Skip to content
Publicly Available Published by Oldenbourg Wissenschaftsverlag April 22, 2021

“Let’s Talk about Gender” – Development of a Card Deck on (Gender) Sensitivity in HCI Research and Practice Based on a Contrasting Literature Review

  • Sabrina Burtscher

    Sabrina Burtscher is a student in TU Wien’s Media and Human Centered Computing Master’s program. As part of her Master’s thesis, she develops a method and tool kit to enable researchers and practitioners to discover gender(ed) aspects and dimensions in their work.

    EMAIL logo
    and Katta Spiel

    Katta Spiel is an FWF Hertha-Firnberg scholar at the HCI Group of TU Wien. They research marginalised perspectives on technology to inform critical design and engineering with a particular focus on trans and intersex as well as disabled experiences.

From the journal i-com

Abstract

Funding bodies in Western societies increasingly require researchers to address gender in their proposals – though often exclusively framed around binary notions. With oppressive power structures being prevalent and persuasive, these seep into current practices of Computer Science and Human-Computer Interaction (HCI) research. However, current curricula rarely provide actors in this space with grounded guidance on gender issues tied to their inquiries. Hence, developing an increased awareness of our societal responsibility towards equity can be challenging without an appropriate starting point. Drawing on a close reading of select literature discussing gender sensitivity in HCI research, we derived practical guidance in the form of recommendations for the design, proposal, conduct and presentation of research. Based on these recommendations, we then present the design of a card deck and initial tests thereof at ditact women’s IT summer university. Our analysis offers a starting point for HCI students and interested researchers to explore questions and issues around gender and to identify how gender relates to their research. This sensitisation may aid them in further reflecting on how they might better account for gendered implications of their work.

1 Introduction

In recent years, public and private funding organisations increasingly expect researchers and (in the case of industry collaborations) practitioners to position their proposals with regards to equity concerns, particularly focused on (binary notions of) gender.[1] As academic funding is more and more driven by third-party money [64], this also relates to research areas not primarily concerned with or particularly educated in matters of equity. The implications of inclusive efforts are tremendous (see, for example, related discussions in the context of machine learning and artificial intelligence [38]), particularly when technological artefacts and algorithms operate on scale and permeate infrastructures [56]. However, while individual Human-Computer Interaction (HCI) researchers have increasingly included marginalised populations (e. g., [19], [59], [17], [53]), most HCI curricula lack concise and actionable guidance in how students might unlearn [28] dominant oppressive paradigms and develop a sensitivity that allows them to assess the impacts of their future research and practice with confidence.

Given the ever-growing corpus of work published on issues on topics of gender pertaining technology research and practice [60], that confidence is difficult to acquire. A query for “gender” in paper abstracts within the Guide to Computing Literature of the Digital Library of the Association for Computing Machinery (ACM DL) matches almost 7500 papers.[2] Such an amount of potential sources can be in itself overwhelming for researchers unfamiliar with the topic who might struggle to identify works relevant to their specific context. Our contribution lies in offering guidance to readers in developing sensitivity as to how to explore the relevance of equity within their work, based on an analysis of different approaches to consider gender in HCI research and practice. In addition, we provide readers with a card-based approach to start a conversation on these issues with collaborators and colleagues.

To synthesise and summarise specialised knowledge regarding gender and technology in HCI, we conducted a close reading of selected papers that function as an introduction into how the field operationalises and articulates gendered aspects. We analysed these papers according to implications for HCI research and knowledge production, derive a set of intermediate level recommendations, and created a card set for structured and collaborative learning. While we, in line with current funding requirements, initially only focused on gender, we expanded our lens to account for the intersectional characteristics (i. e., notions of multi-dimensional oppressive factors having specific effects on gendered experiences such as class, race, sexuality, dis/ability etc. [16], [52]) contextualising the specificity of marginalised experiences. Hence, our efforts, while focused on gender, are intended to spark HCI researchers’ curiosity on what else they might have to consider to adequately address equity in their research.

This paper (and the associated materials) provides guidance around developing (gender) sensitivity for researchers not previously engaged with the topic and to serve as a starting point for further, more situated deliberations in the acquisition, design, conduct and presentation of our inquiries. It offers a synthesising document that combines theories, empirical findings and recommendations for gender-inclusive research practices as related to designing research protocols, acquiring funding, conducting research and presenting it.

To this effect, we start by introducing theoretical contributions to understanding gender stemming from outside as well as inside HCI and revisit existing recommendations for the field. After detailing our methods, we present a close reading of ten HCI publications implicitly or explicitly addressing gender in their work. We then discuss how gender relates to different research contributions in HCI [65] and derive a set of recommendations for gender-inclusive practices before we present the card deck and initial tests thereof with lay people. The paper closes on the limitations of our work with a look towards potentials for future research.

2 Background

Within Gender Studies, diverse conceptualisations of gender and how the notion permeates societies are continuously negotiated. We briefly introduce some of these before attending to related work illustrating how HCI conceptualises and operationalises gender.

2.1 Identity & Intersectionality

We can broadly identify three approaches to gender focusing on essentialism, performance or identity. The most prevalent popular view on gender within society draws on essentialism. It assumes a strictly binary, immutable, complementary and fixed set of genitalia that neatly align with gender, learned behaviour and social roles [5]. In rejecting this alignment, scholars have started to distinguish between sex (assigned at birth) and gender, with gender becoming a mostly aligned reiterated performance. However, the sexed body still determines how gender is supposedly performed and any performance not following suit marks deviant behaviour [43].

Both of these views provide a biologically deterministic perspective on gender where the assignment of genitalia to a specific sex implicates gender. More recently, theorists have started understanding gender (and sex for that matter) as a social construct going far beyond a simple binary [21]. Assuming a gendered identity consciously constitutes an exercise of self-determination [32]. Due to the “sexing” of the body [21], there is no clear distinction between presumably ‘biological’ sex and ‘performed’ gender; instead, both are continuously and consciously upheld within societal practices and customs. Our work is informed by an understanding of gender as self-determined and independent of genital makeup.

Inequality, though, operates not exclusively on gender. The concept of intersectionality articulates how different marginalised identities might overlap and amplify specific modes of oppression. Crenshaw illustrates this vividly for black women. “Feminist efforts to politicize experiences of women and antiracist efforts to politicize experiences of people of color have frequently proceeded as though the issues and experiences they each detail occur on mutually exclusive terrains” [16]. Hence, beyond gender, further markers of identity are likely similarly relevant and important. If gender and other factors remain implicit, researchers risk operating within unmarked norms of, e. g., masculinity, whiteness, ableness and (often upper) middle-class [7].

2.2 Related Work in HCI

With the recent surge of publications discussing how gendered dimensions shape HCI research, we see increased activity regarding gender recommendations in the forms of analytical reviews, software analysis and guidelines.

2.2.1 Analytical Reviews

Roig-Maimó and Mas-Sansó show how researchers conducting user studies need to be aware of problematic generalisations if they focus solely on mathematical parameters of significance without also accounting for the distribution of gender among their participants [48]. Albeit limited itself by a binary conceptualisation of gender, the authors suggest researchers address limitations due to participant samples. With this binary view, their work aligns with most of traditional HCI research. Stumpf et al. recently published a conceptual review on gender-inclusive HCI research with a specific focus on cognitive and behavioural works as well as dangers of stereotyping in technological design [60]. However, even though the authors explicitly attend to issues surrounding a dominant binary notion of gender, their approach risks essentialising gender (cf. [36]) and does not operate from a notion of self-identification.

Our work takes a prior literature review by Schlesinger et al. as its foundation point due to its intersectional approach [52]. The authors identified how existing research is narrowly focused on single markers of identity without structurally attending to overlapping and amplified modes of oppression. The authors studied how HCI research identifies and/or classifies target populations and research participants in a large corpus of 140 papers published between 1982–2016 at the ACM CHI Conference on Human Factors in Computing Systems (CHI). Particularly, they investigated diversity and intersecting categories of identification along gender, ethnicity, race, class, and sexuality (though leaving out, e. g., age, localisation and disability).

2.2.2 Software Analysis

As gender comprises a generally pervasive structure within Western societies, Burnett et al. identified it as a relevant factor in software [10], even if not explicitly concerned with gender [23]. The authors and further collaborators created an application that systematically steps through source code to identify instances in which gender bias occurs [63]. The approach relies on gender-inclusive personas [39] while trying to avoid stereotyping [31]. Recently, some of the authors were also involved in an attempt at generalising the method to other issues of equity [41]. However, this approach requires not only that researchers are already aware and understand gender (or other) identity factors appropriately, it also assumes that these issues can be feasibly addressed by a cognitive walk-through. This can comprise a first step in aiming at (gender) equity in software, however, it presents a formulaic approach to an issue that requires situated sensitivity. Our work presents a potential starting point to developing knowledge relevant to use such methods critically and effectively.

2.2.3 Existing Gender Guidelines

In the context of qualitative settings, Rode calls for more reflexivity (also on gender) in HCI research [47] and Brulé and Spiel developed considerations on how to negotiate gender (and disability) between researchers and participants in participatory design [8]. Jaroszewski et al. conducted research in two different contexts, tumblr members and fantasy football players, using the same form to inquire into participants’ gender: a free form field [33]. The first group provided a range of unique genders and praised the approach whereas the second group was hostile towards the freedom of expression. From that, the authors derive a generalised approach for how to inquire into gender, which has been further refined, including a call for situated nuance [55]. The HCI Gender Guidelines [51] stem from a grassroots initiative of several HCI researchers and provide hands-on guidance for gender equity in writing, research and the organisation of academic events. Our work supplements this approach based on lived experience and activist knowledge by analysing existing works and drawing higher-level recommendations from peer-reviewed publications as different modes of assessing these issues lead to stronger recommendations across them through triangulation.

2.2.4 Research Gap

All of these approaches show that researchers have started attending to gendered issues in their research. However, most of them require a certain level of understanding the issues at hand to apply them. We could not identify work aiming at researchers to develop their sensitivity as it comes to equity around gender and other markers of identity with an intersectionally informed lens. Hence, our work offers a theoretical as well as applied starting point allowing them to probe their research for relevant matters of (gender) equity.

3 Close Reading Approach

We selected a smaller subset of seven papers from Schlesinger et al. [52] as instances covering a range of topics concerning gender and HCI while also exhibiting geographical diversity as it relates to authors’ institutions and research environments. Augmenting these, we added three further papers published since 2017. We have not specifically selected papers addressing methodological guidance regarding gender sensitivity for HCI research (such as the ones in the Related Work Section) but instead aimed to develop our recommendations from works that have implicitly or explicitly dealt with populations which made gender sensitivity relevant to their research and covered a range of different contribution types [65]. Hence, our final close-reading corpus includes the ten papers shown in Table 1.

Table 1

Characteristics of papers part of our close reading corpus. Contribution types have been identified according to [65]. Papers indicated with * have been added outside of Schlesinger et al.’s corpus [52]. DA stands for domestic abuse.

Paper population locale context focus treatment contribution
Blackwell et al. [4] LGBT parents USA social media implicit self-identification empirical
Karuei et al. [34] general CA/DE wearables explicit binary/biology artefact
Otterbacher [45] general CY crowdsourcing explicit binary/external empirical/dataset
Clarke et al. [12] DA survivors UK photo-sharing explicit binary empirical/method
Ahmed et al. [1] general BD online platform explicit binary light empirical/method
Haimson et al. [25] men having sex with men USA dating ads implicit ambiguous empirical
DiSalvo et al. [18] black men USA game development explicit ambiguous empirical/theory
Keyes [35]* trans people USA gender recognition explicit self-identification survey/argument
Metaxa-Kakavouli et al. [42]* general USA web interfaces explicit self-identification empirical
Fernandez and Birnholtz [22]* trans people USA dating platforms explicit self-identification empirical

Figure 1 
            Comparison between systematic and contrasting literature reviews.
Figure 1

Comparison between systematic and contrasting literature reviews.

We analysed these papers using close reading as our method [40]. This is an interpretive and qualitative approach with which we, contrary to systematic reviews, do not aim to illustrate the breadth of what is available. Instead, we sampled papers according to diversity criteria and engaged with them deeply to provide a coherent contextual analysis. This contrasting review allows us to analyse distinctions between different approaches and the motivations behind a smaller range of publications (see also, Figure 1). Hence, we conducted a contrasting review with an in-depth analysis of a decidedly non-exhaustive list of selected papers chosen to represent difference among them.

Each of the instances within the corpus was analysed regarding gender aspects in research questions, methodologies, and language used. In that, we were guided by the overall question: “What can we learn from this?” to ensure our work remains relevant for HCI researchers and practitioners not yet familiar with the topic.

As our work is interpretive, rigour can be established within a notion of partial perspective [29], meaning that we do not claim generalisability of our findings, but rather present an informed analysis leading to a set of situated recommendations. Both authors inhabit marginalised positions within gendered power dynamics (as woman or nonbinary person), which implicates our analysis [14]. Additionally, all authors are white and culturally tied to majority perspectives within central Europe, rendering our analysis one that remains outside of a lived experience of, e. g., race [44] or forced migration [61]. One of the authors further identifies as disabled and neurodivergent. Making these positional markers visible allows readers to reflect on the potential limitations of our analysis and contrast their position. We invite and encourage further work from different perspectives [30].

4 HCI and Gender: Findings from the Close Reading

Here, we discuss each of the corpus papers in detail to illustrate our close reading. We have set best practices in italics which serve as inspiration for researchers and practitioners, and are the base for the recommendations we describe in Section 6.

4.1 Complexities of Disclosure in Blackwell et al., 2016

Blackwell et al. were interested in social media use patterns of LGBT[3] parents [4]. They asked which platforms their participants engaged with, which details they shared and how they restricted their audiences. While this work contains no explicit focus on gender (e. g., by choice of keywords), gender is implicitly relevant. In a series of interviews with 28 participants, they identified three themes relating to (1) continuously gauging safety of disclosure around particular audiences (p. 614), (2) the external attribution of activism to lived experiences (p. 615) and (3) detailed deliberations regarding privacy of self and family members (p. 616).

Methodologically, the authors point out the relevance of respecting participants’ choice of self-identification regarding gender and sexual orientation. Their participants were recruited through peer groups on a range of social media platforms (e. g., Twitter, Facebook). Even though the context was social media use, the authors refrained from seeking out their participants’ online profiles to respect their individual privacy and dignify the desired representation towards the researchers. From their results, we additionally take that researchers should communicate a non-judgemental safe environment to participants, create a space in which it is comfortable to share personal experiences and refrain from probing deeper where participants do not readily disclose aspects of their lives.

4.2 Diversity of Embodiments in Karuei et al., 2011

Karuei et al. were interested in the sensitivity of different body locations to vibrations, how they are affected by movement and whether visual workload, participants’ expectations or gender played a role in task performance and preferences during the use of tactile displays [34]. The authors identified differences in the rate of signal detection as well as response time when analysing according to participants’ gender (which they take as an indicator for body fat distribution). In this, they follow a biologically essentialist understanding of gender which becomes further convoluted as the authors point out themselves that the results are neither consistent nor significant.

While the authors attend to gender explicitly, they miss out on fundamentally gendered issues. For example, they identify that the “thigh was among the least effective and least preferred stimulus site we tested; and yet, front pocket is a common location to stow a mobile device, particularly for men” [34, p. 3274]. In this context, differences in gendered clothing (such as the lack of or diminished size of pockets in clothes aimed at women [9]) and their influence on preferences are not reflected upon. As different embodiments are not specifically discussed, we assume that they operate from an unmarked norm of presumably able-bodied participants.

Methodologically, the authors ensured that participants wore the same type of close-fitting clothes (sportswear) to reduce the effect of different clothing styles. However, they do not disclose how they inquired into gender and only report it for eight participants as male. From their results, we can take that gender only poorly transfers to a concept of specific embodiment and that body characteristics should be explicitly and directly analysed instead of inferred from identity characteristics.

4.3 Development of Gender Bias in Otterbacher, 2015

To crowdsource textual metadata for images, Von Ahn and Dabbish developed ESP,[4] a Game With A Purpose (GWAP). Players are randomly paired and challenged to agree (under a time constraint) on as many labels as possible for a given image [62]. The game requires players to create more and more specific labels once a label has reached a threshold of popularity within the broader player community. Hence, players contribute to the accumulation of big database structures, but also introduce their personal and collective biases into the systems and the algorithms learning from them.

Within this context, Otterbacher studied how bias, particularly gender bias, might be introduced and piped into biases in automated decisions drawing on the data [45]. They operate from a binary notion of gender that can be externally assigned and focus on adjectives, especially strongly subjective ones, and their polarity and frequency in over 33.000 images of people [45, p. 1958f.]. They state that descriptive adjectives are commonly used for ‘women’, whereas active nouns refer more to ‘men’, with a heavy tendency to assign heteronormative descriptors in both cases [45, p. 1961].

Methodologically, the author cautions readers towards a pre-existing bias in online pictures of people as they are pre-shaped by gendered societal expectations. Further, Otterbacher reflects on the limits of assigning gender to pictures, particularly reduced to binary labels. From their results, researchers can take guidance on ensuring equivalence in describing people regardless of gender, i. e., using the same type of words (adjectives, nouns etc.) and classifiers (e. g., jobs, colours etc.) to consistently articulate the same information across participants, case studies, personas or stories.

4.4 Agency Around Trauma in Clarke et al., 2013

Domestic violence is often conceptualised as a gendered problem within a binary concept, where ‘women’ are dominantly seen as survivors and men as perpetrators [3]. However, as class [2] and culture [13] play an additional role, disabled people are overall more at risk to abuse [15] and LGBTQ2IA* people face additional barriers to access services [11], researchers increasingly call for an intersectional lens (see Section 2.1) [46].

On that backdrop, Clarke et al. worked with six women who had recently left an abusive partner [12]. Specifically, they investigated the potential of photographs as a means to (re-)gain agency over survivors’ narrative through storytelling. Their work was conducted over the span of several months and design iterations.

Methodologically, the authors opted for building careful relationships with participants. This included attending to participants’ needs and availability, i. e., not holding workshops during times schools were closed as many of their participants had care duties for young dependants. Additionally, they collaborated with topical professionals to ensure that their research was safe and appropriate. They also disclosed their perspective and motivations on the topic during the sessions to create a trustworthy setting. These close relationships were embodied within the prototypes and deemed essential for the designs to be meaningful and relevant.

4.5 Attending to Everyday Violence in Ahmed et al., 2014

Ahmed et al. created a platform where everyday experiences of sexual violence and harassment could be reported, shared and collectively discussed [1]. They employed user-centred design including surveys, interviews and focus groups. Participants indicated a strong need to share their experiences and find community to work through feelings of shame, sadness and regret as well as defiance and anger to facilitate positive change.

Methodologically, the researchers acknowledge the high prevalence of sexual harassment and violence as well as their gendered impact. Understanding the damaging effect repeated violations might have on some participants can contextualise, e. g., work discussed previously [34]. Hence, Ahmed et al. specifically sought out more privileged participants who might feel safer sharing experiences associated with shame. Attending to economic factors shaping research participation allowed the researchers to more critically reflect on the limitations of their work.

4.6 Language on Sexual Health in a Haimson et al., 2014

Haimson et al. studied the language men who have sex with men (MSM)[5] use to pass on sexual health information in online personal ads [25]. Their intent was to inform linguistic choices within system architectures as well as shedding light on representations of “health conditions, preferences, and activities” [25, p. 1622f.].

Methodologically, the authors use self-identification for sexual preference (analogous to [4]). As they did not directly work with participants, their positional disclosure of two coders identifying as gay men aids the contextualisation of their findings. Additionally, they categorised age groups along life experiences instead of conventional age clusters. However, they do not explicitly disclose how they operationalise gender. As they only talk about ‘men’ without further specifying, they implicitly rely on the dominant binary model.[6] Hence, we suggest that even in cases where the focus lies on one gender only, researchers should briefly reflect the inclusion and exclusion criteria for their understanding of which populations they refer to.

4.7 Intersectional Identities in DiSalvo et al., 2011

DiSalvo et al. developed Glitch Game Testers, a program introducing low-income high school students to computing through video games [18]. Their focus lies on African American[7] men, who are less likely to be interested in technology even though they are passionate about video games. DiSalvo et al. root this in stereotypical representations of “geek” or “nerd” as white young men disregarding personal appearances, whereas African American men tend to construct their identity through athleticism and body work. After participating in a summer programme interning in the games industry, participants were more likely to see themselves and their peers as technically adept.

Methodologically, researchers can take from this research an awareness of how racialised and gendered stereotypes amplify intersectionally (see Section 2.1) [16]. The work (together with more recent examples, e. g., [44]) challenges us to reflect on the unmarked norm of Whiteness in computing. DiSalvo et al. also show how computing professions are not just inaccessible for (white) women. Hence, in thinking about equity and gender, researchers need to critically interrogate their viewpoints and allude to their standpoints [30] to appropriately contextualise the knowledge they produce including the limits thereof.

4.8 Automating Stereotypes in Keyes, 2018

In a theoretically driven analysis, Keyes identified how Automated Gender Recognition (AGR), the “automatic computational identification of a person’s gender from photographs or videos” [35, p88:4] is inherently trans-exclusive due to the presumption that gender could be assigned externally. More recent work has subsequently shown how computers fail to assign gender in a way that aligns with a person’s gender reliably [50]. Keyes “found a remarkably consistent operationalisation of gender within AGR research. Almost every paper with a focus on gender, and many of those without, treated gender in a way aligned with the traditional [exclusively binary, immutable and physiological] view” [35, p88:7]. According to Keyes, such findings likely hold for racialised contexts as well [35], rendering such approaches inherently prone to risk amplifying pre-existing inequalities.

Methodologically, this paper suggests to generally avoid automated identity assignment. Researchers should not externally define aspects of a person’s self-representation. They also suggest to frame gender explicitly and trans-inclusively and call for appropriate resources for researchers. Since then, some such resources have been developed, for example, the HCI Gender Guidelines [51] or specific recommendations for accounting for gender in survey design [55].

4.9 Interfacing in Metaxa-Kakavouli et al., 2018

Metaxa-Kakavouli et al. created two webpages for an introductory course in Computer Science with differing design and layout. In an experimental setup, men did not distinguish much between ‘masculine’ and ‘gender-neutral’ designs, contrary to women [42]. Subsequently, women had lower degrees of belonging in the context of stereotypically masculine designs. The authors explicitly mention that no participants reported to be nonbinary.

Methodologically, the paper operates weakly within a gender model of self-identification. As cited related work predominantly uses a binary model of gender, the authors take care to denaturalise previous work and point out different conceptualisations of gender across cultures, though participants were restricted to US residents only. From their results, researchers can reflect on how the designs of prototypes and software for studies might be perceived as stereotypical within their test population to appropriately control for gendered impressions.

4.10 Identity Disclosure in Fernandez and Birnholtz, 2019

Trans people (particularly trans women of colour [22, p226:6]) can rarely be certain whether it is safe (i. e., not leading to slurs, harassment, hate speech and/or violence by another (cis) person) to disclose their gender in online contexts and have developed strategies that allow them to assess the safety of a given situation. In one-on-one interviews with 20 trans people using dating platforms in the United States of America, Fernandez and Birnholtz identified two forms of disclosure: direct (clear and open) and indirect (with the option of plausible deniability) [22].

Methodologically, the authors recruited through community-based platforms (e. g., an LGBTQI2A* community centre, trans-specific mailing lists) but also employed targeted Facebook advertisements. However, this strategy lead to a disproportionate representation of White trans people volunteering. Participants’ preferences were taken into account for the interview setting. Additionally, Fernandez and Birnholtz explicitly asked about the pronouns each participant had and made clear these would be used in publications. The authors also reflect on their status as cis researchers and identify themselves as allies.

5 Discussion: Gender in Different Types of HCI Contributions

Through our analysis, we illustrated how gender can affect research in many forms regardless of whether gender comprises an implicit or explicit research parameter. We could show that across a range of different publications, an essentialised binary model of gender still is prevalent; with the model of self-identification becoming more prominent in recent works. We welcome this drive towards self-identification in research and expect it to carry over to the implementation and actualisation of (gendered) technical infrastructures. This becomes all the more relevant seeing how digital infrastructures increasingly encode gender (quite literally). They further become substantial – if volatile – resources in finding and articulating gender identities [26]. We now illustrate how gender might play a role in different types of HCI research contributions and delineate a set of recommendations for the design, funding, conduct and presentation of research.

Within the HCI community, Wobbrock and Kientz have identified a range of research contributions [65], and not all of them involve direct interaction with people. Still, gender plays a role as the field concerns humans and their inherent messiness at the core. This can either implicitly be the case (e. g., through representation) or explicitly articulated (e. g., in tackling a gender issue). We now step through contribution types identified by Wobbrock and Kientz [65] and provide brief examples on how to attend to gender in each of them respectively. In doing so, we reference back to the papers within our corpus to illustrate how they inform our analysis along contribution types.

  • In empirical research, gender influences who can participate in experiments, and who is excluded. This affects, for example, gendered care responsibilities (cf. [12]), or gendered differences in recruitment strategies (cf. [4], [25], [18], [22], [42]). Participants and their gender then also limit generalisability (in post-positivist research) or transferability (in constructivist approaches).

  • The artefacts or systems developed in HCI projects may be easier or harder to understand for people with different levels of affinity for technology, or varying previous topical knowledge in an area – both possibly influenced by gender roles and gendered stereotypes. Artefacts or systems also might record gender (e. g., through forms). In such cases, it becomes relevant to reflect on how gender is encoded and operationalised (cf. [34], [42]).

  • For methods to be applicable by others, it is paramount to report on how gender was operationalised and included over the course of development, and why. Such deliberations should extend to reflections on how appropriate a method is for otherwise marginalised populations, e. g., disabled people. Some methods inherently exclude, for example, people with a low or no income due to the reliance on technical infrastructures (cf. [1]).

  • Theories, according to Wobbrock and Kientz, are “not simply observing that but explaining why” [65, p. 41, emphasis in the original]. As we have discussed previously, gendered roles and intersections of identities have extensive effects on people and technologies. Theoretical contributions need to be aware of potential gendered influences, to stay relevant to practical applications. If work does not directly apply gender theory (cf. [35]), marginalised populations may be casually included, e. g. in providing imagery with diverse body types (compare e. g., outside of our corpus, [54]).

  • Following Wobbrock and Kientz, datasets ought to be representative to be useful for testing, measuring, analysis and knowledge construction [65]. Gaps in datasets as they pertain to gender, race, disability and economic status may not only cause nuisances due to incompleteness, but lead to and further support systemic discrimination (see for a further discussion on these tendencies, [20]). Hence, in creating a dataset, researchers can improve its usefulness by reflecting across these dimensions (cf. [45]).

