring individual andgroup privacy management strategies in Facebook.
Computers in Human Behavior
Proceedings of the 2017 ACM conference on computer supportedcooperative work and social computing . ACM, 740–754.[40] Michael A DeVito, Jeremy Birnholtz, Jeffery T Hancock, Megan French, and Sunny Liu. 2018. How people form folktheories of social media feeds and what it means for how we study self-presentation. In
Proceedings of the 2018 CHIConference on Human Factors in Computing Systems . 1–12.[41] Michael A DeVito, Ashley Marie Walker, and Jeremy Birnholtz. 2018. ’Too Gay for Facebook’: Presenting LGBTQ+Identity Throughout the Personal Social Media Ecosystem.
Proceedings of the ACM on Human-Computer Interaction
New Media & Society
18, 6 (2016), 891–907.[44] Casey Fiesler, Alyson Young, Tamara Peyton, Amy S Bruckman, Mary Gray, Jeff Hancock, and Wayne Lutters. 2015.Ethics for studying online sociotechnical systems in a big data world. In
Proceedings of the 18th ACM ConferenceCompanion on Computer Supported Cooperative Work & Social Computing . 289–292.[45] Luciano Floridi. 2014. Open data, data protection, and group privacy.
Philosophy & Technology
27, 1 (2014), 1–3.[46] M Paz Galupo, Tiana E Krum, D Brienne Hagen, Kirsten A Gonzalez, and LA Bauerband. 2014. Disclosure oftransgender identity and status in the context of friendship.
Journal of LGBT Issues in Counseling
8, 1 (2014), 25–42.[47] Shane Gannon. 2007. With Respect to Sex: Negotiating Hijra Identity in South India.
Journal of the History of Sexuality
16, 2 (2007), 328–330.[48] Franca Garzotto and Heidi Schelhowe. 2008. Marginalized young people: inclusion through ICT. In
Proceedings of the7th international Conference on interaction Design and Children . 101–104.[49] Banhishikha Ghosh. 2017. Representation of the Hijra: Need for Social Work Intervention. (12 2017).[50] Tarleton Gillespie. 2014. The relevance of algorithms.
Media technologies: Essays on communication, materiality, andsociety
167 (2014), 167.[51] Barney G Glaser, Anselm L Strauss, and Elizabeth Strutzel. 1968. The discovery of grounded theory; strategies forqualitative research.
Nursing research
17, 4 (1968), 364.[52] Ina Goel. 2016. Hijra communities of Delhi.
Sexualities
19, 5-6 (2016), 535–546.[53] Erving Goffman et al. 1978.
The presentation of self in everyday life . Harmondsworth London.[54] Oliver L Haimson. 2016. Transgender experiences with online harassment. In
Social Computing Symposium .[55] Oliver L Haimson. 2017. The Social Complexities of Transgender Identity Disclosure on Social Network Sites. In
Proceedings of the 2017 CHI Conference Extended Abstracts on Human Factors in Computing Systems . ACM, 280–285.[56] Oliver L Haimson, Jed R Brubaker, Lynn Dombrowski, and Gillian R Hayes. 2015. Disclosure, stress, and supportduring gender transition on Facebook. In
Proceedings of the 18th ACM Conference on Computer Supported CooperativeWork & Social Computing . 1176–1190.[57] Oliver L Haimson, Jed R Brubaker, Courtney Loder, Lynn Dombrowski, and Gillian R Hayes. 2015. User Response toFacebook’s Custom Gender Options. iConference 2015 Proceedings (2015).[58] Oliver L Haimson, Dykee Gorrell, Denny L Starks, and Zu Weinger. 2020. Designing Trans Technology: DefiningChallenges and Envisioning Community-Centered Solutions. In
Proceedings of the 2020 CHI Conference on HumanFactors in Computing Systems . 1–13.[59] Oliver L Haimson and Anna Lauren Hoffmann. 2016. Constructing and enforcing" authentic" identity online: Facebook,real names, and non-normative identities.
First Monday
21, 6 (2016).[60] SM Taiabul Haque, Pratyasha Saha, Muhammad Sajidur Rahman, and Syed Ishtiaque Ahmed. 2019. Of Ulti,’hajano’,and" Matachetar otanetak datam" Exploring Local Practices of Exchanging Confidential and Sensitive Information inProc. ACM Hum.-Comput. Interact., Vol. 5, No. CSCW1, Article 157. Publication date: April 2021.
Urban Bangladesh.
Proceedings of the ACM on Human-Computer Interaction
3, CSCW (2019), 1–22.[61] Eszter Hargittai and Aaron Shaw. 2015. Mind the skills gap: the role of Internet know-how and gender in differentiatedcontributions to Wikipedia.
Information, communication & society
18, 4 (2015), 424–442.[62] Adnan Hossain. 2017. The paradox of recognition: hijra, third gender and sexual rights in Bangladesh.
Culture, Health& Sexuality
19, 12 (2017), 1418–1431.[63] Vikas Jayadeva. 2017. Understanding the Mental Health of the Hijra Women of India.
American Journal of PsychiatryResidents’ Journal
12, 05 (2017), 7–9.[64] Lubna Jebin and Umme Farhana. 2015. The Rights of Hijras in Bangladesh: An Overview. (06 2015).[65] Shagun Jhaver, Sucheta Ghoshal, Amy Bruckman, and Eric Gilbert. 2018. Online harassment and content moderation:The case of blocklists.
ACM Transactions on Computer-Human Interaction (TOCHI)
25, 2 (2018), 1–33.[66] Adam N Joinson. 2008. Looking at, looking up or keeping up with people? Motives and use of Facebook. In
Proceedingsof the SIGCHI conference on Human Factors in Computing Systems . 1027–1036.[67] Os Keyes. 2018. The misgendering machines: Trans/HCI implications of automatic gender recognition.
Proceedings ofthe ACM on Human-Computer Interaction
Journal of health, population, and nutrition
The Centre for Internet and Society
13 (2010).[73] Neha Kumar, Marisol Wong-Villacres, Naveena Karusala, Aditya Vishwanath, Arkadeep Kumar, and Azra Ismail. 2019.Aspirations-based design. In
Proceedings of the Tenth International Conference on Information and CommunicationTechnologies and Development . 1–11.[74] Laura E Kuper and Brian S Mustanski. 2014. Using narrative analysis to identify patterns of Internet influence on theidentity development of same-sex attracted youth.
Journal of Adolescent Research
29, 4 (2014), 499–532.[75] Mark R Leary and Robin M Kowalski. 1990. Impression management: A literature review and two-component model.
Psychological bulletin
Proceedings of the 2020 CHI Conference on Human Factors in ComputingSystems . 1–13.[77] Eden Litt and Eszter Hargittai. 2016. The imagined audience on social network sites.
Social Media+ Society
2, 1 (2016),2056305116633482.[78] Sonia Livingstone. 2007. Strategies of parental regulation in the media-rich home.
Computers in human behavior
LGBT-ParentFamilies . Springer, 209–223.[80] Jose Marichal. 2013. Political Facebook groups: Micro-activism and the digital front stage.
First Monday
18, 12 (2013).[81] Julie McCarthy. 2014. A journey of pain and beauty: On becoming transgender in India.
NPR (National Public Radio) (2014).[82] Elizabeth McConnell, Bálint Néray, Bernie Hogan, Aaron Korpak, Antonia Clifford, and Michelle Birkett. 2018.“Everybody puts their whole life on Facebook”: Identity Management and the online social networks of LGBTQ youth.
International journal of environmental research and public health
15, 6 (2018), 1078.[83] Indrani Medhi, SN Nagasena Gautama, and Kentaro Toyama. 2009. A comparison of mobile money-transfer UIs fornon-literate and semi-literate users. In
Proceedings of the SIGCHI Conference on Human Factors in Computing Systems .1741–1750.[84] Aleksa Milanović. 2017. The Impact of Western Society onto the Identity Politics of Sexual and Gender Minorities inColonial and Post-colonial India. In
Regimes of Invisibility in Contemporary Art, Theory and Culture . Springer, 61–73.[85] Rahul Mitra. 2010. Resisting the spectacle of pride: Queer Indian bloggers as interpretive communities.
Journal ofbroadcasting & electronic media
54, 1 (2010), 163–178.[86] Rahul Mitra and Radhika Gajjala. 2008. Queer blogging in Indian digital diasporas: A dialogic encounter.
Journal ofCommunication Inquiry
32, 4 (2008), 400–423.Proc. ACM Hum.-Comput. Interact., Vol. 5, No. CSCW1, Article 157. Publication date: April 2021. [87] Serena Nanda. 1986. The Hijras of India: Cultural and individual dimensions of an institutionalized third gender role.
Journal of Homosexuality
11, 3-4 (1986), 35–54.[88] David Nemer. 2016. Online favela: The use of social media by the marginalized in Brazil.
Information Technology forDevelopment
22, 3 (2016), 364–379.[89] David Nemer, Shad Gross, and Nic True. 2013. Materializing digital inequalities: the digital artifacts of the marginalizedin Brazil. In
Proceedings of the Sixth International Conference on Information and Communications Technologies andDevelopment: Notes-Volume 2 . 108–111.[90] Alireza Nili, Mary Tate, and David Johnstone. 2017. A framework and approach for analysis of focus group data ininformation systems research.
Communications of the Association for Information Systems
40, Article 1 (2017), 1–21.[91] Sladjana V Nørskov and Morten Rask. 2011. Observation of online communities: A discussion of online and offlineobserver roles in studying development, cooperation and coordination in an open source software environment. In
Forum Qualitative Sozialforschung/Forum: Qualitative Social Research , Vol. 12.[92] Fayika Farhat Nova, Michael Ann DeVito, Pratyasha Saha, Kazi Shohanur Rashid, Shashwata Roy Turzo, Sadia Afrin,and Shion Guha. 2020. Understanding How Marginalized Hijra in Bangladesh Navigate Complex Social MediaEcosystem. In
Conference Companion Publication of the 2020 on Computer Supported Cooperative Work and SocialComputing . 353–358.[93] Fayika Farhat Nova, MD Rifat, Pratyasha Saha, Syed Ishtiaque Ahmed, and Shion Guha. 2019. Online sexual harassmentover anonymous social media in Bangladesh. In
Proceedings of the Tenth International Conference on Information andCommunication Technologies and Development . ACM, 1.[94] Fayika Farhat Nova, Pratyasha Saha, Md Shafiur Raihan Shafi, and Shion Guha. 2019. Sharing Of Public HarassmentExperiences on Social Media in Bangladesh. In
Conference Companion Publication of the 2019 on Computer SupportedCooperative Work and Social Computing . 324–329.[95] Joyojeet Pal, Tawfiq Ammari, Ramaswami Mahalingam, Ana Maria Huaita Alfaro, and Meera Lakshmanan. 2013.Marginality, aspiration and accessibility in ICTD. In
Proceedings of the Sixth International Conference on Informationand Communication Technologies and Development: Full Papers-Volume 1 . 68–78.[96] Yong Jin Park. 2013. Digital literacy and privacy behavior online.
Communication Research
40, 2 (2013), 215–236.[97] Yong Jin Park, Scott W Campbell, and Nojin Kwak. 2012. Affect, cognition and reward: Predictors of privacy protectiononline.
Computers in Human Behavior
28, 3 (2012), 1019–1027.[98] Yong Jin Park and S Mo Jang. 2014. Understanding privacy knowledge and skill in mobile communication.
Computersin Human Behavior
38 (2014), 296–303.[99] Vrushali Patil. 2018. The heterosexual matrix as imperial effect.
Sociological Theory
36, 1 (2018), 1–26.[100] Signe Ravn, Ashley Barnwell, and Barbara Barbosa Neves. 2020. What Is “publicly available data”? Exploring blurredpublic–private boundaries and ethical practices through a case study on Instagram.
Journal of Empirical Research onHuman Research Ethics
15, 1-2 (2020), 40–45.[101] Alan Rubel. 2016. The Black Box Society: The Secret Algorithms that Control Money and Information, by FrankPasquale. Cambridge: Harvard University Press, 2015. 320 pp. ISBN 978–0674368279.
Business Ethics Quarterly
26, 4(2016), 568–571.[102] Noorie Safa. 2016. Inclusion of Excluded: Integrating Need-Based Concerns of Hijra Population in MainstreamDevelopment.
Sociology and Anthropology
Fourteenth Symposium on UsablePrivacy and Security ( { SOUPS } . 127–142.[104] Nithya Sambasivan, Ed Cutrell, Kentaro Toyama, and Bonnie Nardi. 2010. Intermediated technology use in developingcommunities. In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems . 2583–2592.[105] Riddhi Sandil, Matthew Robinson, Melanie E Brewster, Stephanie Wong, and Elizabeth Geiger. 2015. Negotiatingmultiple marginalizations: Experiences of South Asian LGBQ individuals.
