Was This Helpful? Yes or No Design Explorations Of Ethics And Agency In Social Media By Aman Singh Bachelor of Technology (Mechanical Engineering) Dr. A.P.J. Abdul Kalam Technical University, 2016 A CRITICAL AND PROCESS DOCUMENTATION THESIS PAPER SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF DESIGN EMILY CARR UNIVERSITY OF ART + DESIGN 2022 © Aman Singh, 2022 Acknowledgement I would like to express my deep and sincere gratitude to my research supervisor, Dr. Bonne Zabolotney, Associate Professor, Ian Gillespie Faculty of Design + Dynamic Media, Emily Carr University of Art + Design for her invaluable guidance throughout this research. Her vision, sincerity and motivation helped me to carry out the research and present it as clearly as possible. It was a great privilege and honor to work and study under her guidance. I would also like to thank all the faculty members, for their honest feedback throughout the course of this research that helped me push my limits as a designer. A big thank you to everyone in my cohort for their brutally honest critiques and always being there to help me turn ideas into real projects. Also, for the all the FRIYAYs to unwind after a long week. This research would have never been possible without your support. I am extremely grateful to my family for their love, support, and motivation to keep pushing through and do better each day. i Abstract Table of Contents This thesis project focuses on exploring the ways in which designers can facilitate agency and Acknowledgement ......................................................................................................... i ethical understanding in social media users. My research began with the investigation of how Abstract .......................................................................................................................... ii consent forms for online platforms are designed, the rules that govern them, and the amount of Table of Contents ........................................................................................................... iii user information collected through these forms. At the same time, I investigated the definition of Keywords ....................................................................................................................... v true informed consent on these platforms. I conducted various data visualization experiments, List of Figures ................................................................................................................ vi using both simple and complex data, to learn the ways in which information design can make data understandable. Using quantitative methods and design action research, I expanded my I INTRODUCTION ......................................................................................................... 1 investigations to interrogate concepts of data exhaust, inferences, and behavioral modification, Context and Framing ................................................................................................ 1 with the aim to empower users to understand and control the use of their personal data. Formulating the Problem Space ............................................................................... 3 Throughout these investigations and research projects, I understood that information design can Initial Investigations Into Understanding Consent on Digital Platforms .................. 4 be a useful tool to uncover ethical, political, and corporate structures which affect the privacy and Importance of Data Literacy for Users ..................................................................... 6 agency of social media users. Uncovering these structures also assist future designers in understanding their own contributions to these spaces. II INFORMATION DESIGN ............................................................................................ 7 Information Design Theory and Practices ................................................................ 7 First Experiments in Data Visualizations ................................................................. 9 FRIES of Digital Consent ......................................................................................... 17 Formulating the Idea of Visualizing Consent Forms of Social Media Platforms ..... 20 III INFERENCES ............................................................................................................... 34 Data Exhaust, Behavioral Data, Inferences and Behavioral Modifications .............. 34 Why Does Data Visualization Not Help? ................................................................. 36 Formulating the Data and Behavioral Surplus Project ............................................. 37 REB Application ....................................................................................................... 39 ‘Infer’ - Keyword Search on Online Consent Forms ................................................ 45 IV DISSEMINATIONS ...................................................................................................... 49 How do I get People to Engage with All This? ........................................................ 49 Internet Cookies – A Presentation ............................................................................ 50 The Inferences Presentation ...................................................................................... 60 Research Narratives .................................................................................................. 63 ii iii Keywords V APPLICATIONS OF INFORMATION DESIGN ....................................................... 65 MITACS - ioAirFlow .............................................................................................. 65 Temperature v/s Time Graph ................................................................................... 66 VI DISCUSSION AND CLOSING REMARKS .............................................................. 71 Conclusion .............................................................................................................. 71 VII REFERENCES ............................................................................................................ 74 VIII APPENDIX A .............................................................................................................. 80 REB Application ..................................................................................................... 80 B Behavioral Data, 34, 39, 83 Behavioral Modification, 6, 34, 35 E Ethics, 1, 6, 29, 30, 34, 36, 39, 71, 73, 77, 79, 80, 81, 87, 89 D Data Exhaust, 34, 37, 38, 72, 82, 83 Data Literacy, 6 I Inferences, 34, 39, 45, 46, 60, 62, 71, 72, 83 Information Design, 1, 6, 7, 8, 20, 65, 71, 74, 77, 79, 83 iv v List of Figures 1: Tweet by Jake Tapper, March 21, 2018....................................................................... 2: Line chart showing distance traveled on foot and bus for the months of February & March 2021 (Singh, March 2021)............................................................................... 3: I - INTRODUCTION 3 11 Radial chart showing distance traveled for the months of February & March 2021 (Singh, March 2021).................................................................................................... 12 4: Demographics of MDES 2022 (Singh, March 2021).................................................. 14 5: Annotating the first 100 decimals of Pi, following Anwar Aqeel’s work (Singh, March 2021)................................................................................................... 6: 16 Understanding the FRIES of consentful technology, designed using definitions (Singh, March 2021).................................................................................................... 18 7: Visualizing the consent form of Pinterest.com (Singh, May 2021)............................ 21 8: Network visualization of members of my cohort, I met in person for the first time 9: 26 Visualizing the consent form of WhatsApp after being updated in 2021 (Singh, October 2021)................................................................................................. 32 10: Visualizing the self-referencing aspect of online consent forms (Singh, July 2021).. 33 11: Examples of Data Exhaust on LinkedIn.com.............................................................. 38 12–20: Documenting data from LinkedIn.com (Singh, July 2021)......................................... 40-42 21: The inference cycle (Singh, August 2021).................................................................. 44 22: ‘infer’ - keyword search on linkedin.com/legal/user-agreement................................. 47 23: ‘infer’ - keyword search on policy.pinterest.com/en/privacy-policy........................... 48 24–31: Slides from the Cookies presentation (Singh, September 2021)................................. 51-58 32: Screens from the Inferences prototype (Singh, October 2021)................................... 62 33: Screenshots from the Research Narrative.................................................................... 64 34: Older version of the Temperature v/s Time graph (shared with permission from ioAirFlow)................................................................................................................... 35: 36: 67 Redesigned version of the Temperature v/s Time graph with sensors categorized according to location (shared with permission from ioAirFlow)................................ 68 Redesigned version of the Temperature v/s Time graph (isolated view) (shared with permission from ioAirFlow)........................................................................................ vi This thesis project focuses on exploring ways in which information 1. Practice of presenting information in design1 can facilitate agency for data ownership and develop easily understood by users 69 a way that makes it most accessible and transparency and ethics in informed consent. My studies started (Tomboc 2019) with a critical response on the reading Ethics in Design by Clive 2. Moral principles that govern a person’s Dilnot. Throughout the article Dilnot tries to understand the role behavior or conducting of an activity 2 of ethics in design. He does this by referencing the work of other designers like Victor Papanek and Kenneth Frampton. He establishes similarities in the views of these designers and from Building Consentful Tech (Lee & Toliver 2017) (Singh, September 2021)............................................................................................. Context and Framing concludes with considering design’s role as serving the wider interests of subjects rather than the narrow interests of private profit. Dilnot writes about the role of design capabilities in creating a human world. “Capitalist culture organizes people as buyers of commodities and service [and]... transform[s] information and knowledge into commodities. The corporate conglomerates of the culture industry have created a global public sphere which does not offer any scope of discussion of the social and cultural consequences of the ‘free flow of information’ organized by them” (Troon). “Incorporating ethics in design will counter our culture and social inability to designate the dimensions of human good beyond that of the market, bring absolute primacy of the interest of human beings in a human future and open up ways of making and remaking a sustainable world” (Dilnot 2009). Meanwhile in January 2021, mobile messaging platform WhatsApp announced that it will be updating its Privacy Policy starting February 8, 2021 after it was bought by Facebook. This update would allow WhatsApp to share more user data including their private messages with Facebook. It also stated that users would 1 have to agree to these terms if they wished to continue using the Formulating the Problem Space app. ‘This led to a widespread backlash from confused users, resulting in various legal challenges and regulatory investigations’ (TechCrunch 2021). Users immediately started fleeing to other messaging apps like Signal and Telegram. The effect was such that ‘WhatsApp in the UK fell from being the eighth most downloaded app to the 23rd in just a few weeks. By contrast, Signal, which was not even in the top 1,000 downloaded apps, jumped to being one of the most downloaded apps’ (AppAnnie 2021). In order to clear the air around this confusion WhatsApp rolled out a few in-app banners explaining the new terms visually and announced a three month delay to the new terms and conditions which then went into effect in May 2021. Figure 1: Tweet by Jake Tapper, March 21, 2018 Social media platforms that have become a part of our lives and help us ‘stay connected’ to the world never charge us for the services they provide because they sell our data as the product. Sharing user data in the name of customized experiences and targeted ads can sometimes prove to be disastrous. In the past there have been cases of online fraud and financial losses because someone’s personal data was leaked. In a time where everything is linked to your email account, the need to be vigilant while giving out personal data has increased manifold. Everything on an online platform is aesthetically pleasing but the consent page which is black and white and undesigned. The most important page, where users agree to share their data, but most users are not aware of what the terms of sharing this data are. So, this left me with a question of whether the consent is truly informed or not. I realize that neither social media platforms nor the internet can function without the user data, but the entire process of collecting this data and then processing it to make our lives better could be more transparent. This research aims to explore, how the experience of giving consent to share user data be made transparent, inviting and ethical? 2 3 Initial Investigations Into Understanding Consent on Digital ‘Pre-digital sharing was about exchange, sharing with digital Platforms technologies is about exchange and distribution’ (Stalder & Sützl 2011). While users share information via online platforms, there are User agreements are typically ignored by people using online 3. The potential harms of misuse of services, therefore users are not aware of the potential harms3 and is not part of this research work associated with the sharing of their personal information. ‘The biggest threat to personal security in the online realm is the personal data is an entire subject by itself 4. The 7 key rules of the GDPR (One conflicts between the commercial ambitions of these platforms and the intellectual property rights of the shared information. On the internet, users are bound by private consent forms to use any Trust, 2021) at the heart of the law should service and cannot negotiate the terms of use on these contracts. unregulated accumulation of data by numerous social media inform every step of a modern privacy Unethical practices like sharing user data with government companies’ (Zuboff 2019). ‘The emergence of social networks like are as follows: Facebook or disclosure platforms like WikiLeaks have shown that there is a need to go beyond the scientific habits and legal management program. These principles Lawfulness, Fairness and Transparency agencies and third-party advertisers for financial gains are some of the violations of data privacy that users have no control over. There should always be a good reason for standards of sharing knowledge and distributing information to processing personal data understand and govern communicative space and exchange of Purpose Limitation information’ (Stalder & Sützl 2011). clearly established and must be clearly The purpose of processing data must be communicated to individuals through a privacy note Even after all of this, some might argue that there are certain rules and regulations in place to keep check on how companies use and Data Minimization Only collect the smallest amount of data store user data. A good example of this could be the General Data you will need to complete your purposes Protection Regulation (GDPR)4, which is ‘a regulation in EU law Accuracy on data protection and privacy in the European Union and the the data collected and stored and should Companies should ensure the accuracy of European Economic Area General Data Protection Regulation’ always set up checks and balances to (gdpr-info.eu, 2018). incomplete data that comes in. correct, update, or erase incorrect or Storage Limitation You should always justify the length of time you are keeping each piece of data you store. Integrity and Confidentiality You should secure the collected data from internal and external threats and must protect data from unauthorized or unlawful processing and accidental loss, destruction or damage. Accountability You must at all times have appropriate measures and records in place as proof of your compliance with the data processing principles. 4 5 Importance of Data Literacy for Users II - INFORMATION DESIGN Information Design Theory and Practices As people generate and interact with data every minute of their 5. Is a treatment approach that replaces lives, it becomes important for them to know how it is used, who it is shared with and how it affects their lives and of those around Dear Data by Giorgia Lupi and Stefani Posavec (September, 2016) 6. Study of visual representations of desirable ones using the principles of documents a year-long information design project. Each week for a cognition operant conditioning year they collected and shared data about mundane activities with undesirable behaviors with more them. Misuse of this data “...reinforces biases in the society and each other. Sharing this data gave ideas about their days through moreover leads to behavioral modification5 of the masses that is not the pretext of their data collection. It was an attempt to compose a self directed. The availability of vast amounts of data and the tools portrait of the other person through those fragments of their nature to analyze them have come faster than the ethical and legal in the form of data. The authors found that “…data is not only standards could develop regarding the use of such data” scientific and functional but can also be presented in its lateral (Weinhardt 2020). “The protocols of communication are based on truths that can reveal the emotions, wants, needs, personality and the sharing of culture but not on the culture of sharing” so on of the person it belongs to” (Lupi 2020). They conclude that (Castells 2009). Even though there are rules in place to protect a data is not just about percentages and numbers, but are user’s data, not many people bother to think twice before placeholders for meaningful moments and visualizing it was a way consenting. They choose convenience over privacy which is not to reconnect those numbers to what they really stood for. This surprising. conclusion is supported by Rendgen and Widememann, who stated that “Our everyday lives are filled with a massive flow of After reading through these resources I wanted to explore how information that we must interpret to understand the world we live information design could be used as a tool to make the experience in. Considering this complex variety of data floating around us, of giving consent inviting and engaging. Hoping that this would sometimes the best - or even only way to communicate is visually” give users a chance to make informed decisions before sharing (2020). their data online. This led me to my first research question: While Lupi, Posavec, Rendgen and Widememann connect meaning How can information design assist in understanding and to everyday life through information design, other scholars focus developing transparency and ethics in Informed Consent forms? on the cognitive aspects of information design. “Information visualization6 is the use of computer supported interactive, visual representation of abstract data to amplify cognition” (Card et al, 1999). Graphics represent the different elements in our environment and the relations between these elements are represented in the graphic space. Information graphics help in understanding the target information by using graphical patterns. These patterns invoke existing schematic structures and most 6 7 abstract data to reinforce human information graphics are understood in image schematic terms. 