  • One of the goals Wobbrock and Kientz describe for meta-analyses and survey contributions is exposing trends and gaps in current literature. Here, authors can explicitly reflect on several reported characteristics of participants in given studies, including gender (cf. [35]). Even in cases where the focus does not lie on a specific population, analysing how different people are included in the production of knowledge is paramount to identify exclusionary research practices and existing research gaps.

  • Finally, in essays and arguments, the rigour lies in probing the argumentative flow for edge cases and strengthening it by thinking those through. Gender and other identity factors can provide an edge case on which to contrast an argument and increase its cohesion. This is not relegated to constitute an “interesting starting point” but should be seen as a productive engagement with being accountable towards marginalising tendencies within existing research (cf. [22] and, outside of our corpus, [44]).

As our source material touched, at least partially, on all of these contribution types, we argue that an analysis of these works offers the opportunity for recommendations that can guide HCI researchers, students and practitioners further.

6 Recommendations Along Different Research Stages

From our results and the ensuing discussion, we have derived a set of high level recommendations on accounting for gender while designing research, acquiring funding, conducting research and presenting research within HCI. Before we go into the details, we caution readers that creating sensitivity to matters of oppression falls into the realm of slow science [58]; it takes time to attune to experiences that might not resonate with our own. We encourage our readers to dig further, read some of the texts presented here or others in depth and continuously reflect on their practices. Knowledge production attending to power imbalances, oppression and marginalisation comprises a process of ongoing atonement and negotiation.

6.1 Designing Research Protocols

Before funding can be sought, before research can be conducted, researchers have to design their research protocol. To build a strong basis for all subsequent work, researchers should reflect on gender sensitivity early on and be aware of power mechanics affecting potential participants, target populations or informants.

6.1.1 Articulate Gender Explicitly

In our background section (Section 2.1), we detailed three different approaches to understanding gender, namely essentialist, performative and identity based models. Similar to Keyes [35], we encourage researchers to articulate their understanding of gender upfront. This allows them to later communicate ensuing implications more clearly and allows them to reflect on otherwise potentially implicitly made inferences of gender on their work.

6.1.2 Plan for Individual and Collective Reflection on Positionality

To identify potential presumptions and stepping stones influencing a given research project and its proponents, we suggest planning for individual and collective reflection on positionality and personal perspectives early on. Such a plan can be understood as a living document that adapts to emerging needs and practices. Examples for such ongoing reflection can be a shared or private diary for every individual involved, regular meetings with documentation, or informal check-ins with an accountability structure [47]. Different modes might be suitable for different researchers. However, thinking about potential ways to capture the research process from a personal perspective early on makes disclosing and discussing positionality in the presentation of research easier later on (compare [13], [22], [4]).

6.1.3 Probe Methods for Accessibility and Sensitivity

Researchers build their work on prior concepts and research. However, all predecessor technologies have been developed in and for certain contexts, so sometimes adaptations are necessary. This may relate to gender in that body related work might have been tested only on a homogeneous population, e. g. able-bodied cis men. While mishaps may happen and can be met with an honest interest to do better, reflecting on how accessible methods are in a specific context allows researchers to understand more about the limitations of their work. As an example from within the corpus, Karuei et al. ensured that participants wore the same type of clothing as to not occlude bodily characteristics by gendered clothes [34]. We suggest to take deliberations on gender as a starting point for identity factors more broadly to further develop and sharpen one’s sensitivity for marginalising tendencies in research.

6.1.4 Deliberately Assess Exclusions

Any research will exclude some participants and include others. Deliberately assessing exclusions upfront can mitigate situations where this happens inadvertently and in opposition to a given research question. For example, Ahmed et al. acknowledge that the participant base which felt safe and secure enough to talk about topics associated with stigma was fairly small and privileged [1]. Another way to think through this might be by attending to unmarked norms as DiSalvo et al. did this for Whiteness [18].

6.2 Acquiring Funding

To fund research activities, researchers need to engage more and more with funding bodies awarding finances competitively. Especially those tied to political or governmental entities, e. g., the European Union or national science funds such as the Deutsche Forschungsgesellschaft (DFG), distribute available funds according to several criteria. Being themselves funded largely through taxpayer money, they need to attend to the usefulness of research to advance societies to become more just and equal. Hence, they often ask researchers to allude to impacts on gender norms, rules, or stereotypes. Here, researchers have the opportunity to not just push funding bodies into a more progressive position (while simultaneously presenting their projects as progressive as well) but also to extend knowledge about more substantial issues regarding equity to reviewers and boards.

6.2.1 Check Expectations and See How They May Be Extended

When calling for submissions, funding bodies often publish a request to attend to gender issues, often with stereotypical binary ideas tied to an essentialist view on gender (see Section 2). Researchers might inspect the proposal and review limitations thereof. In constructing a statement that attends to issues of intersectionality and equity more generally, researchers will need to argue how their expanded attention to these issues falls under the umbrella of what the funding body expects. Such an understanding makes, e. g., the work on Black men in computing detailed above possible [18].

6.2.2 Understand Proposal Writing as a Teaching Opportunity

As proposal reviewers might be unfamiliar with gender as a self-determined identity, such approaches need to be communicated accessibly across disciplines. As such, proposal writing becomes an opportunity for educating on these topics by providing an excellent example. Hence, in detailing one’s assumptions on gender and equity, clarifying specifically why and where this matters within the proposal creates a best practice. In that regard, being precise in language and acknowledging potential biases in data sets (whether pre-existing or created; cf. [45]) guides reviewers’ understanding of the subject matter.

6.3 Conducting Research

Making research practices more inclusive requires passionate labour and comes in many different shapes and sizes. However, by conducting research accessibly and in line with the needs of marginalised participants, the resulting knowledge has the potential to be more thorough and reflected along its limitations.

6.3.1 Follow Participants’ Choices in Identifying Them

As Fernandez and Birnholtz as well as Haimson et al. illustrated, it is a basic sign of respecting participants to explicitly ask for and use their pronouns [22] (even if not directly interacting with them [25]). Valuing their choice on how they want to self-identify without judgement [4] implies staying close to the data and honest towards participants. For this, researchers need to provide participants with the space and freedom to express themselves in the manner they please. In interviews, this can be managed during dialogue, and smaller scale studies, can provide free form fields for disclosing gender. In larger scale studies it then becomes relevant to know the dominant approach to gender within the target population in order to appropriately design questions into gender identity [55]. We also urge researchers to explicitly offer participants an option not to disclose their gender [55].

6.3.2 Attend to Different Needs and Preferences

Starting with recruiting, researchers can familiarise themselves with the cultural environment a target population might have. This can also help in recruiting through peer groups [4] or community-based platforms [22]. While it is understandable that researchers are focused on their questions and tasks, it helps to be mindful about participants’ lives outside of a given project context. This begins by actively including marginalised groups and goes to offering lunch during workshop days and tweaking the project schedule to accommodate for care responsibilities [12], as well as making counselling available for participants [1], [12]. Additionally, flexibility in the choice of research setting (e. g., conducting interviews either through chat or via telephone) creates an inclusive research environment [22]. To further create a safe and non-judgemental atmosphere, we suggest to refrain from probing deeper into topics where participants do not readily disclose aspects of their lives.

6.3.3 Seek Critical Feedback

Critical feedback can be sought in three ways, as seen in the work by Clarke et al.: by building long lasting engagements and careful relationships with participants (to encourage opposition to ideas), collaborating with topical professionals and actively disclosing one’s own perspective and motivations to participants [12]. However, all of this takes time to appropriately account for where we can do better [58].

6.3.4 Actively Look for What’s Missing

One of the hardest things to do is to be aware of the gaps in ideas, data, and plans we have developed ourselves, and then set out to mend them. Finding those gaps can be difficult if only discussed within one’s own domain; in lieu of (or addition to) critical feedback, researchers can attune themselves to actively attend to what is missing in their research (akin to hunting software bugs in engineering practices). Sometimes, this may result in the realisation that, e. g., not using gender as a variable, might be more appropriate and a different, more nuanced, view on people (cf. [35]) or their bodies (cf. [34]) could be more meaningful to a given research question or context.

6.4 Presenting Research

In writing and talking about research, we need to exercise reflexivity and thoughtfulness on the implications of our research as it pertains to gender and other identity factors.

6.4.1 Provide Appropriate Context Information

To allow readers and fellow researchers to build on existing work, detailing and explaining choices is paramount to provide appropriate context. This pertains motivating choices to include and operationalise gender within a given research context (cf. [35]). We suggest to discuss the origin of pre-collected data, if present, and analyse it for incompleteness – for example, workplace environments likely change across time. Here, we further recommend to acknowledge the specific locale for knowledge, understanding the limits inherent in this and extending solidarity to diversify knowledge constructed in different geographical contexts (see also, [37]). Disclosing positionality can be relevant contextual information (see [25], [22], [4]). As we pointed out above in referencing prior work [29], [8], disclosing identity allows readers to contextualise the viewpoint of given works, which – contrary to dominant practices – remains relevant for post-positivist as well as constructivist approaches. Additionally, if researchers are themselves part of the groups they analyse, this comprises another point of explicit representation and provides an additional level of lived expertise with a given subject matter [35].

