From open assessment to shared understanding: practical experiences

This page is a nugget. The page identifier is Op_en7836
Moderator:Jouni (see all)
Upload data Show results	Unlike most other pages in Opasnet, the nuggets have predetermined authors, and you cannot freely edit the contents. Note! If you want to protect the nugget you've created from unauthorized editing click here

From insight network to open policy practice: practical experiences is a manuscript of a scientific article. The main point is to offer a comprehensive summary of the methods developed at THL/environmental health to support informed societal decision making, and evaluate their use and usability in practical examples in 2006-2018. The manuscript was published in 2020:

Tuomisto, J.T., Pohjola, M.V. & Rintala, T. From insight network to open policy practice: practical experiences. Health Res Policy Sys 18, 36 (2020). https://doi.org/10.1186/s12961-020-00547-3

Title page

From insight network to open policy practice: practical experiences

Short title: From insight network to open policy practice

Jouni T. Tuomisto^1* ORCID 0000-0002-9988-1762, Mikko Pohjola^1,2 0000-0001-9006-6510, Teemu Rintala^1,3 ORCID 0000-0003-1849-235X.

¹ Finnish Institute for Health and Welfare, Kuopio, Finland

² Kisakallio Sport Institute, Lohja, Finland

³ Institute of Biomedicine, University of Eastern Finland, Kuopio, Finland

^* Corresponding author

Email: jouni.tuomisto[]thl.fi

This article describes a decision support method called open policy practice. It has mostly been developed in Finnish Institute for Health and Welfare (THL, Finland) during the last 15 years. Each assessment, case study, and method has been openly described and also typically published in scientific journals. However, this is the first comprehensive summary of open policy practice as a whole (since 2007) and thus gives a valuable overview, rationale, and evaluation for several methodological choices we have made. We have combined methods from several disciplines, including toxicology, exposure sciences, impact assessment, statistical and Bayesian methods, argumentation theory, ontologies, and co-creation to produce a coherent method for scientific decision support.

The article is currently under peer review. You can read about the main topics of the article from Opasnet pages Open policy practice, Shared understanding, Open assessment, and Properties of good assessment.

Abstract

Background

Evidence-informed decision making and better use of scientific information in societal decisions has been an area of development for decades but is still topical. Decision support work can be viewed from the perspective of information collection, synthesis, and flow between decision makers, experts, and stakeholders. Open policy practice is a coherent set of methods for such work. It has been developed and utilised mostly in Finnish and European contexts.

Methods

An overview of open policy practice is given, and theoretical and practical properties are evaluated based on properties of good policy support. The evaluation is based on information from several assessments and research projects developing and applying open policy practice and the authors' practical experiences. The methods are evaluated against their capability of producing quality of content, applicability, and efficiency in policy support, as well as how well they support close interaction among participants and understanding of each other's views.

Results

The evaluation revealed that methods and online tools work as expected, as demonstrated by the assessments and policy support processes conducted. The approach improves the availability of information and especially of relevant details. Experts are ambivalent about the acceptability of openness: it is an important scientific principle, but it goes against many current research and decision making practices. However, co-creation and openness are megatrends that are changing science, decision making and the society at large. Against many experts' fears, open participation has not caused problems in performing high-quality assessments. On the contrary, a key challenge is to motivate and help more experts, decision makers, and citizens to participate and share their views. Many methods within open policy practice have also been used widely in other contexts.

Conclusions

Open policy practice proved to be a useful and coherent set of methods. It guided policy processes toward more collaborative approach, whose purpose was wider understanding rather than winning a debate. There is potential for merging open policy practice with other open science and open decision process tools. Active facilitation, community building and improving the user-friendliness of the tools were identified as key solutions for improving usability of the method in the future.

Keywords

environmental health, decision support, open assessment, open policy practice, shared understanding, policy making, collaboration, evaluation, knowledge crystal, impact assessment

Background

This article describes and evaluates open policy practice, a set of methods and tools for improving evidence-informed policy making. Evidence-informed decision support has been a hot and evolving topic for a long time, and its importance is not diminishing any time soon. In this article, decision support is defined as knowledge work that is performed during the whole decision process (ideating possible actions, assessing impacts, deciding between options, implementing decisions, and evaluating outcomes) and that aims to produce better decisions and outcomes. Here, "assessment of impacts" means ex ante consideration about what will happen if a particular decision is made, and "evaluation of outcomes" means ex post consideration about what did happen after a decision was implemented.

The area is complex, and the key players — decision makers, experts, and citizens or other stakeholders — all have different views on the process, their own roles in it, and how information should be used in the process. For example, researchers often think of information as a way to find the truth, while politicians see information as one of the tools to promote political agendas ultimately based on values. Therefore, a successful method should provide functionalities for each of the key groups.

In the late 1970's, the focus was on scientific knowledge and an idea that political ambitions should be separated from objective assessments especially in the US. Since the 1980's, risk assessment has been a key method to assess human risks of environmental and occupational chemicals. National Research Council specifically developed a process that could be used by all federal US agencies. The report emphasised the importance of scientific knowledge in decision making and scientific methods, such as critical use of data, as integral parts of assessments. Criticism based on observations and rationality is a central idea in the scientific method. The report also clarified the use of causality: the purpose of an assessment is to clarify and quantify a causal path where an exposure to a chemical or other agent leads to a health risk via pathological changes described by the dose-response function of that chemical.

The approach was designed for single chemicals rather than for complex societal issues. This shortcoming was approached in another report that acknowledged this complexity and offered deliberation with stakeholders as a solution, in addition to scientific analysis. An idea was to explicate the intentions of the decision maker but also those of the public. Also, mutual learning about the topic was seen important. There are models for describing facts and values in a coherent dual system. However, practical assessments have found it difficult to successfully perform deliberation on a routine basis. Indeed, citizens often complain that even if they have been formally listened to during a process, the processes need more openness, as their concerns have not contributed to the decisions made.

Western societies have shown a megatrend of increasing openness in many sectors, including decision-making and research. Openness of scientific publishing is increasing and many research funders also demand publishing of data, and research societies are starting to see the publishing of data as a scientific merit in itself. It has been widely acknowledged that the current mainstream of proprietary (as contrast to open access) scientific publishing is a hindrance to spreading ideas and ultimately science. Also governments have been active in opening data and statistics to wide use (data.gov.uk). Governance practices have been developed towards openness and inclusiveness, promoted by international initiatives such as Open Government Partnership (www.opengovpartnership.org).

As an extreme example, a successful hedge fund Bridgewater Associates implements radical openness and continuous criticism of all ideas presented by its workers rather than letting organisational status determine who is heard. In a sense, they are implementing the scientific method in much more rigorous way than what is typically done in science.

In the early 2000's, several important books and articles were published about mass collaboration, wisdom of crowds, crowdsourcing in the government, and co-creation. A common idea of the authors was that voluntary, self-organised groups had knowledge and capabilities that could be much more effectively harnessed in the society than what was happening at the time. Large collaborative projects have shown that in many cases, they are very effective ways to produce high-quality information, as long as quality control systems are functional. In software development, Linux operating system, Git software, and Github platform are examples of this. Also Wikipedia, the largest and most used encyclopedia in the world, has demonstrated that self-organised groups can indeed produce high-quality content.