7. Non-erasable core of the graphic First Experiments in Data Visualization This is done by using knowledge of things that are well understood and superimposing them on something that is Based on these readings, I started experimenting with my first complicated. For instance, using the solar system to understand the visualizations. I did this by visualizing simple and personal data structure of an atom. “Source and target domain representations are that I had access to, from my location history on Google Maps. aligned and features of the source are projected onto the target via After moving to Vancouver and coming out of quarantine in early inference” (Risch 2008). February, I traveled a lot around the city. While traveling, I found myself totally depending on Google Maps to find my way around To get a deeper understanding of data visualization and how it the city. I was learning different bus routes and constantly looking helps the human mind comprehend complex information, I relied out for landmarks to familiarize with the street names. After a few on The Visual Display of Quantitative Information by Edward weeks, I started recognizing those streets but I thought to myself, Tufte. In his chapter “Data Ink7 and Graphical Redesign,” Tufte ‘what if a tool like Google Maps could help me familiarize with a states that ‘‘...data graphics should draw the viewer’s attention to new place rather than me totally depending on it?’ Since I had been the sense and substance of data. Every bit of ink on a graphic going to school a lot, I wanted to explore if Google Maps could requires a reason and nearly always that reason should be that the help me learn this new city in a better way? While traveling on ink presents new information’’ (2001). Tufte refers to the cognitive buses in Vancouver, riders can constantly listen to the qualities of information design where jumping back and forth from announcement of the next stop. I wanted to see if the audio cues text to graphic severs the understanding and the flow of on the bus and the visual cues of the stop could help a newcomer information. He says that ‘‘words and pictures belong together, like me learn the routes. To do this, I took screenshots of the entire genuinely together. Separating text and graphics even on the same route from my home to school and recorded the sounds that I heard page usually requires encoding to link the separate elements. while taking that route. I was able to record the sounds of Attentive readers must repeatedly jump back and forth between text everything starting from the bus PA system to the beep on the and graphic’’ (Tufte 1990). This supports Lupi’s intuitive approach crossing that goes off every time the light turns green and to the to information design where text and graphics are an immersive voices of other people talking on the bus. experience. The idea was to document all of this and make a slideshow that could replicate and communicate my experience of taking the bus to the school. Building on what I had learnt so far reading Dear Data by Giorgia Lupi and Stefanie Posavec, I wanted to explore how much of this data could tell the audience about how I spent a usual day in the city. But as I began documenting this data, my focus shifted towards visualizing it and that led me to my first project in data visualization. 8 9 I started documenting the distance I had covered on foot and the distance I had covered on a bus for a period of 4 weeks. Plotting Figure 2: Line chart showing distance traveled on foot and bus for the months of February & March 2021 (Singh, March 2021) this data on a line chart (Figure 2) and on a radial chart (Figure 3), I could immediately see patterns emerging from them. For instance, the distance covered on a bus on weekdays was much higher than that covered on weekends. This made sense since I was not leaving home on weekends but was going to school during the weekdays. The other pattern that emerged from this experiment was that the distance covered was more during the later weeks of February and early weeks of March than before. This was true too because I had not got my Compass Card which is used to take the bus in Vancouver. All these initial experiments with Data Visualization indicated the fact that visualizing large amounts of data does help in understanding it, as opposed to when it is read in its traditional form. 10 11 To continue working with data visualization, I started using simple data that I had access to. One of which was visualizing the demographics of my cohort at school (Figure 4). Visualizing the split between the males and females in my class, made it clear what the demographic of my cohort was, which otherwise were just numbers before. Even though these numbers made sense, the split in the demographics was never evident. Figure 3: Radial chart showing distance traveled for the months of February and March 2021 (Singh, March 2021) 12 13 After completing these visualizations, I was sure that visualizing data could help in grasping and understanding large amounts of data easily. But my other question of whether data visualization could lead to user engagement was still left unanswered. This led me to my next experiment in data annotation . 8 While continuing to research on Data Annotation I came across an article on the internet titled, ‘‘Visualizing the Beauty of Pi – Towards Data Science’’ (Aqeel 2021). The author had conducted an experiment by annotating the value of Pi to the first 100 decimal places. Each digit was annotated using a specific color and all even numbers were annotated using hollow circles while all odd numbers were annotated using solid circles. These circles were then arranged in a grid of 10x10. Looking at the grid of these annotated digits there were more interesting patterns that emerged. For instance, the first 5 digits from the top in the fourth column were taken by the number 2 and looking at the grid, one could easily count how many times a single digit was repeating itself for the first 100 times. I printed out the final visualization (Figure 5)9 and pasted it on my wall in the studio and soon realized that out of all the visualizations I had done so far, this one was getting the most engagement. Asking my peers as to why they found this visualization so engaging I learnt that most of them were intrigued Figure 4: Demographics of MDes 2022 (Singh, March 2021) by the contrast of color and the placement of the circles. After completing these first experiments, I was able to conclude that large and complex data could achieve engagement using data visualization and data annotation. 14 15 8. Process of labeling information so that machines can use it 9. Figure 5: Annotating the first 100 decimals of Pi, following Anwar Aqeel’s work (Singh, March 2021) – Pg. 16 FRIES of Digital Consent In the online realm, the data we transmit builds our digital 10. Explaining FRIES identities. These digital identities then interact with the identities of to mislead people into doing something other users intermediated by the servers they inhabit. Users on the other hand have no control over the data, the digital identities, or Freely Given: If an interface is designed they normally would not do, the application is not consentful. the servers that occupy it. ‘‘By binding the data into a cell with its Reversible: Users should have the right to own logic, protected by encryption, users could restore autonomy any time. to their digital bodies, allowing interactions to involve them instead of acting upon them’’ (Lee & Toliver 2017). In their zine Building limit access or entirely remove their data Informed: Consentful applications use clear and accessible language to inform Consentful Tech, authors Una Lee and Dann Toliver explain users about the risks they present and the consent in digital technologies with the acronym FRIES10. these important details from the fine print data they are storing, rather than buying of terms and conditions. Enthusiastic: If people are giving up their data because they have to in order to access necessary services and not because they want to, that is not consentful. Specific: A consentful app only uses data the user has directly given, not data acquired through other means like scraping or buying, and uses it only in ways the user has consented to. (Lee & Toliver 2017) 16 17 At this point in my research I had concluded that even though all online services say they have the users’ consent, in reality the way this consent is taken is not often informed. Users are left in the dark when it comes to knowing what goes on with their data and how it is used. But, after conducting some preliminary experiments with data visualization and data annotation, I had evidence that no matter how complex the data is, if it is presented in an inviting manner, it does lead to user engagement which could be the first step towards harnessing the agency to understand it. ‘‘There exists an extensive, if sporadic, body of work on how the structure of diagrams and other visual representations shapes our understanding of their informational content’’ (Ziemkiewicz & Kosara 2008). This led me to the second version of my research question: Would visualizing an entire consent form or a privacy policy statement from an online platform, lead to the same level of engagement from users? Figure 6: Understanding the FRIES of consentful technology, designed using definitions from Building Consentful Tech (Lee & Toliver 2017), (Singh, March 2021) 18 19 Formulating the Idea of Visualizing Consent Forms of Social Media Platforms My aim in this project was to break down consent forms of online 11. Figure 7: Visualizing the consent social media platforms beginning with Pinterest. Visually encode Pg. 22 it and interrogate it to fathom areas where there could be concerns about transparency in data sharing. This was easier said than done. On an average, ‘‘an online consent form is 15 to 20 pages long and could have upto 17,150 words’’ (LePan 2021). To begin visualizing the Privacy Policy of Pinterest, I had to review 4395 words and 48 internal links. This design experiment was based on, ‘‘Visualizing Text Based Data: Identifying the potential of visual knowledge production through design practice’’ (Kasunic & Sweetapple 2015). There is a fine line between Information Visualization and Information Design. While the goal of Information Visualization is to discover the structure of a typically large dataset, Information Design visually expresses a dataset that already has a clear structure. In other words Information Visualization works with data while Information Design works with information. The Privacy Policy of Pinterest as available on the internet has 14 sections. Since the document was complex, I decided to map out the links on the document and track where each link led me. I started with the 14 sections of the consent form as my headings and then documented each internal link that I came across. Some sections on this document had one or no links at all, while some sections had more than 25 links. Breaking down a circle into the first 14 sections gave me a starting point to see how far this visualization could go. After meticulously mapping every link on the document I finally had the visualization of all the internal links on the consent form of Pinterest (Figure 7)11. 