Designing text-based tools for digital deliberation
Understanding and measuring the influence of certain features on the quality of online text-based deliberations can help us make better design decisions
Co-authored by: Anatol Itten
Citizens want to make their voices heard by policy makers on important decisions that affect their everyday lives. After conducting a series of online public policy evaluation experiments last year, we found that while our participation process meant we were able to include the views of (tens of) thousands of citizens on complex subjects like coronavirus exit strategies or energy transition, we had not provided an online environment that was entirely suited to this task and we needed to rethink the design.
While this kind of participation encouraged citizens to reflect on information and while they found it useful to learn about the trade-offs for proposed alternatives, few were comfortable with making a final decision themselves. Some people were more reflective than others and some people found the questions too complex. It was also clear to us that communication skills don’t come naturally to many people, especially online, and this essential element of the policy evaluation process would need to be encouraged within the platform design.
We realised that for particularly complex policy questions, an online platform that encouraged deliberative ideals would be more suitable. Last but not least, where participatory processes can be performed at a massive scale, deliberation requires a different scale to achieve its full potential.
Deliberation requires participants to reflect, engage respectfully with different viewpoints and give rational reasons for arguments. It is particularly suited for complex issues or wicked problems such as climate change, where there is high uncertainty and many diverse viewpoints. The ideal communication space for deliberation is one of openness, inclusiveness, trust, rationality, and political neutrality. However, most online platforms do not live up to these ideals.
Here we discuss different design features from theory and practice that impact the quality of text-based online deliberation in particular, acknowledging that video interaction and how to combine text and video come with other considerations.
Timing: Synchronous or asynchronous discussion?
Choosing between a synchronous or asynchronous environment creates a trade-off between a more ‘real-life’ discussion experience and a more reflective, inclusive, egalitarian or accessible discussion. Real-time chat or video is more spontaneous and dynamic and this helps build rapport between participants. Asynchronous communication, on the other hand, allows more time for self-reflection, removes location or time restrictions, and increases access for people with slower internet speeds. It is a way to “level the playing field” between the more and less informed public. Some research indicates that asynchronous discussions are likely to produce higher quality deliberations overall.
Privacy: Identification or anonymity?
Choosing between identification or anonymity in digital deliberations creates a number of trade-offs. With anonymity, a more egalitarian environment is possible since people feel more freedom to express their honest, even if unpopular, point of view. Harmful social dynamics are reduced and people stay more focussed on the task at hand. Anonymity can also allow civil servants or people with neutrality obligations to participate.
However, anonymity can imply a loss of accountability and the risk of uncivil behaviour. Reducing anonymity has a positive effect on respectfulness and thoughtfulness and increases transparency, but has a negative effect on engagement — people tend to contribute less to the discussion overall when they are identifiable.
Discussion format: conversation or visualisation?
There is a trade-off between user accessibility and an understandable, well-structured discussion. Most online discussions happen on easy-to-use conversation-based platforms like forums, even though their ability to promote fair and transparent discussion is debatable. Posts organised temporally, rather than topically are more difficult to navigate and connect to each other and content tends to be repeated. New platforms like Kialo now visualise discussions and map out arguments, helping participants to clarify their thinking and better connect information. These platforms may require user training or supervision, but they counter sponsored content and promote fair and rational assessment of alternatives. Nevertheless, for complex problems with a broad range of perspectives, rigid pro/con structures may not be appropriate. Other options include mind maps or systems maps.
Moderation: human or machine?
Having an independent moderator can vastly improve the quality of any discussion, since they can enforce social norms. However, larger scale online deliberations are more challenging and resource-intensive. Moderators also suffer from human bias, as well as time and location constraints. Automated facilitation techniques are therefore an important new avenue of research. Machine learning techniques, NLP or algorithms can assist moderators with tedious tasks and give more equal voice to less willing participants. While such algorithms are definitely useful, we must weigh up replacing human bias with the inherent bias of the algorithms behind automated moderators. Transparency is key.
To sum up: designers may not always realise the extent to which their own worldviews, opinions, or assumptions are embedded in the tools they create. Understanding and measuring the influence of certain features on the quality of online text-based deliberations can help us make better design decisions.
Ruth Shortall, PhD, is a postdoctoral researcher at the faculty of Technology, Policy and Management at Delft Technical University. In a previous life she was a computer programmer. Her research is about deliberative policy appraisal methods and she is particularly interested in the design aspects of deliberative online environments.
Anatol Itten is a post-doctoral researcher in co-creation and mass participation of citizens in government decisions at Delft University of Technology. Anatol is also a co-founder of the Disrupted Societies Institute, a think tank to unravel the dynamics of social divisions and polarization and advised the United Nations Climate Conference COP23 and the German Ministry for the Environment in stakeholder engagement and citizens’ participation.
Are you a practitioner delivering a representative deliberative process fully or partially online? The OECD has put together this survey for practitioners about what they are doing, how, and why. Answers are publicly available from the moment they are submitted in this viewable Airtable database (except for the name, job title and email of the individual filling out the form).
Read the articles in the Digital for Deliberation series:
Digital for deliberation: Catching the deliberative wave
We are opening a discussion on the use of digital tools for deliberative processes, in collaboration with our colleagues working on digital government and public sector innovation.
How can digital tools support deliberation? Join the conversation!
The series will focus on three overarching questions: (1) How can digital tools support representative deliberative processes? (2) What are the limits of using digital tools for representative deliberative processes? (3) In what other contexts could these learnings be applied?
Designing an online citizens’ assembly: A practitioner perspective
One of the core elements of a citizens’ assembly is to create the space for people to meet face-to-face. That is where the magic of citizens’ assemblies lies. Why go online then? Marcin Gerwin offers some ideas.
Engineering for deliberative democracy
Just as the architecture of a meeting hall affects whose voice can be heard, the design of our digital tools provides and forecloses certain political possibilities. Jessica Feldman outlines where engineering decisions need to be made.
Online deliberation: Opportunities and challenges
Lyn Carson in conversation with Graham Smith about transferring face-to-face deliberations to an online environment.
The digital participatory process that fed into the French Climate Assembly
The online contributions from the wider public on the Decidim platform enriched the work of the in-person assembly, writes Eloïse Gabadou.
Digital solutions can complement real world participation — but mustn’t exclude
After public meetings can resume, digital participation will likely grow as a complement to offline events. This will broaden citizen engagement — but we have to be careful it doesn’t freeze people out.
Digital parliaments: Adapting democratic institutions to 21st century realities
The coronavirus crisis should be a catalyst for institutionalising the use of digital tools in parliament, argues French MP Paula Forteza.
Public discussions on Covid-19 lockdown in Scotland
Reflections from government on the challenges of digital engagement by Niamh Webster.
Digital tools to open the judiciary: A perspective from Argentina
Pablo Hilaire writes that by promoting conscious uses of digital technologies in favour of open justice, we have learnt that to facilitate and promote deliberation and participation online, we need to put citizens at the centre, from the design to the collection of data and feedback.
Please note that this is a moderated forum that is subject to the Terms and Conditions and Privacy Policy of the OECD website. The views expressed and arguments employed herein are solely those of the author(s) and do not necessarily reflect the views of the OECD or its member countries. The Organisation cannot be held responsible for possible violations of copyright resulting from the posting of any written material on this website. Your comments to posts are valuable; please contribute in a civil and constructive way. OECD reserves the right to delete or edit any comments before they are published. The OECD Secretariat will not respond systematically to each contribution. Note that the views expressed can be personal views.