6.4.2 Reflect on Representation and Prior Work

Representation can refer to examples, pictures and imagery used. Here, researchers can reflect on unmarked norms embedded in their visualisations and textual examples and diversify where possible. As most prior work operates from a binary, essentialist understanding of gender [60], denaturalising prior work can include broadening representation and attending to cultural differences where appropriate (see [42]).

6.4.3 Choose Mindful Language

The use of inclusive and precise language may avoid perpetuating exclusionary stereotypes. For example, we suggest to use singular they when referring to “abstract” individuals (as done e. g., by [35]) and to state all options and choices when presenting statistics (e. g., “45 % identified as women, 44 % as men, 7 % as nonbinary, 4 % chose not to disclose their gender; participants could select multiple options”). Further we recommend asking for and using participants’ self-descriptions and pronouns [4], [25], [22], as well as gathering feedback regarding language used by representatives of the target population.

7 Going Further in Collaboration – The Card Deck

Based on the recommendations drawn from our close reading and analysis and described in Section 6, the first author of this paper developed a card deck intended to be used by project teams and/or collaborators to initiate, facilitate and encourage discussion and reflection regarding (gender) sensitivity in HCI research and practice – and to ease the dissemination of related knowledge.

7.1 Design Goals

Given the abstract characteristics of our recommendations above, we expected conversation cards to offer more flexibility for reflection as facilitating the start of a process (instead of merely a static set of suggestions) and being more conducive to a variety of group settings. Individuals could use the card deck by themselves, while groups have several options for engaging with the material ranging from studious to playful. The basic idea for the card deck was for a project team to sit down together and, instead of going through a checklist, try to answer the questions and discuss the concepts presented on the cards. For example, the group organising a university lecture could use the cards to assess how they could include and acknowledge gender dimensions in their class and exercises. We assumed that the cards’ haptic nature would offer an additional “hands-on” aspect to the process of discovering and discussing new concepts and angles.

The card deck intends to offer HCI researchers and practitioners a range of different perspectives to investigate their project from, independent of the specific context of inquiry. When used in a group setting, e. g. distributing one card per participant, each participant will have to assume a different position from which to discuss a given project from. Each card consists of a text snippet offering information and proposing actions to take. To increase the accessibility of the content, we decided to use plain German as the language of the cards, removing technical terms common within Gender Studies as to not erect artificial barriers to said content. A translation to English is currently in progress.

7.2 Card Development

The recommendations with their verbose and brief descriptions were used as the basis for the text and headers for the first draft of cards. During the development of the cards we used Libre Office Writer, choosing an A5 page format and split-page setup to come close to the intended size of cards. The text then was (a) translated to a more vernacular German and (b) cut down so each recommendation would fit on one card. The first author decided to use a less academical German because the workshop context for initial testing took place at a German-only venue,[8] and because people should be able to use the cards regardless of their level of education or research focus, e. g. facilitating use by non-academic practitioners. This design choice was driven by our commitment to accessibility and a desire to bring the topic closer to those who until now had little to no contact with it, as described in Section 2.2. Subsequently, we planned the initial testing workshop as an opportunity to receive feedback from people who were unfamiliar with our work.

Another difference between the original recommendations, and the deck of cards based on our goal of accessibility for topical novices lies in the amount of footnotes. Many cards feature one or several footnotes which mostly cover recurring concepts and terms (cis, trans, intersectionality) to establish a baseline for everyone engaging with them. We wanted this crucial information to be available on each card, but without cluttering the main text too much. If someone in a group using the cards does not know a term or phrase, they do not have to publicly acknowledge uncertainty to the person holding onto the description. In addition, having the information directly on the card means that readers would not have to handle multiple documents or negotiate the use of the description with other group members.

In addition to the categories we had already established, we identified further categorizations by taking on a closer look at thematic rather than temporal connections. The guiding question here was “How, apart from seen over the life cycle of a project, could these cards be connected?”. The temporal categories in the initial design were color-coded into the title of each card, and later in the background of the title to increase legibility. The thematic categories are displayed using icons on the top left of each card (see also, Figure 2).

Figure 2 
              Schematic of the cards detailing different structuring elements (icons and colours) as well as text built up from high level recommendations via guiding questions and explanations to footnotes.
Figure 2

Schematic of the cards detailing different structuring elements (icons and colours) as well as text built up from high level recommendations via guiding questions and explanations to footnotes.

On the technical side, the base document for the cards is a Libre Office Writer document. Writer was chosen to ensure ease of future editing and potential community contributions. The WYSIWYG character of the program makes alterations like a switch from single-page A6 to double-page A5 immediately visible and better comparable. The initial online version functioned as a proof-of-concept and consisted of screenshots from the PDF generated by Writer. A PHP script would return a random image drawn from a folder with the screenshots. In this form, the card drawing tool was not accessible for people who use screen readers, and the cards did not scale on different screen sizes. To increase the accessibility of the website, we created a script that would convert markdown files to HTML files. Hence, in the final version, the cards are screen reader accessible, and mobile-friendly – and additionally, each card on the online version is located within a separate HTML file with a designated URL, so people can link to and share specific cards. The online version can be accessed (so far, only in German) at https://div.uber.space. To assess the suitability of the cards as physical materials, the first author verified font and overall size through printouts to make it feasible to provide the PDF online for interested parties.

Figure 3 
              Selection of cards for testing font size and suitability for physical workshops.
Figure 3

Selection of cards for testing font size and suitability for physical workshops.

7.3 Initial Testing

In September 2020, we had the chance to present this work in the form of a half-day workshop at the ditact Women’s summer IT studies. Usually, these courses take place in Salzburg, Austria- due to COVID-19, however, most (including our workshop) were moved online.

Initially, we planned to evaluate the card deck, and with it our results, with focus groups of research and teaching staff at TU Wien, but after talking about this work to general and informed audiences (e. g., at 36C3 in Leipzig[9]), a member of the ditact organisational team encouraged the first author to hand in a corresponding proposal. We opted for the workshop format rather than a talk, because we saw the potential of evaluating the cards and gathering critical feedback on our findings and inferences. This would provide us with the option to test our main assumption: that the cards can be used in an interactive setting by encouraging discussion and reflection on (gender) sensitivities in HCI research and practice. As the target group for courses at ditactare both students and professionals working in IT-related sectors, we kept both the title and the course description practical and pragmatic: “Inclusive project design – diversity beyond personas”.[10]

Due to the change to an online format, we prioritized the development of the online version of the card deck. For the workshop itself, we used Zoom. The workshop attracted five participants from various backgrounds. For introductions, the first author asked each participant what their current connections to diversity and inclusion were, and whether they had any current projects where they would like to apply the card deck. We discussed what their experiences with diversity and inclusion measures were in the past, and followed that with a presentation and discussion of the development and content of the card deck, and offered space and time for questions. Thereafter, each participant had the task to draw a card for themselves and (on their own) try to find ways of applying it to their current project (or their occupation in general). Most participants used the first card the tool presented them with. One person reported looking at multiple cards before starting on the task, using the card they were “handed” initially. At the end of the workshop, we discussed each card participants drew, and what their ideas on applying it were. At that point already one workshop participant asked for a PDF file containing the whole card deck to be put up on the website so they could print it and use the card deck at work. This affirmed our assumption that there is a need for this kind of guidance, and that the card deck seemed to work well for (at least) this participant. Hence, their feedback indicates the potential applicability of the card deck for practitioners.

Participants were informed of the exploratory nature of this workshop and the associated intent of testing the card deck with others unfamiliar with the work. The workshop was documented (mostly by the first author) in a collaborative document[11]. We further sent out an online form for feedback via email, though the form was only filled in by one person.

7.3.1 Verbal Feedback

Participants were asked for their feedback on both the deck of cards and the facilitation thereof at the end of the workshop. The first author took notes on what they said in a local notepad for future analysis, which would give us pointers regarding strengths and potential improvements.

The group setting was well received by participants. One participant noted that the mixed group was interesting, another pointed out that the group size worked well for the virtual setting. Another stated that they liked the character of the setting (“great exchange”),[12] and that they approved of keeping to a half-day schedule. One participant remarked that they had hoped for a broader focus of the workshop overall – even though we tried to bring in other dimensions of diversity, we mostly discussed issues of gender.

Regarding the cards, several participants remarked that the content clearly leaned towards research and HCI issues – however, they also pointed out that they could see how to adapt the deck for different settings. One participant said that they found the questions on the cards to be well developed. They described the temporal grouping of the cards as “smart and fascinating”.[13] Two participants explicitly stated that they really liked the idea of using cards.

We find particularly striking that even the participant who during the introduction round had left the impression of being sceptical of the method, and even of the necessity of improving diversity and inclusion, said that they had learned about new aspects. More specifically, they would, in the future, scrutinize implicit assumptions more often.

7.3.2 Written Feedback

After the workshop, we sent participants another email with further information, and the links to our documentation. We included the link to a Nextcloud form (hosted by Chaos Computer Club Wien) asking them for feedback. While only one person filled in the form, they provided us with detailed comments that will guide the next iteration of the cards.