20 form of Pinterest.com (Singh, May 2021) I could see the extent of information these consent forms held and As I read more on network visualization, I was also trying to why users were not bothered to read through them. The visualize all the people I had met in person after coming to interesting outcome of this design experiment was that almost all Vancouver. After spending the fall semester of 2020 in India I was internal links on the document led to Pinterest’s Help Center and excited to meet my colleagues and faculty members in links to it appeared 16 times on the visualization, while other links person. Once again I started documenting the names of the people I like Contact Us, Settings, Cookie Policy and Privacy Policy was meeting for the first time in person, along with other appeared 14, 12, 4 and 2 times respectively. Kasunic and information such as, if they were students or faculty members, Sweetapple’s work was a helpful guide in designing this were they males or females, did I meet them on campus or not and visualization. By mapping out all the internal links on the consent if they were part of my cohort or not? The final visualization is form I was able to create different entry points into the document shown on the next page. and navigate easily through the consent form using these links. Figure 8: Network visualization of members of my cohort, I met ‘‘Any network visualization, which is the visualization of with in-person for the first time (Singh, September 2021) – Pg. 26 relationships (edges or links) between data elements or nodes, in its Macro View should provide a bird’s eye view into the network and highlight certain clusters as well as isolated groups within its structures’’ (Lima 2013). For the next part of this project I decided to go through the visualization and mark areas which could lead me to my answers about how user data is used by these platforms. But again, all the links that I followed in the document ended up on Pinterest’s Help Center which was nothing more than a complaint form that users could fill to get their questions answered. Therefore after completing this experiment I concluded that online consent forms only give information about what user data they use, but not how they use it. 24 25 To complete my initial plan of visualizing and comparing consent At this point in my research, I was struggling to find answers to my forms of different social media platforms, I started mapping out the research questions regarding how user data was being used by links on the consent form of WhatsApp (Figure 9). While doing online platforms and how they were able to target content to their this I also read a chapter in Visual Complexity (Lima. M., 2013) users with such accuracy. Visualizing and scanning the Privacy titled Decoding Networks, to get more information out of Policy statements of two popular social media platforms did not Pinterest’s Privacy Policy visualization. As Jacques Bertin writes help. But in a final attempt to find answers using the visualization I in this chapter, ‘‘the graphic is no longer only the representation returned to Data Ink and Graphical Redesign (Tufte 2002) and was of a final simplification, it is a point of departure for the discovery reminded of Tufte’s advice to draw the viewer’s attention to the of these simplifications and the means for their justification. The sense and substance of data, and to be strategic with every graphic. graphic has become, by its manageability, an instrument of Here he instructs his reader to ‘‘Erase non-data ink, within reason. information processing’’ (Bertin 2013). Ink that fails to depict statistical information does not have much interest to the viewer of a graphic, infact sometimes such non-data In the International Review of Information Ethics (Stalder & Sützl ink clutters up the data’’ (Tufte 2002). 2011) authors Felix Stalder and Wolfgang Sützl, write how in the online realm sharing is often confused with copying. The authors After reading through all this and attempting to redesign my have argued that if a piece of information is shared digitally, it visualizations, I was still convinced that ethics in data sharing lay should go less in quantity as opposed to being multiplied which is at a deeper layer inaccessible to the user. For the next phase of my what happens with data. In other words, every time data is ‘shared’, research I intended to find out what this layer was. all that happens is that it multiplies, and users that it is shared with now have the same amount of data. The other question that they As D’ Ignazio and Klein explain, Donna Haraway describes data raise is that once this data is shared, who has ownership of it? visualization, as ‘‘the God trick of seeing everything from Sharing has a feeling of ownership while distribution commodifies nowhere. It is also the most ethically complicated to navigate for your data. Do we really have control of our data, once it is on the the ways in which it masks the people, the methods, the internet? Or are we being tricked into the whole notion of our data questions and the messiness that lies behind clean lines and being secured and still ours? geometric shapes’’ (D’Ignazio & Klein 2020). Hence I was able to conclude that visualizing consent forms might lead to engagement but not answer any questions around the ethics of the use of data and its sharing. 29 30 Figure 9: Visualizing the Terms & Conditions of WhatsApp after being updated in January 2021 (Singh, October 2021) 32 III - INFERENCES Data Exhaust12, Behavioral Data13, Inferences14 and Behavioral Modification Continuing with my research to find out more on the ethics of data 12. Data Exhaust: Trail of data left use and sharing, I was suggested to refer to The Age of computer system users during their online Surveillance Capitalism : The Fight for a Human Future at the activity, behavior and transactions New Frontier of Power written by Shoshana Zuboff. As Zuboff 13. Behavioral Data: Data that describes explains in her book, ‘‘surveillance capitalism is a new economic the observed actions of users or 15 order that claims human experience as free raw material for hidden commercial practices of extraction, prediction, and sales’’ (2019). She further writes that ‘‘Surveillance capitalists learn that 14. Inferences: A conclusion reached on the basis of evidence and reasoning 15. Surveillance Capitalism: Economic raw-material supplies come from intervening in our experience to commodification of personal data with commercial outcomes’’ (Zuboff 2019). She continues to highlight how big tech companies constantly feed on user data and when you have access to so much data about someone, it becomes easier to make predictions about them. While on the internet, users give out personal data in the form of Name, Phone Number, Date of Birth and Email Addresses. But there is a lot of other information that gets recorded outside of the user’s knowledge. This data is called ‘‘Data Exhaust’’ (Zuboff 2019), which is not given out voluntarily but can be inferred by how users access the internet over a certain period. 33 customers extracting human experience is not enough. The most predictive shape our behavior in ways that favor surveillance capitalists’ Figure 10: Visualizing the self referencing aspect of online consent forms (Singh, October 2021) by the activities of an internet or other 34 system centered around the the core purpose of profit-making (Zuboff 2019) Every action by users on the internet, produces a wake of collateral data. The keywords used for search, the likes, the comments, the 16. Data that goes beyond online product and service use Why Does Data Visualization Not Help? shares and even the duration of a browsing session tells a little bit The Age of Surveillance Capitalism, began to answer why I was not more about the user every time. It acts as a digital identity builder able to find any information around the ethics of data sharing in the that tells detailed stories about the users’ thoughts, feelings and consent forms I had worked with before. ‘‘Paper contracts require a interests. ‘‘Companies utilize this information through data physical signature, limiting the burden a firm is likely to impose on driven personalization to exploit personal weakness with a customer by requiring them to read multiple pages of fine print. calculated efficiency’’ (Zuboff 2019). This data is fed back to push Digital terms in contrast are ‘weightless’. They can be targeted content to the user. ‘‘If it’s free then you are not the expanded, reproduced, distributed and archived at no additional product but it’s the gradual, slight, imperceptible change in your costs’’ (Zuboff 2019). This also started to make sense of the own behavior and perception, that is the product’’ argument that Felix Stalder and Wolfgang Sützl make about (The Social Dilemma, Netflix, 2020). In other words, this is sharing and copying data digitally and how if a piece of behavioral modification or habit forming that is not information is shared digitally, it should diminish in quantity as directed by your own choice. ‘‘Under the regime of surveillance opposed to being multiplied. When data multiplies it is not being capitalism, content is a source of behavioral surplus16, as is the shared by definition but is being copied and distributed which behavior of the people who provide the content, as are their raises concerns around its ownership. patterns of connection, communication and mobility, their thoughts and feelings, and the meta-data, expressed in their emoticons, exclamation points, lists, contractions and salutations’’ (Zuboff 2019). 