The five principles of collaboration, openness, causality, criticism, and intentionality (Table 1) were seen as potentially important for environmental health assessment in Finnish Institute for Health and Welfare (THL; at that time National Public Health Institute, KTL), and they were adopted in the methodological decision support work of the Centre of Excellence for Environmental Health Risk Analysis (2002-2007). Open policy practice has been developed during the last twenty years especially to improve environmental health assessments^a. Developers have come from several countries in projects mostly funded by EU and the Academy of Finland (see Funding and Acknowledgements).

Materials for the development, testing, and evaluation of open policy practice were collected from several sources.

Research projects about assessing environmental health risks were an important platform to develop, test, and implement assessment methods and policy practices. Important projects are listed in Funding. Especially the sixth framework programme of EU and its INTARESE and HEIMTSA projects (2005-2011) enabled active international collaboration around environmental health assessment methods.

Assessment cases were performed in research projects and in support for national or municipality decision making in Finland. Methods and tools were developed side by side with practical assessment work (Appendix S1).

Literature searches were performed to scientific and policy literature and websites. Concepts and methods similar to those in open policy practice were sought. Data was searched from Pubmed, Web of Knowledge, Google Scholar, and the Internet. In addition, a snowball method was used: found documents were used to screen their references and authors' other publications to identify new publications. Articles that describe large literature searches and their results include.

Open risk assessment workshops were organised as spin-offs of several of these projects for international doctoral students in 2007, 2008, and 2009. The workshops offered a place to share, discuss, and criticise ideas.

A master's course Decision Analysis and Risk Management (6 credit points) was organised by the University of Eastern Finland (previously University of Kuopio) in 2011, 2013, 2015, and 2017. The course taught open policy practice and tested its methods in course work.

Finally, general expertise and understanding was developed during practical experiences and long-term follow-up of international and national politics.

The development and selection of methods and tools to open policy practice has roughly followed this iterative pattern, where an idea is improved during each iteration, or sometimes rejected.

• A need is identified for improving knowledge practices of a decision process or scientific policy support. This need typically arises from scientific literature, project work or news media.
• A solution idea is developed in aim to tackle the need.
• It is checked whether the idea fits logically in the current framework of open policy practice.
• The idea is discussed in a project team to develop it further and gain acceptance.
• A practical solution (web tool, checklist or similar) is produced.
• The solution is piloted in an assessment or policy process.
• The solution is added into the recommended set of methods of open policy practice.
• The method is updated based on practical experience.

Development of open policy practice started with focus on opening the expert work in policy assessments. In 2007, this line of research produced a summary report about the new methods and tools developed to facilitate assessments. Later, a wider question about open policy practice^b emerged: how to organise evidence-informed decision making in a situation where the five principles are used as the starting point? The question was challenging, especially as it was understood that societal decision making is rarely a single event, but often consists of several interlinked decisions at different time points and sometimes by several decision-making bodies. Therefore, it was seen more as a leadership guidance rather than advice about a single decision.

This article gives the first comprehensive, peer-reviewed description about the current methods and tools of open policy practice since the 2007 report. Case studies have been published along the way, and the key methods have been described in different articles. Also, all methods and tools have been developed online and the full material has been available at Opasnet (http://en.opasnet.org) for interested readers since each piece was first written.

The purpose of this article is to critically evaluate the performance of open policy practice. Does open policy practice have the properties of good policy support? And does it enable policy support according to the five principles in Table 1?

Table 1. Principles of open policy practice. (COCCI principles)

Principle	Description
Collaboration	Knowledge work is performed together in aim to produce shared information.
Openness	All work and all information is openly available to anyone interested for reading and contributing all the time. If there are exceptions, these must be publicly justified.
Causality	The focus is on understanding and describing the causal relations between the decision options and the intended outcomes. The aim is to predict what impacts will likely occur if a particular decision option is chosen.
Criticism	All information presented can be criticised based on relevance and accordance to observations. The aim is to reject ideas, hypotheses — and ultimately decision options — that do not hold against critique.
Intentionality	The decision makers explicate their objectives and decision options under consideration. Also values of other participants or stakeholders are documented and considered.

Open policy practice

Figure 1. Information flows in open policy practice. Open assessments and web-workspaces have an important role as information hubs. They collect relevant information for particular decision processes and organise and synthesise it into useful formats especially for decision makers but also for anyone. The information hub works more effectively if all stakeholders contribute to one place, or alternatively facilitators collect their contributions there.

In this section, open policy practice is described in its current state. First, an overview is given, and then each part is described in more detail.

Open policy practice is a set of methods to support and perform societal decision making in an open society, and it is the overarching concept covering all methods, tools, practices, and terms presented in this article. Its theoretical foundation is on the graph theory and systematic information structures. Open policy practice especially focuses on promoting the openness, flow and use of information in decision processes (Figure 1). Its purpose is to give practical guidance for the whole decision process from ideating possible actions to assessing impacts, deciding between options, implementing decisions, and finally to evaluating outcomes. It aims to be applicable to all kinds of societal decision situations in any administrative area or discipline. An ambitious objective of open policy practice is to be so effective that a citizen can observe improvements in decisions and outcomes, and so reliable that a citizen is reluctant to believe claims that are in contradiction with shared understanding produced by open policy practice.

Open policy practice is based on the five principles presented in Table 1. The principles can be met if the purpose of policy support is set to produce shared understanding (a situation where different facts, values, and disagreements related to a decision situation are understood and documented). The description of shared understanding (and consequently improved actions) is thus the main output of open policy practice (see also Figure 2). It is a product that guides the decision and is the basis for evaluation of outcomes.

This guidance is formalised as evaluation and management of the work and knowledge content during a decision process. It defines the criteria against which the knowledge process needs to be evaluated and managed. It contains methods to look at what is being done, whether the work is producing the intended knowledge and outputs, and what needs to be changed. Each task is evaluated before, during, and after the actual execution, and the work is iteratively managed based on this.

The execution of a decision process is about collecting, organising and synthesising scientific knowledge and values in order to achieve objectives by informing the decision maker and stakeholders. A key part is open assessment that typically estimates the impacts of the planned decision options. Assessment and knowledge production is also performed during the implementation and evaluation steps. Execution also contains the acts of making and implementing decisions; however, they are so case-specific processes depending on the topic, decision maker, and the societal context that they are not discussed in this article.

Figure 2. The three parts of open policy practice. The timeline goes roughly from left to right, but all work should be seen as iterative processes. Shared understanding as the main output is in the middle, expert-driven information production is a part of execution. Evaluation and management gives guidance to the execution.

Shared understanding

Shared understanding is a situation where all participants' views about a particular topic have been understood, described and documented well enough so that people can know what facts, opinions, reasonings, and values exist and what agreements and disagreements exist and why. Shared understanding is produced in collaboration by decision makers, experts, and stakeholders. Each group brings in their own knowledge and concerns. Shared understanding aims to reflect all the five principles of open policy practice. This creates requirements to the methods that can be used to produce shared understanding.