They had specific feedback for two cards, #11 and #1. For #11 (“Actively looking for what’s missing”[14]), they recommended to add more examples, to include other dimensions of diversity than gender. They also noted that it might be useful to discuss the concept of positionality on this card. Their feedback for #1 (“Talking explicitly about gender”[15]) has a similar direction: the card itself “is great, but I wondered why other dimensions don’t exist as individual cards”.[16] They continued to list some examples like background (German: Herkunft) and race/racialization (German: Hautfarbe), or to change the card to “Talking explicitly about identity”.[17]

In general, they said they really liked the idea, and would love to see more cards. For the online drawing tool, they mused about an option to draw more than one card.

Asked for specific use cases, the respondent noted planning events (e. g. conferences), as well as scientific projects. Prompted for whether they would advise the card deck to their colleagues, or what would have to be improved on for them to do so, they responded they already had told others about the deck.

8 Conclusion

We conducted a close reading of ten HCI papers across a range of contribution types with a primary or secondary focus on gender. Through using an interpretative and qualitative approach, we thereby engaged deeply with the material to provide a contextual analysis. We contrasted different approaches and the motivations for different modes of knowledge production. Each paper from the corpus was analysed for gender aspects in research questions, methodologies, and language used, guided by the overall question “What can we learn from this?”. The recommendations we derived from this contrasting review can be applied to various contribution types as described by [65], and span the whole life cycle of HCI research from designing research protocols to acquiring funding, and conducting research to presenting research. Drawing on these recommendations, we further developed a card deck for the initiation and facilitation of discussion and structured reflection of gender issues in HCI research and practice, which we tested with five participants in an exploratory setting.

Our view on marginalisation is specific, and can never be all-encompassing. As white members of academia, culturally tied to majority perspectives within central Europe, we can hardly talk about marginalisation due to e. g. race or forced migration. However, our view on the world and existing HCI research is informed by our personal experiences of marginalisation, which have implicitly informed our analysis of existing works, and should be taken into account by readers and fellow researchers.

All papers we reviewed were published by SIGCHI or in ACM journals on HCI. This means that while we were able to gather an impression on what is mainstream regarding gendered issues in HCI, we might have missed some developments already addressing our critique in parts or fully. We encourage fellow researchers to pursue this angle from their perspective, amplifying voices, views and requests that, to date, remain unattended to.

By conducting a close reading, we engaged only with a limited set of (mostly empirical) papers. While the papers were selected to represent a diverse set of contribution types and issues, they only represent a limited set of individual instances on these parameters. Hence, our subsequent recommendations are derived not only from the situatedness of our reading of these papers but also by the specific make-up of our corpus. Other analyses of the same or different sets of papers might come to different conclusions in what they might deem important to focus on. Hence, we suggest to our readers to not take these guidelines as a monolithic entity, but as a potential (starting) point inviting critical engagement and a continuous process aiming at developing adequate (gender) sensitivity.

To test the recommendations listed here in practice and to provide researchers with more support materials, we developed a deck of informative cards communicating these recommendations more succinctly. However, due to the limited testing conducted so far, further work remains to be done in solidifying the validity and appropriateness of the card deck. Additionally, we plan to create a secondary set that allows for the explicitly playful exploration of gender issues in HCI. The cards can be used as teaching material or as management tools in most phases of HCI projects. They may assist with (1) identifying and developing more specific and intricate research questions; (2) planning and organising a project’s sessions and meetings to be inclusive and supportive of marginalised individuals (no matter their level of participation) and (3) creating project materials that provide researchers, practitioners and their collaborators with a deep understanding of the project’s context.

We determined a gap in the existing body of work on gender in HCI (analytical reviews, software analyses and guidelines), namely a lack of guidance for HCI researchers interested in how to conduct research in a (gender) inclusive manner to develop their sensitivity. To further narrow this gap, we contributed our critical analysis, recommendations and the ensuing card deck communicating those.

Funding source: Austrian Science Fund

Award Identifier / Grant number: T 1146-G

Funding statement: Part of this work has been financially supported by the Austrian Science Fund (FWF) under grant T 1146-G.

About the authors

Sabrina Burtscher

Sabrina Burtscher is a student in TU Wien’s Media and Human Centered Computing Master’s program. As part of her Master’s thesis, she develops a method and tool kit to enable researchers and practitioners to discover gender(ed) aspects and dimensions in their work.

Katta Spiel

Katta Spiel is an FWF Hertha-Firnberg scholar at the HCI Group of TU Wien. They research marginalised perspectives on technology to inform critical design and engineering with a particular focus on trans and intersex as well as disabled experiences.

Acknowledgment

We thank all the authors who came before us in identifying issues around gender in HCI. Additionally, we are indebted to Kristina Weinberger for feedback on previous versions of this draft as well as Brigitte Ratzer for soliciting and guiding the initial work. Matthias Steinböck and Alex Steiner were instrumental in supporting the digital implementation, accessibility improvements and maintenance of the card deck.

References

[1] Syed Ishtiaque Ahmed, Steven J. Jackson, Nova Ahmed, Hasan Shahid Ferdous, Md. Rashidujjaman Rifat, A.S.M., Rizvi Shamir Ahmed, and Rifat Sabbir Mansur. 2014. Protibadi: A Platform for Fighting Sexual Harassment in Urban Bangladesh. In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (CHI ’14) (Toronto, Ontario, Canada). Association for Computing Machinery, New York, NY, USA, 2695–2704. https://doi.org/10.1145/2556288.2557376.Search in Google Scholar

[2] Dan Anderberg, Helmut Rainer, Jonathan Wadsworth, and Tanya Wilson. 2015. Unemployment and domestic violence: Theory and evidence. The Economic Journal 126, 597, 1947–1979. https://doi.org/10.1111/ecoj.12246. arXiv: http://oup.prod.sis.lan/ej/article-pdf/126/597/1947/25841725/ej1947.pdf.Search in Google Scholar

[3] Rosanna Bellini, Angelika Strohmayer, Patrick Olivier, and Clara Crivellaro. 2019. Mapping the Margins: Navigating the Ecologies of Domestic Violence Service Provision. In Proceedings of the 2019 CHI Conference on Human Factors in Computing Systems (CHI ’19) (Glasgow, Scotland, UK). Association for Computing Machinery, New York, NY, USA, Article 122, 13 pages. https://doi.org/10.1145/3290605.3300352.Search in Google Scholar

[4] Lindsay Blackwell, Jean Hardy, Tawfiq Ammari, Tiffany Veinot, Cliff Lampe, and Sarita Schoenebeck. 2016. LGBT Parents and Social Media: Advocacy, Privacy, and Disclosure during Shifting Social Movements. In Proceedings of the 2016 CHI Conference on Human Factors in Computing Systems (CHI ’16) (San Jose, California, USA). Association for Computing Machinery, New York, NY, USA, 610–622. https://doi.org/10.1145/2858036.2858342.Search in Google Scholar

[5] Janis S. Bohan. 1993. Regarding gender: Essentialism, constructionism, and feminist psychology. Psychology of Women Quarterly 17, 1, 5–21.Search in Google Scholar

[6] Cathy A. R. Brant. 2016. Teaching Our Teachers: Trans* and Gender Education in Teacher Preparation and Professional Development. In Teaching, Affirming, and Recognizing Trans and Gender Creative Youth. Springer, 47–61.Search in Google Scholar

[7] Wayne Brekhus. 1998. A sociology of the unmarked: Redirecting our focus. Sociological Theory 16, 1, 34–51.Search in Google Scholar

[8] Emeline Brulé and Katta Spiel. 2019. Negotiating Gender and Disability Identities in Participatory Design. In Proceedings of the 9th International Conference on Communities & Technologies – Transforming Communities (C & T ’19) (Vienna, Austria). Association for Computing Machinery, New York, NY, USA, 218–227. https://doi.org/10.1145/3328320.3328369.Search in Google Scholar

[9] Barbara Burman. 2002. Pocketing the difference: Gender and pockets in nineteenth-century Britain. Gender & History 14, 3, 447–469.Search in Google Scholar

[10] Margaret Burnett, Anicia Peters, Charles Hill, and Noha Elarief. 2016. Finding Gender-Inclusiveness Software Issues with GenderMag: A Field Investigation. In Proceedings of the 2016 CHI Conference on Human Factors in Computing Systems (CHI ’16) (San Jose, California, USA). Association for Computing Machinery, New York, NY, USA, 2586–2598. https://doi.org/10.1145/2858036.2858274.Search in Google Scholar

[11] Jenna M. Calton, Lauren Bennett Cattaneo, and Kris T. Gebhard. 2016. Barriers to help seeking for lesbian, gay, bisexual, transgender, and queer survivors of intimate partner violence. Trauma, Violence, & Abuse 17, 5, 585–600. https://doi.org/10.1177/1524838015585318, arXiv: https://doi.org/10.1177/1524838015585318, PMID: 25979872.Search in Google Scholar

[12] Rachel Clarke, Peter Wright, Madeline Balaam, and John McCarthy. 2013. Digital Portraits: Photo-Sharing after Domestic Violence. In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (CHI ’13) (Paris, France). Association for Computing Machinery, New York, NY, USA, 2517–2526. https://doi.org/10.1145/2470654.2481348.Search in Google Scholar

[13] Rachel Clarke, Peter Wright, and John McCarthy. 2012. Sharing Narrative and Experience: Digital Stories and Portraits at a Women’s Centre. In CHI ’12 Extended Abstracts on Human Factors in Computing Systems (CHI EA ’12) (Austin, Texas, USA). ACM, New York, NY, USA, 1505–1510. https://doi.org/10.1145/2212776.2223663.Search in Google Scholar

[14] Patricia Hill Collins. 1997. Comment on Hekman’s “Truth and method: Feminist standpoint theory revisited”: Where’s the power? Signs: Journal of Women in Culture and Society 22, 2, 375–381.Search in Google Scholar

[15] Elizabeth P. Cramer and Sara-Beth Plummer. 2010. Social work practice with abused persons with disabilities. Domestic Violence: Intersectionality and Culturally Competent Practice, 128–154.Search in Google Scholar

[16] Kimberle Crenshaw. 1990. Mapping the margins: Intersectionality, identity politics, and violence against women of color. Stan. L. Rev. 43, 1241.Search in Google Scholar

[17] Hilary Davis and Jane Farmer. 2016. Digital Participation: Engaging Diverse and Marginalised Communities. In Proceedings of the 28th Australian Conference on Computer-Human Interaction (OzCHI ’16) (Launceston, Tasmania, Australia). Association for Computing Machinery, New York, NY, USA, 672–675. https://doi.org/10.1145/3010915.3011866.Search in Google Scholar

[18] Betsy James DiSalvo, Sarita Yardi, Mark Guzdial, Tom McKlin, Charles Meadows, Kenneth Perry, and Amy Bruckman. 2011. African American Men Constructing Computing Identity. In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (CHI ’11) (Vancouver, BC, Canada). Association for Computing Machinery, New York, NY, USA, 2967–2970. https://doi.org/10.1145/1978942.1979381.Search in Google Scholar

[19] Ana Maria Bustamante Duarte, Nina Brendel, Auriol Degbelo, and Christian Kray. 2018. Participatory design and participatory research: An HCI case study with young forced migrants. ACM Trans. Comput.-Hum. Interact. 25, 1, Article 3, 39 pages. https://doi.org/10.1145/3145472.Search in Google Scholar

[20] Virginia Eubanks. 2018. Automating Inequality: How High-Tech Tools Profile, Police, and Punish the Poor. St. Martin’s Press, Inc., USA.Search in Google Scholar

[21] Anne Fausto-Sterling. 2000. Sexing the Body: Gender Politics and the Construction of Sexuality. Basic Books.Search in Google Scholar

[22] Julia R. Fernandez and Jeremy Birnholtz. 2019. “I don’t want them to not know”: Investigating decisions to disclose transgender identity on dating platforms. Proc. ACM Hum.-Comput. Interact. 3, CSCW, Article 226, 21 pages. https://doi.org/10.1145/3359328.Search in Google Scholar

[23] Carina S. González-González, Rosa M. Gil Iranzo, Patricia Paderewski, and Natalia Padilla-Zea. 2018. Gender Design Methods for Engineering, Responsible Innovation and Interaction. In Proceedings of the XIX International Conference on Human Computer Interaction (Interacción 2018) (Palma, Spain). Association for Computing Machinery, New York, NY, USA, Article 40, 2 pages. https://doi.org/10.1145/3233824.3233868.Search in Google Scholar

[24] Oliver Haimson. 2018. Social media as social transition machinery. Proc. ACM Hum.-Comput. Interact. 2, CSCW, Article 63, 21 pages. https://doi.org/10.1145/3274332.Search in Google Scholar

[25] Oliver L. Haimson, Jed R. Brubaker, and Gillian R. Hayes. 2014. DDF Seeks Same: Sexual Health-Related Language in Online Personal Ads for Men Who Have Sex with Men. In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (CHI ’14) (Toronto, Ontario, Canada). Association for Computing Machinery, New York, NY, USA, 1615–1624. https://doi.org/10.1145/2556288.2557077.Search in Google Scholar

[26] Oliver L. Haimson, Avery Dame-Griff, Elias Capello, and Zahari Richter. 2019a. Tumblr was a trans technology: the meaning, importance, history, and future of trans technologies. Feminist Media Studies, 1–17. https://doi.org/10.1080/14680777.2019.1678505, arXiv: https://doi.org/10.1080/14680777.2019.1678505.Search in Google Scholar

[27] Oliver L. Haimson, Bryan Semaan, Brianna Dym, Joey Chiao-Yin Hsiao, Daniel Herron, and Wendy Moncur. 2019b. Life Transitions and Social Technologies: Research and Design for Times of Life Change. In Conference Companion Publication of the 2019 on Computer Supported Cooperative Work and Social Computing (CSCW ’19) (Austin, TX, USA). Association for Computing Machinery, New York, NY, USA, 480–486. https://doi.org/10.1145/3311957.3359431.Search in Google Scholar

[28] Jack Halberstam. 2012. Unlearning. Profession, 1, 9–16.Search in Google Scholar

[29] Donna Haraway. 1988. Situated knowledges: The science question in feminism and the privilege of partial perspective. Feminist Studies 14, 3, 575–599.Search in Google Scholar

[30] Nancy C.M. Hartsock. 2017. The feminist standpoint: Developing the ground for a specifically feminist historical materialism. In Karl Marx. Routledge, 565–592.Search in Google Scholar

[31] Charles G. Hill, Maren Haag, Alannah Oleson, Chris Mendez, Nicola Marsden, Anita Sarma, and Margaret Burnett. 2017. Gender-Inclusiveness Personas vs. Stereotyping: Can We Have It Both Ways? In Proceedings of the 2017 CHI Conference on Human Factors in Computing Systems (CHI ’17) (Denver, Colorado, USA). Association for Computing Machinery, New York, NY, USA, 6658–6671. https://doi.org/10.1145/3025453.3025609.Search in Google Scholar

[32] Sarah Hunt. 2018. Embodying self-determination: beyond the gender binary. In Determinants of Indigenous Peoples’ Health, 22–39.Search in Google Scholar

[33] Samantha Jaroszewski, Danielle Lottridge, Oliver L. Haimson, and Katie Quehl. 2018. “Genderfluid” or “Attack Helicopter”: Responsible HCI Research Practice with Non-Binary Gender Variation in Online Communities. In Proceedings of the 2018 CHI Conference on Human Factors in Computing Systems (CHI ’18) (Montreal QC, Canada). Association for Computing Machinery, New York, NY, USA, Article 307, 15 pages. https://doi.org/10.1145/3173574.3173881.Search in Google Scholar

[34] Idin Karuei, Karon E. MacLean, Zoltan Foley-Fisher, Russell MacKenzie, Sebastian Koch, and Mohamed El-Zohairy. 2011. Detecting Vibrations Across the Body in Mobile Contexts. In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (CHI ’11) (Vancouver, BC, Canada). Association for Computing Machinery, New York, NY, USA, 3267–3276. https://doi.org/10.1145/1978942.1979426.Search in Google Scholar

[35] Os Keyes. 2018. The misgendering machines: Trans/HCI implications of automatic gender recognition. Proc. ACM Hum.-Comput. Interact., 2, CSCW, Article 88, 22 pages. https://doi.org/10.1145/3274357.Search in Google Scholar

[36] Os Keyes, Burren Peil, Rua Williams, and Katta Spiel. 2020. Reimagining (women’s) health: HCI, gender and essentialised embodiment. ACM Trans. Comput.-Hum. Interact., 27, 4, 25. https://doi.org/10.1145/3404218.Search in Google Scholar

[37] Neha Kumar, Naveena Karusala, Azra Ismail, Marisol Wong-Villacres, and Aditya Vishwanath. 2019. Engaging feminist solidarity for comparative research, design, and practice. Proc. ACM Hum.-Comput. Interact., 3, CSCW, Article 167, 24 pages. https://doi.org/10.1145/3359269.Search in Google Scholar

[38] Susan Leavy. 2018. Gender Bias in Artificial Intelligence: The Need for Diversity and Gender Theory in Machine Learning. In Proceedings of the 1st International Workshop on Gender Equality in Software Engineering (GE ’18) (Gothenburg, Sweden). Association for Computing Machinery, New York, NY, USA, 14–16. https://doi.org/10.1145/3195570.3195580.Search in Google Scholar

[39] Nicola Marsden, Julia Hermann, and Monika Pröbster. 2017. Developing Personas, Considering Gender: A Case Study. In Proceedings of the 29th Australian Conference on Computer-Human Interaction (OZCHI ’17) (Brisbane, Queensland, Australia). Association for Computing Machinery, New York, NY, USA, 392–396. https://doi.org/10.1145/3152771.3156143.Search in Google Scholar

[40] JR Martin. 2005. Close reading: functional linguistics as a tool for critical discourse analysis. In Researching Language in Schools and Communities: Functional Linguistic Perspectives, 275.Search in Google Scholar