35 36 Formulating the Data and Behavioral Surplus Project As I read more about Data Exhaust and Behavioral Surplus, everything that I could not find in the visualization of the consent forms of Pinterest and WhatsApp was now making sense. As I was scrolling through my LinkedIn feed, I realized that all the posts appearing on my wall, had information about which of my connections had interacted with these posts. For example, which of my connections had liked the post, commented on it or found it insightful. Still not knowing what could be done with this data, I started to document these data traces or tags under each of my connection’s names. Soon, I realized that with just this small piece of information, I was now learning so much more about my connections without ever interacting with them. For instance, I now knew the interests of my connections from the posts they had interacted with, which of my connections had been very active on LinkedIn and ones who had not. Using this information I could also predict what type of content could lead to any kind of interaction from these connections. Figure 11: Examples of Data Exhaust on LinkedIn.com 37 38 REB Application To continue my research in this direction, I planned another experiment of documenting these interactions for all my connections on LinkedIn and then analyzing it to see how much more could I learn about my connections and if I could predict certain interactions from them? Since I was going to work with data from human participants, I had to apply for a Research Ethics Board approval before I could start this project. This thesis research was done to validate the theory of interpreting behavioral data to make inferences about users. The research was conducted using information from my LinkedIn account’s connections which is available to me and has been consented to before sharing. Names of all connections were masked and a written consent was taken from all connections prior to the start of the research to give them a free choice of opting out. The information collected from LinkedIn.com was the basis of the content analysis. The participants were informed of the duration of the project and their occupations were replaced with non-identifiable, generic categories of the industry sector they worked in. The application to the Research Ethics Board can be found in Appendix A. Once the project was approved by the REB, I was able to get consent from 15 of my connections on LinkedIn. I started documenting (Figure 12 – 20) all their interactions that I had access to, through LinkedIn’s ‘See all activity’ feature. To document these interactions, I was taking note of what the connection had posted or shared on their wall, how old was this interaction or post, posts that my connections had liked, commented on or shared and the hashtags they had used (if any). All of this was being done making sure that no data was personally identifying any of my connections. 39 40 Figure 12 – 20: Documenting data from LinkedIn.com (Singh, July 2021) 41 42 started to recognize certain interaction patterns from my connections. For instance, one of my connections always liked a post and then shared it on their wall, one connection was always using the same set of hashtags on everything that they were posting on LinkedIn. Some connections also had specific reactions to posts depending on who had posted it. I was also able to learn about the interaction habits of some of my connection’s connection by going 17. The Inference Cycle: Index Available Data: All online platforms keep a record of their user’s profiles and interactions that they do on these platforms. Patterns emerge from Data: Based on these records there are certain behavioral patterns that emerge for every user. through their posts. Apart from learning about the interaction habits Interpret Data Patterns to make of my connections I was also able to recognize certain changes Analyzing these behavioral interaction being suggested to me by LinkedIn based on my recent activity. One of these was LinkedIn suggesting that I add ‘Open To Work’ to my profile even when I had recently updated my job just weeks before starting this project. I assume that this suggestion was made since I was looking at many different profiles regularly while working on this project. Assumptions: patterns platforms are then able to make predictions about what content would Figure 21: The inference cycle (Singh, August 2021) As these interactions were being documented, I had already lead to more interactions. Compare aggregated meta-data: Having access to meta-data from all users, platforms are also able to predict how two or more users could connect based on their interaction patterns. All these insights proved helpful in understanding how certain prediction algorithms might work for all their users on different online platforms. Based on the findings of this project, this is how I understood the inference cycle17 being used by online platforms (Figure 21). 43 44 ‘‘Designing is a process of pattern synthesis, rather than pattern The log data includes your Internet Protocol address (which we recognition. The solution is not simply lying there among the use to infer your approximate location), the address of and activity data… it has to be actively constructed by the designer’s own on websites you visit that incorporate Pinterest features, (like the efforts’’ (Cross 2006). By the end of this project I began to ‘Save Button’), searches, browser types and settings, the date and understand how online platforms were utilizing user data to study time of the request. We use your activity - such as pins you click, behavioral patterns and make predictions about their users, while boards you create, and any text you add in comment or description also trying to understand where ethical boundaries lay that should - along with information you provided when you first signed up and not be crossed to manipulate users for monetary interests. information form our partners and advertisers to make inferences about you and your preferences. If you create a board about travel, ‘Infer’ – Keyword Search on Online Consent Forms we may infer you are a travel enthusiast. We may also infer information about your education of professional experience based Now that I had learned how inferences were being made based on on your activity (Pinterest Privacy Policy, 2021). user data, I started to investigate if consent forms ever let users know about these inferences beforehand. Doing a quick keyword Completing this keyword search I was sure that this is where the search for ‘infer’ and ‘inferences’ on the consent forms of gaps about the use of data were. ‘‘When the user who is not LinkedIn.com (Figure 22) and Pinterest.com (Figure 23), I found professionally trained to deal with data, becomes the analyst, it is the following texts on them: important he or she knows what is going on’’ (Yau 2013). Information inferred from data described above (e.g. using job titles from a profile to infer industry, seniority and compensation bracket; using graduation dates to infer age or using first names or pronoun usage to infer gender; using your feed activity to infer your interests or using device data to recognize you as a member (LinkedIn Privacy Policy, 2020). 45 46 47 48 Figure 23: ‘infer‘ - keyword search on policy.pinterest.com/en/privacy-policy Figure 22: ‘infer‘ - keyword search on linkedin.com/legal/user-agreement IV - DISSEMINATIONS Internet Cookies – A Presentation How do I get People to Engage with all This? During one of our studio classes we were asked to pitch our research as a product to our class. The idea was to market it and Looking back at my research at this point I realized how much I sell it to our studio mates who were acting as investors. With my had discovered about data, privacy and the ethical use of user data. experience from the Open Studio still fresh in my mind, I decided But the one question that remained unanswered was, how do I to start simple and sell the idea of a game that would help players invite people to engage with what I had learned, so that they could understand how cookies worked on the internet. We have all come have more agency to understand their data? It was around this time across instances when we have been asked to accept cookies on when we had our Open Studios at school, a perfect entering a website. Most of us do not really understand how these opportunity for me to showcase my research and findings to my cookies work and accept it anyways to start using the website. peers and faculty members. After carefully displaying all my work that I had done for the last year, I was expecting engagement, The idea of this game was that the player acts like the content questions, and discussions from people who visited my studio. server to different browsers linked to it. Each browser had a However, on the day of the Open Studio, I realized that most of my session timer and as the server the player had to keep sending work was just making sense to me while everyone else was still content based on the cookies it received from each of these finding it hard to engage with. The most common feedback that I browsers, to keep them engaged for as long as possible. The got was that, this was still too much information which was hard to information collected through the cookies is annotated using understand and was missing context. A similar feedback was different shapes and colors. The slides are shown on the next page. received during my studio class in the Fall of 2021. Taking all this feedback into account I now understood that even though my research was important, it was still too much information to grasp for others and since data and privacy is something that has a very large audience, I had to come up with simpler ways to invite readers into my research. 49 50 51 52 53 54 55 56 57 58 Figure 24 - 31: Slides from the Cookies presentation (Singh, September 2021) The feedback from this presentation was enlightening. Everyone The Inferences Presentation who was part of this presentation engaged and understood what kind of information is collected through internet browser cookies After getting a positive response to the last presentation I decided and why are they collected. to pitch an idea of another game called Inferences in my studio class. For this game the player had to make inferences about users Getting this feedback confirmed that was important to make based on data that was provided to them. Every time the inference complex and abstract concepts around data and privacy tangible was correct, the cutout of a human body would fill up, indicating before people could start engaging with it. This also gave me a that they had learnt more about the user. For every incorrect answer chance to reflect on the why I was not able to strike up interactions the player would be given another chance. The game ends with the during the Open Studio. entire human cutout filled up. A few screens from the prototype are shown on the next page. 59 60 Scan the QR code to view the prototype Link to the project: shorturl.at/ouxBR Figure 32: Screens from the Inferences prototype (Singh, October 2021) After getting positive feedback from both these pitches, I understood that to get people to engage with my research I had to come up with simple and easy ways to give them an entry point into my research. 61 62 Research Narratives As a final project for the Fall semester, I made short narratives with illustrations that would act as precursors to the different findings of my research. Every small project that I had done during my research had led me to new findings and research questions. I wrote this narrative as an attempt to invite my reader into my research. The protagonist of the narrative is Anthony who is a college student and wants to buy himself a new pair of headphones. The different chapters tell what goes on behind the Scan the QR code to view the narrative Link to the project: shorturl.at/agGZ0 scenes as Anthony looks for headphones and what happens once he buys those. After presenting these narratives to my class and getting yet another positive response I was finally able to give my audience an entry point into my research and let them engage with it. Figure 33: Screenshots from the Research Narrative 63 64 V - APPLICATIONS OF INFORMATION DESIGN To date, I have worked on studying existing graphs on ioAirFlow’s online portal and suggest certain changes that could lead to better MITACS - ioAirFlow reception by its clients. These suggested changes have been based on my prior research on using data visualization as a design tool to In November 2021, I started working on a research project with make complex data comprehensive. I have also been carrying out ioAirFlow through MITACS, to visualize complex building Quality Assurance tests on their software to identify areas where analysis data to increase comprehension of the public audience. there are gaps and suggest changes to attain better user experience. This research study focused on the communication and Few graphs that were worked on are as follows: transmission of technical and industrial analytic data, developed by our research partner, ioAirFlow. ioAirFlow tests the internal Temperature v/s Time Graph: environment of commercial buildings by placing sensors The Temperature v/s Time Graph (Figure 34) on the ioAirFlow throughout them that collect data on several variables related to portal represents the temperature over time for the selected thermal comfort and air quality. This raw sensor data identifies location. But when displaying data for ‘All Locations’ it was hard trends related to a building’s indoor environmental quality. Because to read any data as the resulting graph was clustered. the current recipients of ioAirFlow’s data reports are not always data- or engineering-trained professionals, complex data is not To counter this problem all sensors were grouped based on their always understood or acted upon. This research internship focused location (Figure 35), this not only made differentiating between the on ways to develop information design to communicate sensors easy, but also made it possible to isolate (Figure 36) certain ioAirFlow’s data more effectively and efficiently. Using design sensors to make the data readable. methodologies of mixed methods and action research, I developed information design and data visualization to communicate, inform, visually analyze or compare, and prototype scenarios and challenges within ioAirFlow systems. Information design and data visualization can generate greater readability and accessibility in data interpretation and analysis. Communicating the ill health and inefficiencies of building environments through information design, creates agency for its users or receivers, allowing building inhabitants to better understand their own health in relation to air quality. 65 66 67 68 Figure 35: Redesigned version of the Temperature v/s Time graph with sensors categorized according to location (shared with permission from ioAirFlow) Figure 34: Older version of the Temperature v/s Time graph (shared with permission from ioAirFlow) As of writing this thesis I am still working on developing surveys and questionnaires to help ioAirFlow organize and categorize user Figure 36: Redesigned version of the Temperature v/s Time graph (isolated view) (shared with permission from ioAirFlow) feedback. So far, working on this research project has given me the 69 opportunity to test all my research findings from summer in real world scenarios. 70 VI - DISCUSSION AND CLOSING REMARKS 3. Inferences: What are ways in which design can reveal where inferences exist for the benefit of the users rather than for the Conclusion benefit of the corporations? I began this research in information design, hoping that it would This thesis is a work in ‘‘…quantitative research which is defined reveal areas of concerns around the use of data in online consent as empirical research that uses numeric and quantifiable data to forms. But as the research progressed, it moved towards arrive at conclusions. This type of research uses data which can be quantitative research and more expansive practices that measured and independently verified. Such conclusions are based repositioned me as an information and communications designer. on experimentation or on objective and systematic observations and statistics’’ (Muratovski 2016). This research is applicable to anyone who is bound by online consent forms and therefore I am using information Such research is done by measuring attitudes, behaviors and visualization and information design as a means to disseminate perceptions based on observations, or by collection of numerical its findings. I believe that where ever there is design its primary data. This data is then used to prove or disprove ideas or purpose is to communicate with its users and viewers, so being assumptions and the analysis and conclusions are based on a graphic designer in my understanding was solutionist as I was ‘‘...deductive reasoning - a logical process where repeated always designing something that was a solution to something else observations of a certain phenomenon will lead to conclusions and at the same time I wanted to communicate that solution with based on high probability of occurrence’’ (Lewis-Beck et al, 2004). my audience, but through this research that practice has On the internet, users may have the agency to understand the use of transformed into being more discursive. their data but, they are often not aware of where to find it and that is where the gap is. Redesigning these systems to fill these gaps Upon reflection I see my research evolve through 3 stages, where will be the first step towards harnessing that agency which will each stage has been an open enquiry led by a research question. allow users to control their data and understand how it affects their The 3 stages are as follows: decisions and of those around them. 1. Ethics in Social Media: This stage is an investigation into ethical This research is not solutionist but discursive that “…reveals and technology, how could we achieve it and what would it lead to? contests hegemonies” (Kraff 2020) in the online realm and leaves it up to the reader to make the final choice. 2. Understanding Digital Consent: How are consent forms designed, where are the areas of concerns around the ethical use of My research demonstrates that although information design is data and how can Information Visualization help in making large useful in making concepts and actions such as data exhaust, data amounts of data consumable for the public audience? mining and inferences transparent, it does not stop the use of such services. My design practice invites users to understand their role 71 72 in making informed decisions about the use of their data, VII - REFERENCES providing agency and ethical transparency. It uses information design as a tool to visualize crucial but invisible aspects of social Anwar, Aqeel. “Visualizing the Beauty of Pi.” Medium, Towards Data Science, 7 June 2021, media. This can assist designers to understand the structures in https://towardsdatascience.com/visualizing-the-beauty-of-pi-cfeb1dfdd749. which they can contribute to future design and social media users in understanding true consent and the impact of their participation Big Bang Data: Dear Data. (2016, March 3). [Video]. YouTube. https://www.youtube.com/ in social media spaces. watch?v=iqaVe1MCTlA Blower, J. D., Haines, K., Santokhee, A., & Liu, C. L. (2009). GODIVA2: Interactive Visualization of Environmental Data on the Web. Philosophical Transactions: Mathematical, Physical and Engineering Sciences, 367(1890), 1035–1039. http://www.jstor.org/stable/40485759 Burnham, B. (2013). Three Dimensional Visualization of Complex Environmental Data Sets of Variable Resolution. Leonardo, 46(3), 290–291. http://www.jstor.org/stable/23468288 Castells, M. (2009). Communication Power. Oxford, Oxford University Press Catherine D’Ignazio, & Lauren F. Klein. (2020). Data Feminism. The MIT Press. Cross, N. (2006). Designerly Ways of Knowing (2006th ed.). Springer. Dilnot, 2009, Ethics in Design: 10 Questions. In Design Studies: A Reader (p. 180-190). Engebretsen, M., & Kennedy, H. (Eds.). (2020). Data Visualization in Society. Amsterdam University Press. https://doi.org/10.2307/j.ctvzgb8c7 Facebook Is Even Worse Than Anyone Imagined. (2021, October 25). The New Republic. https:// newrepublic.com/article/164142/facebook-whistleblower-broken-beyond-repair General Data Protection Regulation (GDPR) – Official Legal Text. (2019, September 2). General Data Protection Regulation (GDPR). https://gdpr-info.eu/ 73 74 General Data Protection Regulation. (2021, March 19). Wikipedia. https://en.wikipedia.org/wiki/ Imaginaries Lab. (n.d.). New Metaphors – A creative toolkit for generating ideas and reframing General_Data_Protection_Regulation problems. New Metaphors – A Creative Toolkit for Generating Ideas and Reframing Problems. Retrieved February 27, 2021, from http://newmetaphors.com/t General Data Protection Regulation (GDPR) – Official Legal Text. (2019, September 2). General Data Protection Regulation (GDPR). https://gdpr-info.eu/ Ingold, T. (2007). Lines: A Brief History. Routledge. Giaccardi, E. (2005). Metadesign as an Emergent Design Culture. In Leonardo (4th ed., Vol. 38, Jo, J. (2020, January 