Tricky Design Probes: Triggering Reflection on Design Research Methods Anaëlle Beignon, Emeline Brulé, Jean-Baptiste Joatton, Aurélien Tabard To cite this version: Anaëlle Beignon, Emeline Brulé, Jean-Baptiste Joatton, Aurélien Tabard. Tricky Design Probes: Trig- gering Reflection on Design Research Methods. Designing Interactive Systems, Jul 2020, Eindhoven, Netherlands. 10.1145/3357236.3395572. hal-02869220 HAL Id: hal-02869220 https://hal.archives-ouvertes.fr/hal-02869220 Submitted on 25 Jun 2020 HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. Tricky Design Probes Triggering Reflection on Design Research Methods Anaëlle Beignon1,2, Emeline Brulé3 *, Jean-Baptiste Joatton1 *, Aurélien Tabard2 * * Listed by alphabetical order - the authors contributed equally 1Pôle supérieur de design 2Univ Lyon, Université Lyon 1, 3University of Sussex Villefontaine, France CNRS, LIRIS, UMR5205 Falmer, United Kingdom [email protected] F-69621, France [email protected] [email protected] [email protected] ABSTRACT INTRODUCTION However, this does not bring the critique to the Design research methods are increasingly used as In the last decades, design toolkits for research everyday design process itself - where values are being ready-made recipes for success in a variety of fields or ideation have found their way into businesses, generated and negotiated-therefore missing out on and multidisciplinary teams. Yet as any tools, they educational settings, urban planning [25], and social supporting designers in opening spaces of contestation. shape the gaze, attitudes, and behaviors of designers. policy organisations [1, 22]. In order to be useful across To do this, we propose to use probes [13]. Probes are a Moreover, their generic nature tends to obscure the domains, these toolkits have to be generic, detached generative and narrative method, often used to inspire specific situations in which they were created. In from their context of creation and application, or what the designer, or reveal potential new practices when reaction, grounding our work in adversarial design, Suchman would call “design from nowhere” [28]. deploying voluntarily unfinished artifacts [7]. They are usually used to learn and reflect about participants and we propose four tricky probes: believable design These toolkits often aim at rapidly synthesizing and the design context, but not for designers to learn about tools, which appear to be innocuous, but progressively building consensus on a given issue, relying on carefully themselves. engage designers in crossing boundaries of what calibrated empathy and participation [21]. In doing so should be acceptable. This is done by slowly derailing they tend to avoid critical considerations that could In this pictorial, we propose probes that build on design research activities, leading to trigger reflection undermine the projects’ expected outcomes [17, 25]. critical design to question designers’ own practices. on the part of designers on their beliefs, practice, For instance, projects might ignore the point of view We call them tricky design probes: seemingly genuine and the tools they use. Our probes raise issues at the of marginalised communities [16]. The participatory and innocuous design tools through which designers intersection of design research and gender in urban design scholarship has long highlighted the challenges explore, or encourage others to explore, boundaries and service design, such as the use of pre-made algorithms of supporting social change [6] while creating space in controversial aspects of their methods for the problem to understand gendered patterns in urban movements. which different experiences, as well as the tensions they aim to tackle. We design and deploy four probes to Authors Keywords and disagreements that can not be resolved, can be address designers’ socio-political constructs regarding gender, and question design tools used in urban design tools; design research; critical design; discussed [4, 5]. planning and urban informatics projects: (1) Gender adversarial design; gender; probes. To ignite and sustain these discussions, some classification and large-scale sensor systems (2) Politics researchers and designers have embraced adversarial Permission to make digital or hard copies of all or part of this work for personal of risk and the convenience of design workshops (3) design approaches. Instead of using design methods in or classroom use is granted without fee provided that copies are not made or Street