Shared understanding is always about a particular topic and produced by a particular group of participants. Depending on the participants, the results might differ, but with an increasing number of participants, it putatively approaches a shared understanding of the society as a whole. Ideally, each participant agrees that the written description correctly contains their own thinking about the topic. Participants should even be able to correctly explain what other thoughts there are and how they differ from their own. Ideally any participant can learn, understand, and explain any thought represented in the group. Importantly, there is no need to agree on things, just to agree on what the disagreements are about. Therefore, shared understanding is not the same as consensus or agreement.

Shared understanding has potentially several purposes that all aim to improve the quality of societal decisions. It helps people understand complex policy issues. It helps people see their own thoughts from a wider perspective and thus increase acceptance of decisions. It improves trust in decision makers; but it may also deteriorate trust if the actions of a decision maker are not understandable based on shared understanding. It dissects each difficult detail into separate discussions and then collects statements into an overview; this helps to allocate the time resources of participants efficiently to critical issues. It improves awareness of new ideas. It releases the full potential of the public to prepare, inform, and make decisions. How well these purposes have been fulfilled in practice in assessments are discussed in Results.

Test of shared understanding

Test of shared understanding can be used to evaluate how well shared understanding has been achieved. In a successful case, all participants of a decision process give positive answers to the questions in Table 2. In a way, shared understanding is a metric for evaluating how well decision makers have embraced the knowledge base of the decision situation.

Table 2. Test of shared understanding.

Question	Who is asked?
Is all relevant and important information described?	All participants of the decision processes (including knowledge gathering processes)
Are all relevant and important value judgements described? (Those of all participants, not just decision makers.)
Are the decision maker's decision criteria described?
Is the decision maker's rationale from the criteria to the decision described?

Everything that is done aims to offer better understanding about impacts of the decision related to the decision maker's objectives. However, conclusions may be sensitive to initial values, and ignoring stakeholders' views may cause trouble at a later stage. Therefore, other values in the society are also included in shared understanding.

Shared understanding may have different levels of ambition. On an easy level, shared understanding is taken as general guidance and an attitude towards other people's opinions. Main points and disagreements are summarised in writing, so that an outsider is able to understand the overall picture.

On an ambitious level, the idea of documenting all opinions and their reasonings is taken literally. Participants' views are actively elicited and tested to see whether a facilitator is able to reproduce their thought processes. The objective here is to document the thinking in such a detailed way that a participant's views on the key questions of a policy can be anticipated from the description they have given. This is done by using insight networks, knowledge crystals, and other methods (see below). Written documentation with an available and usable structure is crucial, as it allows participation without being physically present. It also spreads shared understanding to decision makers and to those who were not involved in discussions.

Good descriptions of shared understanding are able to quickly and easily incorporate new information or scenarios from the participants. They can be examined using different premises, i.e., a user should be able to quickly update the knowledge base, change the point of view, or reanalyse how the situation would look like with alternative valuations. Ideally, a user interface would allow the user to select input values with intuitive menus and sliders and would show impacts of changes instantly.

Shared understanding as the key objective gives guidance to the policy process in general. But it also creates requirements that can be described as quality criteria for the process and used to evaluate and manage the work.

Evaluation and management

Evaluation is about following and checking the plans and progress of the decisions and implementation. Management is about adjusting work and updating actions based on evaluation to ensure that objectives are reached. Several criteria were developed in open policy practice to evaluate and describe the decision support work. Their purpose is to help participants focus on the most important parts of open policy practice.

Guidance exists about crowdsourced policymaking, and similar ideas have been utilised in open assessment.

Properties of good policy support

There is a need to evaluate an assessment work before, during, and after it is done. A key question is, what makes good policy support and what criteria should be used (see Table 3).

Fulfilling all these criteria is of course not a guarantee that the outcomes of a decision will be successful. But the properties listed have been found to be important determinants of the success of decision processes. In projects utilising open policy practice, poor performance of specific properties could be linked to particular problems observed. Evaluating these properties before or during a decision process could help to analyse what exactly is wrong, as problems with such properties are by then typically visible. Thus, using this evaluation scheme proactively makes it possible to manage the decision making process towards higher quality of content, applicability, and efficiency.

Table 3. Properties of good policy support. Here, "assessment" can be viewed as a particular expert work producing a report about a specific question, or as a wider description of shared understanding about a whole policy process. Assessment work is done before, during, and after the actual decision.

Category	Description	Guiding questions	Related principles
Quality of content	Specificity, exactness and correctness of information. Correspondence between questions and answers.	How exact and specific are the ideas in the assessment? How completely does the (expected) answer address the assessment question? Are all important aspects addressed? Is there something unnecessary?	Openness, causality, criticism
Applicability	Relevance: Correspondence between output and its intended use.	How well does the assessment address the intended needs of the users? Is the assessment question good in relation to the purpose of the assessment?	Collaboration, openness, criticism, intentionality
	Availability: Accessibility of the output to users in terms of e.g. time, location, extent of information, extent of users.	Is the information provided by the assessment available when, where and to whom is needed?	Openness
	Usability: Potential of the information in the output to generate understanding among its user(s) about the topic of assessment.	Are the intended users able to understand what the assessment is about? Is the assessment useful for them?	Collaboration, openness, causality, intentionality
	Acceptability: Potential of the output being accepted by its users. Fundamentally a matter of its making and delivery, not its information content.	Is the assessment (both its expected results and the way the assessment is planned to be made) acceptable to the intended users?	Collaboration, openness, criticism, intentionality
Efficiency	Resource expenditure of producing the assessment output either in one assessment or in a series of assessments.	How much effort is needed for making the assessment? Is it worth spending the effort, considering the expected results and their applicability for the intended users? Are the assessment results useful also in some other use?	Collaboration, openness

Quality of content refers to the output of an assessment, typically a report, model or summary presentation. Its quality is obviously an important property. If the facts are plain wrong, it is more likely to misguide than lead to good decisions. Specificity, exactness, and correctness describe how large the remaining uncertainties are and how close the answers probably are to the truth (compared to some golden standard). In some statistical texts, similar concepts have been called precision and accuracy, although with decision support they should be understood in a flexible rather than strictly statistical sense. Coherence means that the answers given are those to the questions asked.

Applicability is an important aspect of evaluation. It looks at properties that affect how well the decision support can and will be applied. It is independent of the quality of content, i.e. despite high quality, an assessment may have very poor applicability. The opposite may also be true, as sometimes faulty assessments are actively used to promote policies. However, usability typically decreases rapidly if the target audience evaluates an assessment to be of poor quality.

Relevance asks whether a good question was asked to support decisions. Identification of good questions requires lots of deliberation between different groups, including decision makers and experts, and online forums may potentially help in this.

Availability is a more technical property and describes how easily a user can find the information when needed. A typical problem is that a potential user does not know that a piece of information exists even if it could be easily accessed.