14). Development of an IoT-Based Indoor Air Quality Monitoring Platform. pp. 342-349). The MIT Press Www.Hindawi.Com. https://www.hindawi.com/journals/js/2020/8749764/ Giorgia Lupi: Data Humanism. (2018, March 13). [Video]. YouTube. https://www.youtube.com/ Karppinen, K., & Puukko, O. (2020). Four Discourses of Digital Rights: Promises and watch?v=6wzu0Kvmvw4&t=1s Problems of Rights-Based Politics. Journal of Information Policy, 10, 304-328. doi:10.5325/jinfopoli.10.2020.0304 Giorgia Lupi: Finding Humanity in Data. (2019, June 26). [Video]. YouTube. https://www.youtube.com/watch?v=IYRhCZ0vvFQ Kasunic, J. L., & Sweetapple, K. (2015). Visualising text-based data: Identifying the potential of visual knowledge production through design practice. Studies in Material Thinking, 13. Giorgia Lupi - Information Designer. (2016, May 3). [Video]. YouTube. https://www.youtube. http://www.materialthinking.org/sites/default/files/papers/0143_SMT_Vol13_P07_ com/watch?v=OaYU-gStjUs Kasunic-Sweetapple_FA.pdf Glancy, R. (2017, February 21). Will you read this article about terms and conditions? You Kim, S., & Paulos, E. (2009). inAir: measuring and visualizing indoor air quality. Proceedings of really should do. The Guardian. https://www.theguardian.com/commentisfree/2014/apr/24/ the 11th international conference on Ubiquitous computing. terms-and-conditions-online-small-print-information Kim, S., & Paulos, E. (2010). InAir: sharing indoor air quality measurements and visualizations. Halpern, O. (2015). Beautiful Data: A History of Vision and Reason since 1945 (Experimental Proceedings of the SIGCHI Conference on Human Factors in Computing Systems. Futures) (Illustrated ed.). Duke University Press Books. Lee, U., & Toliver, D. (2017). Building Consentful Tech. https://www.andalsotoo.net/wp-content/ Hern, A. (2021, January 24). WhatsApp loses millions of users after terms update. The Guardian. uploads/2018/10/Building-Consentful-Tech-Zine-SPREADS.pdf https://www.theguardian.com/technology/2021/jan/24/whatsapp-loses-millions-of-users-afterterms-update LePan, Nicholas. “Visualizing the Length of the Fine Print, for 14 Popular Apps.” Visual Capitalist, 25 Jan. 2021, https://www.visualcapitalist.com/terms-of-service-visualiz- How we can find ourselves in data | Giorgia Lupi. (2017, May 4). [Video]. YouTube. ing-the-length-of-internet-agreements/. https://www.youtube.com/watch?v=sFIDCtRX_-o 75 76 Lima, M. (2017). The Book of Circles: Visualizing Spheres of Knowledge: (with over 300 Sarah, P. (2021, February 18). Following backlash, WhatsApp to roll out in-app banner to better beautiful circular artworks, infographics and illustrations from across history) (Illustrated ed.). explain its privacy update. TechCrunch. https://techcrunch.com/2021/02/18/following-back- Princeton Architectural Press. lash-whatsapp-to-roll-out-in-app-banner-to-better-explain-its-privacy-update/#:~:text=Last%20 month%2C%20Facebook%2Downed%20WhatsApp,India%20and%20various%20regulato- Lima, M. (2013). Visual Complexity: Mapping Patterns of Information (history of information ry%20investigations.&text=But%20the%20backlash%20is%20a,trust%20Facebook%20has%20 and data visualization and guide to today’s innovative applications) (Reprint ed.). Princeton since%20squandered Architectural Press. Singh, M. (2021, January 14). WhatsApp faces legal challenge over privacy in its biggest market. Lupi, G., & Posavec, S. (2016). Dear Data. Adfo Books. TechCrunch. https://techcrunch.com/2021/01/14/whatsapp-faces-legal-challenge-over-privacy-inits-biggest-market/ Lupi, G., & Posavec, S. (2021, December 6). Giorgia Lupi & Stefanie Posavec - Dear Data: A friendship in data, drawing and postcards. Vimeo. https://vimeo.com/157474716 Stalder, F., & Sützl, W. (2011). Ethics of Sharing. The International Review of Information Ethics, 15, 2-2. Retrived from https://informationethics.ca/index.php/irie/article/view/225 Meruyert, N. (2019). CAVISAP: CONTEXT-AWARE VISUALIZATION OF AIR POLLUTION WITH IOT PLATFORMS [Masters Thesis, LuleaUniversity of Technology]. Lulea University of Statt, N. (2021, January 15). WhatsApp to delay new privacy policy amid mass confusion about Technology Publications. https://ltu.diva-portal.org/smash/get/diva2:1353255/FULLTEXT01.pdf Facebook data sharing. The Verge. https://www.theverge.com/2021/1/15/22233257/whatsapp-privacy-policy-update-delayed-three-months Mumtaz, R., Zaidi, S. M. H., Shakir, M. Z., Shafi, U., Malik, M. M., Haque, A., Mumtaz, S., et al. (2021). Internet of Things (IoT) Based Indoor Air Quality Sensing and Predictive T., & R., E. (2001). The Visual Display of Quantitative Information (2nd ed.). Graphics Press. Analytic—A COVID-19 Perspective. Electronics, 10(2), 184. MDPI AG. Retrieved from http://dx.doi.org/10.3390/electronics10020184 Tapper, J. (2018, March 21). Tweet. Twitter.com Muratovski, G. (2016). Research for Designers: A Guide to Methods and Practices (1st ed.). The Art of Data Visualization | Off Book | PBS Digital Studios. (2013, May 9). [Video]. You- SAGE Publications Ltd. Tube. https://www.youtube.com/watch?v=AdSZJzb-aX8&t=371s One Trust. “Understanding the 7 Principles of the GDPR - Blog.” OneTrust, 17 May 2021, Tomboc, K. (2020, September 24). What is Information Design and Why It Matters Now More https://www.onetrust.com/blog/gdpr-principles/. Accessed 21 January 2022. Than Ever. Simple Infographic Maker Tool by Easelly. https://www.easel.ly/blog/what-is-information-design/ Rendgen, S., & Wiedemann, J. (2020). Information Graphics (Multilingual ed.). TASCHEN. Tufte, E. R. (1990). Envisioning Information. Graphics Pr. Risch, John. (2008). On the role of metaphor in information visualization. Weinhardt, M. (2020). Ethical Issues in the Use of Big Data for Social Research. Historical Social Research / Historische Sozialforschung, 45(3), 342-368. doi:10.2307/26918416 77 78 Appendix A Zuboff, S. (2019). The Age of Surveillance Capitalism: The Fight for a Human Future at the New REB Application Frontier of Power (1st ed.). PublicAffairs. ROMEO - Researcher Portal Application for Human Research Ethics - REVISED 2017 Project Info. File No: Ref No : 1898 Project Title: DATA AND BEHAVIORAL SURPLUS Principal Investigator: Dr. Bonne Zabolotney (Faculty of Design + Dynamic Media) Start Date: 2021/10/01 End Date: 2021/12/31 Keywords: Information Design, Ethics Project Team Info. Principal Investigator Prefix: Dr. Last Name: Zabolotney First Name: Bonne Affiliation: Faculty of Design + Dynamic Media Position: Associate Professor Email: bzabolot@ecuad.ca Phone1: Phone2: Fax: Primary Address: Emily Carr University of Art + Design, 520 East 1st Avenue, Vancouver, BC V5T 0H2 Institution: Emily Carr University of Art and Design Country: Canada 79 80 Other Project Team Members 1.7 Are you a student (graduate or undergraduate) applying for ethics approval for a thesis Prefix: Mr. project? Last Name: Singh Yes First Name: Aman Affiliation: Faculty of Graduate Studies 1.8 If you answer ‘Yes’ in 1.7, ensure that the Principal Investigator in this application is your Role In Project: Co-Investigator thesis supervisor and add your name to the “Other Project Member” category at the bottom of the Email: asingh44878@ecuad.ca page. Please click the info button for important instructions. Common Questions 1.9 Have all of the named researchers completed the TCPS2:CORE (Course on Research 1. Project Ethics Details Ethics)? If yes, upload each of the certificates using the attachments tab of this application. No 1.1 Anticipated date that work with participants will begin. application will be processed until all of these certificates are supplied. If you have comparable 2021/10/01 certification from another site, please upload the certification with an explanation. Yes 1.2 Anticipated date that work with participants will end. 2021/12/31 1.10 If you have uploaded comparable certification from a source other than the TCPS2:CORE, please describe here. (Provide a link to the program or an institutional description, if available) 1.3 Type of Project Graduate Thesis Project or Dissertation 1.11 For student researchers (if you answered ‘yes’ in 1.7) - describe any potential conflicts of interest for the researchers such as non-academic benefits, (eg. financial remuneration, patent 1.4 If you have chosen “Other” in the selection above, please ownership, employment, consultancies, board membership, share ownership, stock options, etc.) describe here expected by the researchers, partner organizations, or collaborators as a result of the research. Also describe any non-disclosure agreements or any other restrictions anticipated to affect the 1.5 Does the research fall within the jurisdiction of another research ethics board or body? If so, research all approvals need to be in place before participant research can begin. No 1.6 If you answer ‘Yes’ to question 1.5, please list the names of all of the Research Ethics Board(s) to which you have applied for this project.Include the approval date(s). These dates must match the dates in the certification documents that you attach to this application. Please follow this format: UBC GREB | January 15, 2017 to January 15, 2018 81 82 1.12 Attachments checklist - Ensure that the following documents are attached to this application This data exhaust, when shared with third party advertisers and partners is what fuels the targeted using the Attachments tab. Incomplete application will not be reviewed.- TCPS2:CORE ads industry and gives access to study user behavior.This thesis research aims to validate the certificates (or equivalent) for each of the researchers.