harassment and empathy (4) Street harassment distributed for profit or commercial advantage and that copies bear this notice political processes aiming at identifying a consensus, and the full citation on the first page. Copyrights for components of this work and design guides. We reflect on how each probe owned by others than the author(s) must be honored. Abstracting with credit is adversarial design seeks to “open a space for enabled to question the project framing and methods, permitted. To copy otherwise, or republish, to post on servers or to redistribute contestation; and it suggests new practices of design.” to lists, requires prior specific permission and/or a fee. Request permissions in both expected and unexpected ways. Tricky design from [email protected]. [10 p.9] This includes critical and speculative design probes are helpful to practitioners wishing to highlight DIS ‘20, July 6–10, 2020, Eindhoven, Netherlands approaches, which most often involve creating thought- © 2020 Copyright is held by the owner/author(s). Publication rights licensed to ethical issues to colleagues and participants, and can ACM. provoking design pieces, then exhibited to foster ACM ISBN 978-1-4503-6974-9/20/07...$15.00 open new areas of inquiry for research on probes and reflection on beliefs, values and problem framing [22]. https://doi.org/10.1145/3357236.3395572 critical design. GENDER CLASSIFICATION AND LARGE- SCALE SENSOR SYSTEMS 2 Design methods in question The probe Designers increasingly rely on tracking data to This probe engages designers in confronting how they generate insights into people’s behaviors. This ranges stereotype users, and to discuss whether surveillance from analytics tools, to emerging bio-signal and technology can be ethically leveraged to gain insights. biometric monitoring based on video analysis, e.g., We collected snapshots from four public surveillance emotion inference or face-recognition. This large scale cameras recording squares or streets from around the data collection is presented as a highly effective way for world. We set-up the probe as an installation. And we designers to get reliable insights on user behavior. asked designers to classify people on snapshots in two categories: male or female. The problem 1 To set the stage, a strip hangs from the ceiling As data gets collected on a large scale, it must be pro- enabling to scroll through 24 hours of snapshots cessed and synthesized to draw any form of insights. captured from a street camera. On the wall, further This algorithmic treatment involves aggregation, filter- snapshots reinforce the feeling for monitoring and ing, clustering, and summarization techniques in order surveillance. to recognize stereotypical traits or personas. With this 2 On the table, a notebook lets designers enact the probe, we focus on the issue of identifying and catego- gender classification algorithm by grading gender- rizing activities of women. When it comes to gender, stereotypical criteria used to assign a gender to people activists have long recognized both the needs to identify in the video. statistical gender differences [26] but also to go beyond ©Anaëlle Beignon the gender binary [15]. By offering a probe that enables designers to experience this tension we seek to make this problem tangible. Moreover, we also seek to raise the issue of data being collected without full informed consent, and how it can be leveraged by external actors to secure their say on the design of services or even 1 future cities [8]. ©Anaëlle Beignon ©Anaëlle Beignon Training the recognition algorithm 2 FIRST AUHOR’S INITIAL FIRST AUHOR’S VERSION VERSION OF THE In order to process data collected on a large scale, VERSION OF THE ALGORITHM: AFTER THE TRAINING: PARTICIPANTS automated analysis becomes mandatory. These (2 DESIGNERS): automated methods typically rely on machine learning SHORT HAIR: +40 SHORT HAIR: +40 techniques that are trained on a subset of the data and LONG HAIR: -30 LONG HAIR: -25 SHORT HAIR: +50 will propagate the gaps, limitations and biases of the HEIGHT > 1M75: +13 HEIGHT > 1M75: +15 LONG HAIR: -25 initial categorization and training. HEIGHT < 1M75: 0 WEARING A SKIRT: -40 FACIAL HAIR: +40 We structured the experience in two stages, with the WEARING A SKIRT: -80 WEARING PANTS: +10 DYE: -30 first author and two design students manually training WEARING PANTS: +10 BY BIKE: +20 SHIRT: +20 BY BIKE: +20 HANDBAG: -25 HEADGEAR: +20 a gender recognition algorithm. CRITERIA DELETED IN THE HANDBAG: