Anti-Dashboard Design to Combat Online Harassment for a Non-Profit Organisation
Outline
How did I get involved?
I was looking for opportunities to work with visual and interaction designers on a digital product. A cybersecurity friend told me that the Hague Hacks Festival (2018) was an event that matches designers with human rights problems. In one session I participated, four Bangladeshi activists shared their deadly experience with online harassment that escalated to real-life threats and deaths.
If a solution has been agreed upon with activists, why do user research again?
An agreement, even by potential users themselves, does not mean it is the right solution. It was clear from the festival that there was an urgent need for timely information for vulnerable activists to act upon. The dashboard was an attractive option. However, it is unclear whether a new dashboard can address the needs of the activists. Some considerations include the numerous existing products in the market, and the business, technical, and manpower resources required for its development. Is the dashboard a good product fit for activists and the Hague Peace Project organisation hosting the Hague Hacks?
Client
The Hague Peace Projects (HPP) is a volunteer network that supports human rights defenders through education, capacity building, and conflict resolution.
🚴🏼♀️ Step 1: Research Starting Point
Client: We want the dashboard.
My Questions: (Product Agnostic)
I included a broader range of questions to understand the tasks and device use of activists. The dashboard is in my rearview mirror.
- What are the gaps in our understanding of the campaign ecosystem of Bangladeshi high-risk human rights defenders?
- How can we holistically understand the online campaign ecosystem experience of activists?
👨👩👧👦 Step 2: Method and Recruitment
I adapted my research process to suit the fast and slow work cycles in a voluntary organisation.
- Contextual Interviews: five remote interviews were quick, cost-effective, and preferred by participants who could maintain their comfort and security; it allowed us to reach activists from outside the Netherlands
The client mindset shift also required examining existing products in the market and the experiences of software developers and other experts in bringing products to market.
- Competitor Research: heuristic analysis of 233 monitoring tools and expert discussions were conducted in-between interviews and recruitment period lulls (discussed separately)
- Subject Matter Expert Interviews: informational interviews with two specialists: a data manager and a leading cybersecurity consultant who has consulted with non-profits shared their views on data privacy and the business of product development (discussed separately)
For this discussion, I will focus on the five contextual interviews that we did.
Recruitment: For participants, Shucheesmita, HPP’s Bangladesh workgroup leader, tapped into her networks for referrals and participants. The main consideration is the willingness to speak because of the inherently traumatic nature of the topic.
🤹🏽♀️ Step 3: Team Composition and Roles
Alena and I joined up as a team. She spent two months on the project prior to departing for her overseas exchange programme. We were both involved in the project planning, interview execution, transcription, and initial analysis.
Team Work
- Dual remote and in-person interviews
- First level affinity mapping
Solo Work (Melanie)
- Divvied interviews for transcription (two-three each)
- Second-level analysis: persona, device use, risk perception mapping
- Report writing
- Client presentation
🎙️ Step 4: Doing the Interview
(Two Hour) Remote Interviews:
Our dynamic includes topical prompts and role-switching to steer the conversation during emotional periods.
Alena’s enthusiasm is best to establish trust while Melanie’s responsibility was to boomerang the conversation back to project deliverables: device use, campaign process, and risk self-perception.
Sample “Show Me” Prompts:
- Show me an example of a socially acceptable message. What are the different types of messages that you see, name them. What makes them different?
- Can you show me other actors and their networks? How would you describe or label them?
🔎Step 5: Initial Analysis: Affinity Mapping
Key Finding
Our first-level affinity mapping uncovered that activists valued authenticity and identity consistency (offline and online) despite dangerous threats.
Other Results
- The general activist campaign journey was identified.
- Activists were self-censoring themselves as part of their risk prevention and mitigation strategies.
- Identity crises following harassment, bans, and exile dominated activist concerns.
- The security vulnerability of activists is partly due to their commitment to a single identity offline and online.
Learnings:
- Summary and full transcript findings converged for top-level categories in the behaviour maps.
🔎 Step 6: Second Level Analysis - Personas
The personas here present a behaviour spectrum of average tech security and advanced tech-savvy security-aware online users.
Main Task or Need:
To be able to find the relevant and appropriate resources to troubleshoot whatever situation they find themselves in (ranging from Facebook taking down posts, submitting a complaint, how to report to police, etc.)
Desired Outcomes:
- Reduce the feelings of anxiety and uncertainty
- Increase control and security of their online identity and presence
- Quickly implement safe and secure digital practices
Expected Knowledge:
- They are familiar with the Facebook platform
- They are familiar with their mobile phones, laptops, and tablets
Missing Knowledge:
- They don’t know who to turn to for help in a situation
- They don’t know helpful features on their phones or laptops
- They can’t find information and need to ask a friend
State of Mind:
- Uncertain
- Anxious
- Despairing or Agitated
Learning: Demographic data is independent of task behaviour.
🔎 Step 7: Second Level Analysis: Campaign Journey and Device Flow
High-risk activists primarily use mobile for posting and monitoring online posts but switch to laptops when it comes to long-form writing, recording, and reporting harassment screenshots
Activists switch between mobile and laptop use depending on where they are when they first posted their reaction or advocacy.
The activist campaign journey is simplified into four steps:
- online posting
- monitoring
- recording of harassment
- reporting of harassment
High-risk activists accelerated their switching patterns during the monitoring, recording, and reporting phases.
- For mobile-first users, the switch to a laptop happens when they duplicate and export a screenshot of an offending post and proceed to file and store it.
- For mobile-first users, if they don’t want to use their laptop, they proceed to send a copy to themselves, their friends, or back up using other communication apps.
- For reporting to authorities, mobile-first users finally switch to the desktop to compile and print out a report for authorities
- For laptop-first users, they switch to mobile when they want to also send a back-up to their friends on other social media communication apps
Recommendation: Any design intervention requires improving the seamless switching between mobile and laptop devices.
🔎 Step 8: Second Level Analysis: Risk Self Perception
Any perception of serious or mild security or harassment risk is relative. High-risk activists tend to normalise what would be considered dangerous (threats to one’s life and safety) by non-activists.
Low to Medium Risks
These range from posts or replies using vulgar language about a person’s looks or gender or sexual body parts. The mass reporting of their account is serious but has since become commonplace online.
High Risk
The perception of high risk refers to how much emotional distress the activists face. These range from when Facebook blocks their account to receiving offline death threats.
Reaction
The immediate reaction to low to medium-risk harassment is to block the individual or quickly record the incident for sharing. The situation and emotional state of the activist escalate when they are de-platformed and seek human support and tech help for troubleshooting or to use their friend’s pages.
💡 Step 9: Two Quick Solutions
These are quickly actionable, require minimal resource outlay, and tap into the large volunteer manpower that HPP has available.
- Wikipedia Warriors. HPP can mobilise its volunteers to update and detail information in troubleshooting Facebook use and complaints on Wiki sites and Wikipedia.
- YouTube Quick Tips. HPP can also mobilise its volunteers to create short, informational, step-by-step Facebook troubleshooting videos targeting security average users.
💡Step 10 - Pivoting from a Dashboard
I identified three urgent steps for the client to do and pivot from the web-based social media monitoring tool.
- Supporting Evidence. Further research needs to be undertaken to build evidence that looks into other solutions and switch the mindset of the client away from a dashboard.
- From Researcher to Consultant. I needed to switch to a consulting role to identify and align the business and organisational needs of the internal stakeholders with the project.
- Recruiting a Design Team. In the latter half of 2019, I took the initiative to search for designers to build a working prototype.
Project Roadmap
🧭 It was clear that more work is required to finding a workable solution
- a heuristic analysis of other existing platforms and applications
- expert informational interviews on resource planning and business lessons
- wireframing and prototyping a solution
- development partnership with other firms or businesses
Design Principles for Solution
🛠️ Inspired by my current (2021) work with two designers, I’ve come up with two principles and a metric that can be detailed further.
- Seamless switching between mobile and desktop
- Personalisation - let the activist determine what constitutes urgent or dangerous and what information would be relevant
- One metric - measuring when average security users quickly apply help advice to their own practice.