Usability may differ from user to user, depending on e.g. background knowledge, interest, or time available to learn the content.

Acceptability is a very complex issue and most easily detectable when it fails. A common situation is that stakeholders feel that they have not been properly heard and therefore any output from decision support is perceived faulty. Doubts about the credibility of the assessor also fall into this category.

Efficiency evaluates resource use when performing an assessment or other decision support. Money and time are two common measures for this. Often it is most useful to evaluate efficiency before an assessment is started. Is it realistic to produce new important information given the resources and schedule available? If more/less resources were available, what value would be added/lost? Another aspect in efficiency is that if assessments are done openly, reuse of information becomes easier and the marginal cost and time of a new assessment decreases.

All properties of decision support, not just efficiency or quality of content, are meant to guide planning, execution, and evaluation of the whole decision support work. If they are always kept in mind, they can improve daily work.

Settings of assessments

Sometimes, a decision process or an assessment may be missing a clear understanding of what should be done and why. An assessment may even be launched in a hope that it will somehow reveal what the objectives or other important factors are. Settings of assessments (Table 4) are used to explicate these so that useful decision support can be provided. Examining the sub-attributes of an assessment question can also help:

• Research question: the actual question of an open assessment
• Boundaries: temporal, geographical, and other limits within which the question is considered
• Decisions and scenarios: decisions and options to assess and scenarios to consider
• Timing: the schedule of the assessment work
• Participants: people who will or should contribute to the assessment
• Users and intended use: users of the final assessment report and purposes of the use

Table 4. Important settings for environmental health and other impact assessments within the context public policy making.

Attribute	Guiding questions	Example categories
Impacts	Which impacts are addressed in assessment? Which impacts are the most significant? Which impacts are the most relevant for decision making?	Environment, health, cost, equity
Causes	Which causes of impacts are recognized in assessment? Which causes of impacts are the most significant? Which causes of impacts are the most relevant for decision making?	Production, consumption, transport, heating, power production, everyday life
Problem owner	Who has the interest, responsibility and/or means to assess the issue? Who actually conducts the assessment? Who has the interest, responsibility and/or power to make decisions and take actions upon the issue? Who are affected by the impacts?	Policy maker, industry, business, expert, consumer, public
Target users	Who are the intended users of assessment results? Who needs the assessment results? Who can make use of the assessment results?	Policy maker, industry, business, expert, consumer, public
Interaction	What is the degree of openness in assessment (and management)? (See Table 5.) How does assessment interact with the intended use of its results? (See Table 6.) How does assessment interact with other actors in its context?	Isolated, informing, participatory, joint, shared

Interaction and openness

In open policy practice, the method itself is designed to facilitate openness in all its dimensions. The dimensions of openness help to identify if and how the work deviates from the ideal of openness, so that the work can be improved in this respect (Table 5).

Table 5. Dimensions of openness in decision making.

Dimension	Description
Scope of participation	Who is allowed to participate in the process?
Access to information	What information about the issue is made available to participants?
Timing of openness	When are participants invited or allowed to participate?
Scope of contribution	Which aspects of the issue are participants invited or allowed to contribute to?
Impact of contribution	How much are participant contributions allowed to have influence on the outcomes? How much weight is given to participant contributions?

Openness can also be examined based on how intensive it is and what kind of collaboration between decision makers, experts, and stakeholders is aimed for. Different approaches are described in Table 6.

Table 6. Categories of interaction within the knowledge-policy interaction framework.

Category	Description
Isolated	Assessment and use of assessment results are strictly separated. Results are provided for intended use, but users and stakeholders can not interfere with the making of the assessment.
Informing	Assessments are designed and conducted according to specified needs of intended use. Users and limited groups of stakeholders may have a minor role in providing information to the assessment, but mainly serve as recipients of assessment results.
Participatory	Broader inclusion of participants is emphasized. Participation is, however, treated as an add-on alongside the actual processes of assessment and/or use of assessment results.
Joint	Involvement and exchange of summary-level information among multiple actors is emphasised in scoping, management, communication, and follow-up of assessment. On the level of assessment practice, actions by different actors in different roles (assessor, manager, stakeholder) remain separate.
Shared	Different actors engage in open collaboration upon determining assessment questions, seeking answers to them, and implementing answers in practice. However, the actors involved in an assessment retain their roles and responsibilities.

These evaluation methods guide the actual execution of a decision process.

Execution and open assessment

Execution is the work during a decision process, including ideating possible actions, assessing impacts, deciding between options, implementing decisions, and evaluating outcomes. Execution is guided by information produced in evaluation and management. The focus of this article is on knowledge processes that support decisions. Therefore, methods to reach or implement a decision are not discussed here.

Open assessment is a method for performing impact assessments using insight networks, knowledge crystals, and web-workspaces (see below). Open assessment is an important part of execution and the main knowledge production method in open policy practice.

An assessment aims to quantify important objectives, and especially compare differences in impacts resulting from different decision options. In an assessment, current scientific information is used to answer policy-relevant questions that inform decision makers about the impacts of different options.

Open assessments are typically performed before a decision is made (but e.g. the city of Helsinki has used both ex ante and ex post approaches with its climate strategy). The focus is by necessity on expert knowledge and how to organise it, although prioritisation is only possible if the objectives and valuations of the decision maker and stakeholders are known. For a list of major open assessments, see Appendix S1.

As a research topic, open assessment attempts to answer this question: "How can factual information and value judgements be organised for improving societal decision making in a situation where open participation is allowed?" As can be seen, openness, participation, and values are taken as given premises. This was far from common practice but not completely new, when the first open assessments were performed in the early 2000's.

Since the beginning, the main focus has been to think about information and information flows, rather than jurisdictions, political processes, or hierarchies. So, open assessment deliberately focuses on impacts and objectives rather than questions about procedures or mandates of decision support. The premise is that if the information production and dissemination are completely open, the process can be generic, and an assessment can include information from any contributor and inform any kind of decision-making body. Of course, quality control procedures and many other issues must be functional under these conditions.

Co-creation

Co-creation is a method for producing open contents in collaboration, and in this context specifically knowledge production by self-organised groups. It is a discipline in itself, and guidance about how to manage and facilitate co-creation can be found elsewhere. Here, only a few key points are raised about facilitation and structured discussion.

Information has to be collected, organised, and synthesised; facilitators need to motivate and help people to share their information. This requires dedicated work and skills that are not typically available among experts nor decision makers. Co-creation also contains practices and methods, such as motivating participation, facilitating discussions, clarifying and organising argumentation, moderating contents, using probabilities and expert judgement for describing uncertainties, or developing insight networks (see below) or quantitative models. Sometimes the skills needed are called interactional expertise.

Facilitation helps people participate and interact in co-creation processes using hearings, workshops, online questionnaires, wikis, and other tools. In addition to practical tools, facilitation implements principles that have been seen to motivate participation. Three are worth mentioning here, because they have been shown to significantly affect the motivation to participate.