Cultural Diversity and Ethnic MinorityPsychology
21, 1 (2015), 76.[106] Morgan Klaus Scheuerman, Stacy M Branham, and Foad Hamidi. 2018. Safe spaces and safe places: Unpackingtechnology-mediated experiences of safety and harm with transgender people.
Proceedings of the ACM on Human-Computer Interaction
2, CSCW (2018), 155.[107] Morgan Klaus Scheuerman, Jacob M Paul, and Jed R Brubaker. 2019. How Computers See Gender: An Evaluation ofGender Classification in Commercial Facial Analysis Services.
Proceedings of the ACM on Human-Computer Interaction
3, CSCW (2019), 1–33.[108] Syeda S. Shawkat. 2016.
Construction of the Hijra Identity . Ph.D. Dissertation. BRAC University. http://dspace.bracu.ac.bd/xmlui/bitstream/handle/10361/7804/16217005_ESS.pdf?sequence=1&isAllowed=y[109] GC Spivak. 1988. Can the subaltern speak? Can the subaltern speak? Reflections on the history of an idea (pp. 21–78).Proc. ACM Hum.-Comput. Interact., Vol. 5, No. CSCW1, Article 157. Publication date: April 2021. [110] Jennifer Jiyoung Suh, Miriam J Metzger, Scott A Reid, and Amr El Abbadi. 2018. Distinguishing group privacy frompersonal privacy: The effect of group inference technologies on privacy perceptions and behaviors.
Proceedings of theACM on Human-Computer Interaction
2, CSCW (2018), 1–22.[111] Sharifa Sultana and Syed Ishtiaque Ahmed. 2019. Witchcraft and hci: Morality, modernity, and postcolonial computingin rural bangladesh. In
Proceedings of the 2019 CHI Conference on Human Factors in Computing Systems . 1–15.[112] Catalina L Toma and Cassandra L Carlson. 2015. How do Facebook users believe they come across in their profiles?: Ameta-perception approach to investigating Facebook self-presentation.
Communication Research Reports
32, 1 (2015),93–101.[113] Kentaro Toyama. 2017. Design, needs, and aspirations in international development. In
International Conference onSocial Implications of Computers in Developing Countries . Springer, 24–32.[114] Zeynep Turan, Hasan Tinmaz, and Yuksel Goktas. 2013. The reasons for non-use of social networking websites byuniversity students.
Comunicar
21, 41 (2013), 137–145.[115] Jessica Vitak and Jinyoung Kim. 2014. You can’t block people offline: examining how Facebook’s affordances shapethe disclosure process. In
Proceedings of the 17th ACM conference on Computer supported cooperative work & socialcomputing . ACM, 461–474.[116] Ashley Marie Walker and Michael A DeVito. 2020. "’More gay’fits in better": Intracommunity Power Dynamics andHarms in Online LGBTQ+ Spaces. In
Proceedings of the 2020 CHI Conference on Human Factors in Computing Systems .1–15.[117] Joe Wallen. 2019. Transgender community in Bangladesh finally granted full voting rights. https://bit.ly/2MjbEqr[118] Yanbo Wang, Qingfei Min, and Shengnan Han. 2016. Understanding the effects of trust and risk on individual behaviortoward social media platforms: A meta-analysis of the empirical evidence.
Computers in Human Behavior
56 (2016),34–44.[119] Bradley M Weisz, Diane M Quinn, and Michelle K Williams. 2016. Out and healthy: Being more “out” about aconcealable stigmatized identity may boost the health benefits of social support.
Journal of Health Psychology
21, 12(2016), 2934–2943.[120] Yan Zhang and Barbara M Wildemuth. 2009. Qualitative analysis of content.
Applications of social research methods toquestions in information and library science
308 (2009), 319.[121] Qinfeng Zhu, Marko Skoric, and Fei Shen. 2017. I shield myself from thee: Selective avoidance on social media duringpolitical protests.
Political Communication
34, 1 (2017), 112–131.[122] Michael Zimmer. 2010. “But the data is already public”: on the ethics of research in Facebook.
Ethics and informationtechnology
12, 4 (2010), 313–325.[123] Diane L Zosky and Robert Alberts. 2016. What’s in a name? Exploring use of the word queer as a term of identificationwithin the college-aged LGBT community.
Journal of Human Behavior in the Social Environment
26, 7-8 (2016),597–607.