[41] C. Mendez, L. Letaw, M. Burnett, S. Stumpf, A. Sarma, and C. Hilderbrand. 2019. From GenderMag to InclusiveMag: An Inclusive Design Meta-Method. In 2019 IEEE Symposium on Visual Languages and Human-Centric Computing (VL/HCC). 97–106. https://doi.org/10.1109/VLHCC.2019.8818889.Search in Google Scholar

[42] Danaë Metaxa-Kakavouli, Kelly Wang, James A. Landay, and Jeff Hancock. 2018. Gender-Inclusive Design: Sense of Belonging and Bias in Web Interfaces. In Proceedings of the 2018 CHI Conference on Human Factors in Computing Systems (CHI ’18) (Montreal QC, Canada). Association for Computing Machinery, New York, NY, USA, Article 614, 6 pages. https://doi.org/10.1145/3173574.3174188.Search in Google Scholar

[43] Linda Nicholson. 1994. Interpreting gender. Signs: Journal of Women in Culture and Society 20, 1, 79–105.Search in Google Scholar

[44] Ihudiya Finda Ogbonnaya-Ogburu, Angela D. R. Smith, Alexandra To, and Kentaro Toyama. 2020. Critical Race Theory for HCI. In Proceedings of the 2020 CHI Conference on Human Factors in Computing Systems (CHI ’20) (Honolulu, HI, USA). Association for Computing Machinery, New York, NY, USA, 1–16. https://doi.org/10.1145/3313831.3376392.Search in Google Scholar

[45] Jahna Otterbacher. 2015. Crowdsourcing Stereotypes: Linguistic Bias in Metadata Generated via GWAP. In Proceedings of the 33rd Annual ACM Conference on Human Factors in Computing Systems (CHI ’15) (Seoul, Republic of Korea). Association for Computing Machinery, New York, NY, USA, 1955–1964. https://doi.org/10.1145/2702123.2702151.Search in Google Scholar

[46] Janice Ristock and Norma Timbang. 2005. Relationship violence in lesbian/gay/bisexual/transgender/queer [LGBTQ] communities. Violence Against Women Online Resources.Search in Google Scholar

[47] Jennifer A. Rode. 2011. Reflexivity in Digital Anthropology. In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (CHI ’11) (Vancouver, BC, Canada). Association for Computing Machinery, New York, NY, USA, 123–132. https://doi.org/10.1145/1978942.1978961.Search in Google Scholar

[48] Maria Francesca Roig-Maimó and Ramon Mas-Sansó. 2019. The Female Effect: The Case of Gender Parity on User Studies. In Proceedings of the XX International Conference on Human Computer Interaction (Interacción ’19) (Donostia, Gipuzkoa, Spain). Association for Computing Machinery, New York, NY, USA, Article 46, 6 pages. https://doi.org/10.1145/3335595.3335606.Search in Google Scholar

[49] Koustuv Saha, Sang Chan Kim, Manikanta D. Reddy, Albert J. Carter, Eva Sharma, Oliver L. Haimson, and Munmun De Choudhury. 2019. The language of LGBTQ+ minority stress experiences on social media. Proc. ACM Hum.-Comput. Interact., 3, CSCW, Article 89, 22 pages. https://doi.org/10.1145/3361108.Search in Google Scholar

[50] Morgan Klaus Scheuerman, Jacob M. Paul, and Jed R. Brubaker. 2019a. How computers see gender: An evaluation of gender classification in commercial facial analysis services. Proc. ACM Hum.-Comput. Interact., 3, CSCW, Article 144, 33 pages. https://doi.org/10.1145/3359246.Search in Google Scholar

[51] Morgan Klaus Scheuerman, Katta Spiel, Oliver Haimson, Foad Hamidi, and Stacy M. Branham. 2019b. HCI Guidelines for Gender Equity and Inclusivity. https://www.morgan-klaus.com/sigchi-gender-guidelines.Search in Google Scholar

[52] Ari Schlesinger, W. Keith Edwards, and Rebecca E. Grinter. 2017. Intersectional HCI: Engaging Identity through Gender, Race, and Class. In Proceedings of the 2017 CHI Conference on Human Factors in Computing Systems (CHI ’17) (Denver, Colorado, USA). Association for Computing Machinery, New York, NY, USA, 5412–5427. https://doi.org/10.1145/3025453.3025766.Search in Google Scholar

[53] Katta Spiel, Emeline Brulé, Christopher Frauenberger, Gilles Bailley, and Geraldine Fitzpatrick. 2020. In the details: the micro-ethics of negotiations and in-situ judgements in participatory design with marginalised children. CoDesign 16, 1, 45–65. https://doi.org/10.1080/15710882.2020.1722174, arXiv: https://doi.org/10.1080/15710882.2020.1722174.Search in Google Scholar

[54] Katta Spiel and Kathrin Gerling. 2019. The Surrogate Body in Play. In Proceedings of the Annual Symposium on Computer-Human Interaction in Play (CHI PLAY ’19) (Barcelona, Spain). Association for Computing Machinery, New York, NY, USA, 397–411. https://doi.org/10.1145/3311350.3347189.Search in Google Scholar

[55] Katta Spiel, Oliver L. Haimson, and Danielle Lottridge. 2019a. How to do better with gender on surveys: A guide for HCI researchers. Interactions 26, 4, 62–65. https://doi.org/10.1145/3338283.Search in Google Scholar

[56] Katta Spiel, Os Keyes, and Pınar Barlas. 2019b. Patching Gender: Non-Binary Utopias in HCI. In Extended Abstracts of the 2019 CHI Conference on Human Factors in Computing Systems (CHI EA ’19) (Glasgow, Scotland, UK). Association for Computing Machinery, New York, NY, USA, Article alt05, 11 pages. https://doi.org/10.1145/3290607.3310425.Search in Google Scholar

[57] Denny L. Starks, Tawanna Dillahunt, and Oliver L. Haimson. 2019. Designing Technology to Support Safety for Transgender Women & Non-Binary People of Color. In Companion Publication of the 2019 on Designing Interactive Systems Conference 2019 Companion (DIS ’19 Companion) (San Diego, CA, USA). Association for Computing Machinery, New York, NY, USA, 289–294. https://doi.org/10.1145/3301019.3323898.Search in Google Scholar

[58] Isabelle Stengers. 2018. Another Science Is Possible: A Manifesto for Slow Science. John Wiley & Sons.Search in Google Scholar

[59] Angelika Strohmayer, Cayley MacArthur, Velvet Spors, Michael Muller, Morgan Vigil-Hayes, and Ebtisam Alabdulqader. 2019. CHInclusion: Working Toward a More Inclusive HCI Community. In Extended Abstracts of the 2019 CHI Conference on Human Factors in Computing Systems (CHI EA ’19) (Glasgow, Scotland, UK). Association for Computing Machinery, New York, NY, USA, Article W27, 10 pages. https://doi.org/10.1145/3290607.3299012.Search in Google Scholar

[60] Simone Stumpf, Anicia Peters, Shaowen Bardzell, Margaret Burnett, Daniela Busse, Jessica Cauchard, and Elizabeth Churchill. 2020. Gender-inclusive HCI research and design: A conceptual review. Foundations and Trends® in Human-Computer Interaction, 13, 1, 1–69. https://doi.org/10.1561/1100000056.Search in Google Scholar

[61] Reem Talhouk, Ana Bustamante, Konstantin Aal, Anne Weibert, Koula Charitonos, and Vasilis Vlachokyriakos. 2018. HCI and refugees: Experiences and reflections. Interactions 25, 4, 46–51. https://doi.org/10.1145/3215846.Search in Google Scholar

[62] Luis Von Ahn and Laura Dabbish. 2005. ESP: Labeling Images with a Computer Game. In AAAI Spring Symposium: Knowledge Collection from Volunteer Contributors, Vol. 2.Search in Google Scholar

[63] Mihaela Vorvoreanu, Lingyi Zhang, Yun-Han Huang, Claudia Hilderbrand, Zoe Steine-Hanson, and Margaret Burnett. 2019. From Gender Biases to Gender-Inclusive Design: An Empirical Investigation. In Proceedings of the 2019 CHI Conference on Human Factors in Computing Systems (CHI ’19) (Glasgow, Scotland, UK). Association for Computing Machinery, New York, NY, USA, Article 53, 14 pages. https://doi.org/10.1145/3290605.3300283.Search in Google Scholar

[64] Melanie Wiener, Daniela Maresch, and Robert J. Breitenecker. 2020. The shift towards entrepreneurial universities and the relevance of third-party funding of business and economics units in Austria: A research note. Review of Managerial Science, 14, 345–363.Search in Google Scholar

[65] Jacob O. Wobbrock and Julie A. Kientz. 2016. Research contributions in human-computer interaction. Interactions, 23, 3, 38–44. https://doi.org/10.1145/2907069.Search in Google Scholar

Published Online: 2021-04-22
Published in Print: 2021-04-27

© 2021 Walter de Gruyter GmbH, Berlin/Boston

Downloaded on 22.2.2024 from https://www.degruyter.com/document/doi/10.1515/icom-2021-0001/html
Scroll to top button