• Grouping: Facilitation methods are used to promote the participants' feeling of being important members of a group that has a meaningful, shared purpose.
• Trust: Facilitation builds trust among people that they can safely express their ideas and concerns, and that other members of the group support participation even if they disagree on the substance.
• Respect: Contributions are systematically evaluated according to their merit so that each participant receives the respect they deserve based on their contributions as individuals or members of a group.

Structured discussions are synthesised and reorganised discussions, where the purpose is to highlight key statements, and argumentations that lead to acceptance or rejectance of these statements. Discussions can be organised according to pragma-dialectical argumentation rules or argumentation framework, so that arguments form a hierarchical thread pointing to a main statement or statements. Attack arguments are used to invalidate other arguments by showing that they are either untrue or irrelevant in their context; defend arguments are used to protect from attacks; and comments are used to clarify issues. For an example, see Figure S2-5 in Appendix S2 and links thereof.

The discussions can be natural discussions that are reorganised afterwards or online discussions where the structure of contributions is governed by the tools used. A test environment exists for structured argumentation, and Opasnet has R functions for analysing structured discussions written on wiki pages.

Insight networks

Insight networks are graphs as defined by the graph theory. In an insight network, actions, objectives, and other issues are depicted with nodes, and their causal and other relations are depicted with arrows (aka edges). An example is shown in Figure 3, which describes a potential dioxin-related decision to clean up emissions from waste incineration. The logic of such a decision can be described as a chain or network of causally dependent issues: Reduced dioxin emissions to air improve air quality and dioxin deposition into the Baltic Sea; this has a favourable effect on concentrations in the Baltic herring; this reduces human exposures to dioxins via fish; and this helps to achieve an ultimate objective of reduced health risks from dioxin. Insight networks aim to facilitate understanding, analysing, and discussing complex policy issues.

Figure 3. Insight network about dioxins, Baltic fish, and health as described in the BONUS GOHERR project. Decisions are shown as red rectangles, decision makers and stakeholders as yellow hexagons, decision objectives as yellow diamonds, and substantive issues as blue nodes. The relations are written on the diagram as predicates of sentences where the subject is at the tail of the arrow and the object is at the tip of the arrow. For other insight networks, see Appendix S2.

Causal modelling and causal graphs as such are old ideas, and there are various methods developed for them, both qualitative and quantitative. However, the additional ideas with insight networks were that a) also all non-causal issues can and should be linked to the causal core in some way, if they are relevant to the decision, and therefore b) they can be effectively used in clarifying one's ideas, contributing, and then communicating a whole decision situation rather than just the causal core. In other words, a participant in a policy discussion should be able to make a reasonable connection between what they are saying and some node in an insight network developed for that policy issue. If they are not able to make such a link, their point is probably irrelevant.

The first implementations of insight networks were about toxicology of dioxins and restoration of a closed asbestos mine area^c. In the early cases, the main purpose was to give structure to discussion about and examination of an issue rather than to be a backbone for quantitative models. In later implementations, such as in the composite traffic assessment or BONUS GOHERR project, diagrams have been used for both purposes. Most open assessments discussed later (and listed in Appendix S1) have used insight networks to structure and illustrate their content.

Knowledge crystals

Knowledge crystals are web pages where specific research questions are collaboratively answered by producing rationale with any data, facts, values, reasoning, discussion, models, or other information that is needed to convince a critical, rational reader (Table 7).

Knowledge crystals have a few distinct features. The web page of a knowledge crystal has a permanent identifier or URL and an explicit topic, or question, which does not change over time. A user may come to the same page several times and find an up-to-date answer to the same topic. The answer changes as new information becomes available, and anyone is allowed to bring in new relevant information as long as certain rules of co-creation are followed. In a sense, the answer of a knowledge crystal is never final but it is always usable.

Knowledge crystal is a practical information structure that was designed to comply with the principles of open policy practice. Open data principles are used when possible. For example, openness and criticism are implemented by allowing anyone to contribute but only after critical examination. Knowledge crystals differ from open data, which contains little to no interpretation, and scientific articles, which are not updated. Rationale is the place for new information and discussions, and resolutions about new information may change the answer.

The purpose of knowledge crystals is to offer a versatile information structure for nodes in an insight network that describes a complex policy issue. They handle research questions of any topic and describe all causal and non-causal relations from other nodes (i.e. the nodes that may affect the answer of the node under scrutiny). They contain information as necessary: text, images, mathematics, or other forms, both quantitative and qualitative. They handle facts or values depending on the questions, and withstand misconceptions and fuzzy thinking as well. Finally, they are intended to be found online by anyone interested, and their main message to be understood and used even by a non-expert.

Table 7. The attributes of a knowledge crystal.

Attribute	Description
Name	An identifier for the knowledge crystal. Each page has a permanent, unique name and identifier or URL.
Question	A research question that is to be answered. It defines the scope of the knowledge crystal. Assessments have specific sub-attributes for questions (see section Settings of assessments)
Answer	An understandable and useful answer to the question. It is the current best synthesis of all available data. Typically it has a descriptive easy-to-read summary and a detailed quantitative result published as open data. An answer may contain several competing hypotheses, if they all hold against scientific critique. This way, it may include an accurate description of the uncertainty of the answer, often in a probabilistic way.
Rationale	Any information that is necessary to convince a critical rational reader that the answer is credible and usable. It presents to a reader the information required to derive the answer and explains how it is formed. It may have different sub-attributes depending on the page type, some examples are listed below. Data tell about direct observations (or expert judgements) about the topic. Dependencies tell what is known about how upstream knowledge crystals (i.e. causal parents) affect the answer. Dependencies may describe functional or probabilistic relationships. In an insight network, dependencies are described as arrows pointing toward the knowledge crystal. Calculations are an operationalisation of how to calculate or derive the answer. It uses algebra, computer code, or other explicit methods if possible. Discussions are structured or unstructured discussions about the details of the substance, or about the production of substantive information. On a wiki, discussions are typically located on the talk page of the substance page.
Other	In addition to attributes, it is practical to have clarifying subheadings on a knowledge crystal page. These include: See also, Keywords, References, Related files

There are different types of knowledge crystals for different uses. Variables contain substantive topics such as emissions of a pollutant, food consumption or other behaviour of an individual, or disease burden in a population (for examples, see Figure 3 and Appendix S2.) Assessments describe the information needs of particular decision situations and work processes designed to answer those needs. They may also describe whole models (consisting of variables) for simulating impacts of a decision. Methods describe specific procedures to organise or analyse information. The question of a method typically starts with "How to...". For a list of all knowledge crystal types used at Opasnet web-workspace, see Appendix S3.

Openness and collaboration are promoted by design: knowledge crystals are modular, re-usable, and readable for humans and machines. This enables their direct use in several assessment models or internet applications, which is important for the efficiency of the work. Methods are used to standardise and facilitate the work across assessments.

Open web-workspaces

Insight networks, knowledge crystals, and open assessments are information objects that were not directly applicable at any web-workspace available at the time of development. Therefore, web-workspaces have been developed specifically for open policy practice. There are two major web-workspaces for this purpose: Opasnet (designed for expert-driven open assessments) and Climate Watch (designed for evaluation and management of climate mitigation policies).