- All Recruitment materials (also see theory of interpreting behavioral data to make inferences about users. The research will be ‘Research Participants and Recruitment’ tab)- All consent and release materials (also see the conducted using information from the Co-Investigator’s LinkedIn account’s connections which ‘Consent’ tab)- Other relevant documents is publicly available and has been consented to before sharing. Names of all connections will be masked with the use of pseudo names. A written consent will also be taken from all connections 2. Risk & Review prior to the start of the research to give them a free choice of opting out. 2.1 Researchers are invited to complete the “Risk and Review” assessment to determine the level of risk and review required for their project. An optional tool is 3.2 Methodology (200 words max):Describe this project’s methodological approach to available here: http://www.connect.ecuad.ca/research/reb/applications participant research activities. Include details on what will be expected of participants. Attach survey, interview questions and other documents related to the research methods. Include a 2.2 From the ‘Risk and Review’ assessment, the proposed research project is expected to require timetable for participant research activities. the following (choose one). Do not attach the ‘Risk and Review’ assessment Level 2 - Low Risk Quantitative Data Collection: Using information design models and communication materials we 3. Summary of Proposed Research will develop surveys and questionnaires directed towards users to develop quantitative 3.1 Summary of Proposed Research: Describe the purpose of the proposed research project in feedback and user data. The information collected data scraped from LinkedIn.com will be the non-technical language (200 words max, please see info button for details) Systems that make basis for content analysis. This analysis may also be subject to the development of information decisions about people based on their data, produce substantial adverse effects that can design, as a way to systematically organize and categorize user feedback.Prototyping and massively limit their choices, opportunities, and life chances. The most predictive raw-material User-testing. The content analysis, developed from quantitative data collection, will also be supplies come from intervening in our experiences to shape our behavior in ways that favor reviewed to take user feedback and testing into account, and to make improvements from a user’s capitalists commercial outcomes. While using the internet, we give out personal data in the perspective. User testing will also take place with communication materials, including printed form of Name, Phone number and email address. But there is a lot of other information that gets material and websites. This prototyping phase may require multiple iterations and refinements of recorded outside of our knowledge. This data is called data exhaust, which we do not give out materials, and could lead to further user testing and surveys. Participant expectation: The voluntarily but can be inferred by the way we access the internet over a certain period. participants will be informed of the duration of the project and consent will be taken from each connection before starting the project. Names of the participants will be anonymized and no participant will be identifiable. The occupation of the participants will be replaced with non-identifiable, generic categories of the industry sector they work in. 83 84 3.3 Professional Expertise / Qualifications:If any of the research activities require professional 4.3 Recruitment: Describe how participants will be recruited, and by whom. Attach any materials expertise or recognized qualifications (eg. first aid certification, registration as a clinical that might be used for recruitment (eg. email text, posters, fliers, advertisements, letters, psychologist or counsellor, health practitioner qualifications or expertise, etc), describe them telephone scripts). here. Participants will be informed about the project in the form of a consent form that will be sent to them via email. The consent form will include all the details about the project, including: 4. Research Participants and Recruitment Timeline of the project, Steps to mitigate the risks involved for the participants, Information 4.1 Participants: Indicate the groups that will be targeted for recruitment in the research project. about where the findings will be stored and how the results will be shared with the participants. Describe any specific inclusion or exclusion criteria (example: undergraduate students, specific The participants will be given a free choice to not be a part of the project and data from the age ranges, genders, etc) profile of such participants will not be used for the research. Graduate Students Working Professionals 4.4 Incentives: Will participants be offered incentives to encourage their participation? Researchers No Professors 4.5 If yes to above, describe the incentive plans and the rationale for using incentives. 4.2 Number of Participants: What is the expected number of participants? 30 4.6 Participants and vulnerability: Are there circumstances that cause the participants or participant group(s)to be vulnerable in the context of research? No 4.7 If yes to above, describe the particular way participant vulnerability may be affected by the research and any measures that are planned to address potential risks associated with these vulnerabilities. 85 86 4.8 Are people from First Nations, Inuit, Metis or other Indigenous backgrounds being 5. Risk vs Benefit specifically invited to participate in this research? 5.1 Describe any known or anticipated direct or indirect benefits to the research community or No society that may emerge from the proposed research. This research work will feature in the Co- Investigator’s published thesis and will exist as an 4.9 If yes to above, describe any additional reviews/approvals/consultations/cultural protocols available resource for students and researchers in the field of design, data, ethics and privacy. required to complete this research. Ensure your rationale for engaging with specific individuals or communities is described in 3.1. 5.2 Risks of Research: Check any that apply - list all risks likely to be faced by participants in the proposed research. Personal/Sensitive information: the proposed research involves the disclosure 4.10 Research Locations : Select all locations where participant research will occur. of information that is intimate or sensitive in nature. Emily Carr University 5.3 Describe the risks identified and contextualize them related to risks faced by participants in 4.11 Provide details of the locations listed above everyday activities. See info button for details. The research will be conducted on the Emily Carr University of Art + Design campus in Although the Name, Age, and Pronouns of participants will not be included in the documentation Vancouver, Canada. All the findings and results of the research will be stored in a password of their interactions on LinkedIn and any information that could personally, the participants will protected laptop. Only the Principal Investigator and Co- Investigator will have access to this be masked. However, as a requirement of the research, participant’s area of work will be included information. All the data collected throughout the research will be stale dated after a certain point in the research. All precautions will be taken to keep this information as generic as possible. For of time. example, if the participant is a Product Designer working with Honda Motors, their area of work 4.12 Participant Access to Research Results: Describe your plans to provide or share results of your research with participants. This might include invitations to final presentations or exhibitions, or copies of publications produced. Content here should be consistent with descriptions in the consent forms provided. All the results and findings pertaining to each participant of the research will be shared with them via email in the form of a .pdf document. 87 will appear as Automobile Industry. 5.4 Risk Mitigation: Describe how the researchers will mitigate the risks identified above. Describe whether the researchers have the skills to deal with identified risks or whether additional experts will be recruited. Describe any resources that will be made available to participants. 88 To mitigate the risks involved, Names of the participants will be anonymized and occupations All data collected as part of the research will be stored in a password protected laptop of the will be replaced with non-identifiable generic categories. All the findings and results coming out Co-Investigator. Only the Co-Investigator and Principal Investigator will have access to this data. of this research will be shared with participants in the form of a pdf document via email. It will Any files remaining on the laptop will be permanently deleted. be ensured that results shared will be pertaining to that particular participant only. 7.4 Location of Data: Describe the location for long-term storage of confidential materials 6. Consent 6.1 Consent Documents: Check all of the following consent and release documents that will be All data collected as part of the research will be stored in a password protected laptop of the used in this project. Co-Investigator. Only the Co-Investigator and Principal Investigator will have access to this data. Combined Invitation and Consent Form 8. Monitoring 6.2 Describe any special consent provisions selected above 8.1 Once REB approval has been obtained, it is the responsibility of the PI to maintain the ethics file in up-to-date good standing and make appropriate reports (such as Severe Adverse Event reporting) and amendments (please see Info button for more details). Is it expected that the 7. Confidentiality and Security proposed research will require additional monitoring beyond the minimum annual requirement? 7.1 Confidentiality: Indicate the level of confidentiality built into the research design. No Coded - direct identifiers are removed from the research materials (data) and replaced with coded identifiers. There exists the possibility that with access to the codes it may be possible for a third 8.2 If you answered yes to the above, please describe your plans for this. party to re-identify participants. 8.3 Is it expected that the proposed research will continue beyond the conclusion of this project? 7.2 Describe the rationale for the collection of identifiable research materials No 7.3 Storage & Destruction of Confidential Material: Describe in detail how identifiable materials/ data will be collected, stored, retained and destroyed at the end of the data life cycle. 89 90 8.4 If yes to above, describe in detail. Attachments Doc / Agreement: Approval/Certification from other institutions or partner organizations Version Date: 2021/03/19 File Name: tcps2_core_certificate.pdf Description: TCPS2_Core_Certificate Doc / Agreement: Consent Materials Version Date: 2021/09/26 File Name: Combined Invitation & Consent Form.pdf Description: Combined Invitation and Consent Form for the recruitment of participants. Figure A-1: Combined Invitation and Consent Form for the Data & Behavioral Surplus project 91 92 Figure A-2: TCPS 2: CORE Certificate 93