Opasnet

Opasnet is an open wiki-based web-workspace and prototype for performing open policy practice, launched in 2006. It is designed to offer functionalities and tools for performing open assessments so that most if not all work can be done openly online. Its name is a short version of Open Assessors' Network and also from Finnish word for guide, "opas". The purpose was to test and learn co-creation among environmental health experts and start opening the assessment process to interested stakeholders.

Opasnet is based on MediaWiki platform because of its open-source code, wide use and abundance of additional packages, long-term prospects, functionalities for good research practices (e.g. talk pages for meta-level discussions), and full and automatic version control. Two language versions of Opasnet exist. English Opasnet (en.opasnet.org) contains all international projects and most scientific information. Finnish Opasnet (fi.opasnet.org) contains mostly project material for Finnish projects and pages targeted for Finnish audiences. A project wiki Heande (short for Health, the Environment, and Everything) requires a password and contains information that can not (yet) be published, but the open alternatives are preferred.

Opasnet facilitates simultaneous development of theoretical framework, assessment practices, assessment work, and supporting tools. This includes e.g. information structures, assessment methods, evaluation criteria, and online software models and libraries.

For modelling functionalities, the statistical software R is used via an R–Mediawiki interface. R code can be written directly to a wiki page and run by clicking a button. The resulting objects can be stored to the server and fetched later by a different code. Complex models can be run with a web browser without installing anything. The server has automatic version control and archival of the model description, data, code, and results.

An R package OpasnetUtils is available (CRAN repository cran.r-project.org) to support knowledge crystals and impact assessment models. It contains the necessary functions and information structures. Specific functionalities facilitate reuse and explicit quantitation of uncertainties: Scenarios can be defined at a wiki page or via a model user interface, and these scenarios can then be run without changing the model code. If input values are uncertain, uncertainties are automatically propagated through the model using Monte Carlo simulation.

For data storage, Opasnet Base, a MongoDB no-sql database, is used. Each dataset must be linked to a single wiki page, which contains all the necessary descriptions and metadata about the data. Data can be uploaded to the database via a wiki page or a file uploader. The database has an open application programming interface for data retrieval.

For more details, see Appendix S4.

Climate Watch

Figure 4. System architecture of the Climate Watch web-workspace.

Climate Watch is a web-workspace primarily for evaluating and managing climate mitigation actions (Figure 4). It was originally developed in 2018-2019 by the city of Helsinki for its climate strategy. Already from the beginning, scalability was a key priority: the web-workspace was made generic enough so that it could be easily used by other municipalities in Finland and globally, and used for evaluation and management of other topics than climate mitigation.

Climate Watch is described in more detail by Ignatius and coworkers. In brief, Climate Watch consists of actions that aim to reduce climate emissions, and indicators that are supposedly affected by the actions and give insights about progress. Actions and indicators are knowledge crystals, and they are causally connected, thus forming an insight network. Each action and indicator has one or more contact people who are responsible for the reporting of progress (and sometimes for actually implementing the actions).

The requirements for choosing the technologies were wide availability, ease of development, and an architecture based on open application programming interfaces or APIs. The public-facing user interface uses the NextJS framework (https://nextjs.org/). It provides support for server-side rendering and search engine optimisation which is based on the React user interface framework (https://reactjs.org/). The backend is built using the Django web framework (https://www.djangoproject.com/) which provides the contact people with an administrator user interface. The data flows to the Climate Watch interface over a GraphQL API (https://graphql.org/). GraphQL is a standard that has the most traction in the web development community because of its flexibility and performance.

Opasnet and Climate Watch have functional similarities but different technical solutions. The user interfaces for end-users and administrators in Climate Watch have similar purposes as MediaWiki in Opasnet; and while impact assessment and model development are performed by using R at Opasnet, Climate Watch uses Python, Dash, and Jupyter.

Open policy ontology

Open policy ontology is used to describe all the information structures and policy content in a systematic, coherent, and unambiguous way. The ontology is based on the concepts of open linked data and resource description framework (RDF) by the World Wide Web Consortium.

The ontology is based on vocabularies with specified terms and meanings. Also the relations of terms are explicit. Resource description framework is based on the idea of triples, which have three parts: subject, predicate (or relation), and object. These can be thought as sentences: an item (subject) is related to (predicate) another item or value (object), thus forming a claim. Claims can further be specified using qualifiers and backed up by references. Insight networks can be documented as triples, and a set of triples using this ontology can be visualised as diagrams of insight network. Triple databases enable wide, decentralised linking of various sources and information.

Open policy ontology (see Appendix S3) describes all information objects and terms described above, making sure that there is a relevant item type or relation to every critical piece of information that is described in an insight network, open assessment, or shared understanding. "Critical piece of information" means something that is worth describing as a separate node, so that it can be more easily found, understood, and used. A node itself may contain large amounts of information and data, but for the purpose of producing shared understanding about a particular decision, there is no need to highlight the node's internal data on an insight network.

The ontology was used with indicator production in the climate strategy of Helsinki and a visualisation project of insight networks.

For a full description of the current vocabulary in the ontology, see Appendix S3 and Figures S2-3 and S2-4 in Appendix S2.

Novel concepts

This section presents novel concepts that have been identified as useful for a particular need and conceptually coherent with open policy practice. However, they have not been thoroughly tested in practical assessments of policy support.

Value profile is a documented list of values, preferences, and choices of a participant. Voting advice applications are online tools that ask electoral candidates about their values, world views, or decisions they would make if elected. The voters can then answer the same questions and analyse which candidates share their values. Nowadays, such applications are routinely developed by all major media houses for every national election in Finland. Thus, voting advice applications produce a kind of value profile. However, these tools are not used to collect value profiles from the public for actual decision making or between elections although such information could be used in decision support. Value profiles are mydata, i.e. data of an individual where they themself can decide who is able see and use it. This requires trusted and secure information systems.

Archetype is an internally coherent value profile of an anonymised group of people. Coherence means that when two values are in conflict, the value profile describes which one to prefer. Archetypes are published as open data describing the number of supporters but not their identities. People may support an archetype in full or by declaring partial support to some specific values. Archetypes aim to save effort in gathering value data from the public, as when archetypes are used, not everyone needs to answer all possible questions. It also increases security since there is no need to handle individual people's potentially sensitive value profiles, when open aggregated data about archetypes suffices.

Political strategy papers typically contain explicit values of that organisation, aggregated in some way from their members' individual values. The strategic values are then used in the organisation in a normative way, implying that the members should support these values in their membership roles. An archetype differs from this, because it is descriptive rather than normative and a "membership" in an archetype does not imply any rights or responsibilities. Yet, political parties could use also archetypes to describe the values of their members.

The use of archetypes is based on an assumption that although their potential number is very large, most of a population's values relevant for a particular policy can be covered with a manageable amount of archetypes. As a comparison, there are usually from two to a dozen significant political parties in a democratic country rather than hundreds. There is also research on human values showing that they can be systematically evaluated using a fairly small amount (e.g., 4, 10, or 19) of different dimensions.

Paradigms are collections of rules to describe inferences that participants would make from data in the system. For example, scientific paradigm has rules about criticism and a requirement that statements must be backed up by data or references. Participants are free to develop paradigms with any rules of their choosing, as long as they can be documented and operationalised within the system. For example, a paradigm may state that when in conflict, priority is given to the opinion presented by a particular authority. Hybrid paradigms are also allowed. For example, a political party may follow the scientific paradigm in most cases but when economic assessments are ambiguous, the party chooses an interpretation that emphasises the importance of an economically active state (or alternatively market approach with a passive state).

Destructive policy is a policy that a) is actually being implemented or planned, making it politically relevant, b) causes significant harm to most or all stakeholder groups, as measured using their own interests and objectives, and c) has a feasible, less harmful alternative. Societal benefits are likely to be greater if a destructive policy is identified and abandoned, compared with a situation where an assessment only focuses on showing that one good policy option is slightly better than another one.

There are a few mechanisms that may explain why destructive policies exist. First, a powerful group can dominate the policymaking to their own benefit, causing harm to others. Second, the "prisoner's dilemma" or "tragedy of commons" makes a globally optimal solution to be unoptimal for each stakeholder group, thus draining support from it. Third, the issue is so complex that the stability of the whole system is threatened by changes. Advice about destructive policies may produce support for paths out of these frozen situations.

An analysis of destructive policies attempts to systematically analyse policy options and identify, describe, and motivate rejection of those that appear destructive. The tentative questions for such an analysis include the following.

• Are there relevant policy options or practices that are not being assessed?
• Do the policy options have externalities that are not being assessed?
• Are there relevant priorities among stakeholders that are not being assessed?
• Is there strong opposition against some options among the experts or stakeholders? What is the reasoning for and science behind the opposition?
• Is there scientific evidence that an option is unable to reach the objectives or is significantly worse than another option?

The current political actions to mitigate the climate crisis are so far from the global sustainability goals that there must be some destructive policies in place. Identification of destructive policies often requires that an assessor looks out of the box and is not restricted to default research questions. In this example, such questions could be: "What is such a policy B that fulfils the objectives of the current policy A but with less climate emissions?", and "Can we reject the null hypothesis that A is better than B in the light of data and all major archetypes?" This approach has a premise that rejection is more effective than confirmation; an idea that was already presented by Karl Popper.

Parts of open policy practice have been used in several assessments. In this article, we will evaluate how these methods have performed.

Methods

The methods of open policy practice were critically evaluated. The open assessments performed (Appendix S1) were used as the material for evaluation. The properties of good policy support (Table 3) were used as evaluation criteria in a similar way as in a previous evaluation. In addition, open policy practice as a whole was evaluated using the categories of interaction (Table 6) and the test of shared understanding (Table 2) as criteria. Key questions in the evaluations were the following. Does open policy practice have the properties of good policy support? And does it enable policy support according to the five principles of open policy practice in Table 1? For each method within open policy practice, these questions were asked: In what way could the method materialise improvements in the property considered? Are there evidence or experiences showing that improvement has actually happened in practice? Has the method shown disadvantages or side effects when implemented?

Results

Different methods of open policy practice were evaluated for their potential or observed advantages and disadvantages according to the properties of good policy support. Major advantages are listed on Table 8. Some advantages, as well as disadvantages and problems, are discussed in more detail in the text. The text is organised along the properties of good policy support, categories of interaction, and test of shared understanding.

Table 8. Methods evaluated based on properties of good policy support. Background colours: white: no anticipated benefit, yellow: potential benefit, green: actual benefit observed in open policy practice materials. Numbers in parentheses refer to the assessments in Appendix S1, Table S1-1. The last row contains general suggestions to improve policy support with respect to each property.

Method	Quality of content	Relevance	Availability	Usability	Acceptability	Efficiency
Co-creation	Participants bring new info (2, 3, 25, 26)	New questions are identified during collaborative work (6, 11)	Draft results raise awareness during work (2, 8, 27)	Readers ask clarifying questions and learn and create understanding through collaboration	Participants are committed to conclusions (2, 8, 27)	Collaboration integrates communication to decision makers and stakeholders (users) into the making, which saves time and effort
Open assessment	It combines functionalities of other methods and enables peer-reviewed assessment models (4, 5, 16)	End-user discussions improve assessment (16, 26, 27)	It is available as draft since the beginning	Standard structure facilitates use (8-9)	Openness was praised (3, 8, 9, 21)	Scope can be widened incrementally (12-16)
Insight network	It brings structure to assessment and helps causal reasoning (8, 9, 11, 16, 17)	It helps and clarifies discussions between decision makers and experts (8, 9)		Readers see what is excluded	It helps to check whether important issues are missing
Knowledge crystal	They streamline work and provide tools for quantitative assessments (e.g. 3, 23, 24)	They clarify questions (1, 6)	It is mostly easy to see where information should be found	Summaries help to understand	They make the intentionality visible by describing the assessment question	Answers can be reused across assessments (12–16, 23-24)
Web-workspace	Its structure supports high-quality content production when moderated (8, 9)	It combines user needs and open policy practice (8, 9)	It offers an easy approach to and archive of materials (16, 21, 23, 26)	The user needs guided the functions developed (8)		It offers a place to document shared understanding and distribute information broadly.
Structured discussion	It helps to moderate discussion and discourages low-quality contributions (2, 30)	It guides focus on important topics (16, 30)		Threads help to focus reading	User feedback has been positive: it helps to focus on key issues (8, 30)	Structure discourages redundancy
Open policy ontology		It gives structure to insight networks and structured discussions (8, 16, 30)		Ontology clarifies issues and relations
Value profile and archetype		Value profiles help to prioritise (8)		Voting advice applications may offer an example	Stakeholders' values are better heard	Archetypes are effective summaries
Paradigm	It motivates clear reasoning	It systematically describes conflicting reasonings			Stakeholders' reasonings are better heard	It helps to analyse inferences of different groups
Analysis of destructive policies		It widens the scope (3, 8)		It emphasises mistakes to be avoided	Focus is on everyone's problems	Lessons learned can be reused in other decisions
Suggestions by open policy practice	Work openly, invite criticism. Use tools and moderation to encourage high-quality contributions (Table 1.)	Acknowledge the need for and potential of co-creation, discussion, and revised scoping. Invite all to policy support work. Characterize the setting (Table 4.)	Design processes and information to be open from the beginning. Use open web-workspaces. (Table 5.)	Invite participation from the problem owner and user groups early on. Use user feedback to visualise, clarify, and target content (Table 6.)	Be open. Clarify reasoning. Acknowledge disagreements. Use the test of shared understanding (Table 2.)	Combine information production, synthesis, and use to a co-creation process to save time and resources. Use shared information objects with open license, e.g. knowledge crystals.

Quality of content

Open policy practice aims at high-quality information for decision makers. One of the ideas is that openness and co-creation enable external experts to see and criticise the content at all times so that corrections can be made. Participation among decision makers, stakeholders, and experts outside an assessment team is typically less common than ideally and requires special effort. The participation has been remarkably higher in projects where special emphasis and effort has been put to dissemination and facilitation, such as the Climate Watch and the Transport and communications strategy (assessments 8 and 26 in Table S1-1). Resources should be allocated to facilitation already when planning a policy process to ensure useful co-creation.

Participation is a challenge also in Wikipedia, where only a few percent of readers ever contribute, and the fraction of active contributors is even smaller. Indeed, the quality of content in Wikipedia is better in topics that are popular and have a lot of contributors.

Active participation did not solve quality control on behalf of the assessors, and it had to be taken care of by usual means. In any case, open policy practice does not restrict the use of common quality control methods and therefore it has at least the same potential to produce high-quality assessments as those using the common methods. The quality of open assessments has been acceptable for publishing in peer-reviewed scientific journals.

Relevance

What is relevant for a decision process can be a highly disputed topic. The shared interaction implies that stakeholders can and should participate in discussions about relevance and revision of scoping when necessary. In other words, everyone is invited to policy support work. The setting of an assessment (Table 4) helps participants to see, what the assessment is about.

The analysis of destructive policies can be used as a method to focus on critical aspects of an assessment and thus increase relevance. For example, Climate Watch has an impact assessment tool that dynamically simulates the total greenhouse gas emissions of Helsinki based on scenarios provided by the user. The tool is able to demonstrate destructive policies: for example, if the emission factor of district heating production does not significantly decrease in ten years, it will be impossible to reach the emission targets of Helsinki. Thus, there are sets of solutions that could be chosen because of their appealing details but that would not reduce the emission factor. The tool explicitly demonstrates that these solutions fail to reach the objectives. It also demonstrates that the emission factor is a critical variable that must be evaluated and managed carefully to avoid destructive outcomes.

Other examples include the Helsinki energy decision assessment (assessment 3 in Table S1-1). It showed that residential wood combustion was a devastating way to heat houses in urban areas and health risks were much larger than with any other heating method. Yet, this is a popular practice in Finland, and there is clearly a need for dissemination about this destructive practice. Also, a health benefit–risk assessment showed that whatever policy is chosen with dioxins and young women, it should not reduce Baltic fish consumption in other population subgroups (assessment 16 in Table S1-1). This is because the dioxin health risk, while small, is concentrated in the population subgroup of young women, while all other subgroups would clearly benefit from increased fish intake.

Availability

The tools and web-workspaces presented in this article facilitated availability of information. In addition, many policy processes were designed in such a way that information was open from the beginning. Increased openness in the society has increased demands to make information available in situations where experts used to keep details to themselves. For example, source codes of assessment models have increasingly been made openly available, and Opasnet made that possible for these assessments.

Timing of availability is critical in a policy process, and assessment results are preferably available early on. This is a major challenge, because political processes may proceed rapidly and change focus, and quantitative assessments take time. A positive example of agility was a dioxin assessment model that had been developed in several projects during a few years (assessment 16 in Table S1-1). When European Food Safety Authority released their new estimates about dioxin impacts on sperm concentration, the assessment model was updated and new sperm concentration results were produced within days. This was possible because the existing dioxin model was modular and using knowledge crystals, so it was rerun after updates in just one part about sperm effects.

Availability of previous versions may be critical. Many experts were reluctant to make their texts available in draft assessments if other people were able edit them, but this fear was often alleviated by the fact that previous versions were always available if needed in Opasnet version control. Availability was also improved as information was produced in a proper format for archiving, backups were produced automatically, and it was easy to produce a snapshot of a final assessment. It was not necessary to copy information from one repository to another, but in a few cases, the final assessments were stored in external open data repositories.

In structured discussion, hierarchical threads increased availability, because a reader did not need to read further if they agreed with the topmost arguments (assessment 30 in Table S1-1). On the other hand, any thread could be individually scrutinised to the last detail if needed.

Usability

Co-creation activities demonstrated the utility of participation and feedback (assessments 6, 8, Table S1-1). Even with good substance knowledge, an assessor cannot know the aspects and concerns a decision maker may have. Usability of information was clearly improved when problem owners and user groups were invited to participate early on. User feedback proved to be very useful to visualise, clarify, and target content.

The climate strategy of Helsinki (assessment 8, Table S1-1) took the usability challenge seriously and developed Climate Watch website from scratch based on open source code modules and intensive user testing and service design. Insight networks and knowledge crystals were basic building blocks of the system architecture. It received almost exclusively positive feedback from both users and experts. Also, a lot of emphasis was put on building a user community, and city authorities, other municipalities, and research institutes have shown interest in collaboration. In contrast, Opasnet was designed as generic tool for all kinds of assessments but without an existing end-user demand. As a result, the penetration of Climate Watch has been much quicker.

Insight network provides a method to illustrate and analyse a complex decision situation, while knowledge crystals offer help in describing quantitative nuances within the nodes or arrows, such as functional or probabilistic relations or estimates. There are tools with both graphical and modelling functionalities, e.g. Hugin (Hugin Expert A/S, Aalborg, Denmark) for Bayesian belief networks and Analytica® (Lumina Decision Systems Inc, Los Gatos, CA, USA) for Monte Carlo simulation. However, these tools are designed for a single desktop user rather than for open co-creation. In addition, they have limited possibilities for adding non-causal nodes and links or free-format discussions about the topics.

Insight networks were often complex and therefore better suited for detailed expert or policy work rather than for general dissemination. Other dissemination methods were needed as well. This was true also for knowledge crystals, although page summaries helped dissemination.

A knowledge crystal is typically structured so that it starts with a summary, then describes a research question and gives a more detailed answer, and finally provides a user with relevant and increasingly detailed information in a rationale. This increased the usability of a page among different user groups. On the other hand, some people found this structure confusing as they did not expect to see all the details of an assessment. Users were unsure about the status of a knowledge crystal page and whether some information was up to date or still missing. This was because many pages were work in progress rather than finalised products. This was clarified by adding status declarations on the tops of pages. Declaring drafts as drafts helped also experts who were uncomfortable in showing their own work before it was fully complete.

Voting advice applications share properties with value profiles and archetypes, and offer material for concept development. The popularity of these applications implies that there is a societal need for value analysis and aggregation. The data has been used to understand differences between individuals and political groups in Finland. With more nuanced data, a set of archetypes can probably be developed to describe common and important values in the population. Some of them may have potential to increase in popularity and form kind of virtual parties that represent population's key values.

Value profiles and paradigms were tested on structured discussions and shared understanding descriptions (assessment 30, Table S1-1). Also Helsinki tested value profiles in prioritising the development of Climate Watch. They were found to be promising and conceptually sound ideas in this context. Data that resembles value profiles are being collected by social media companies, but the data are used to inform marketing actions, often without the individual's awareness, so they are not mydata. In contrast, the purpose of value profile data is to inform societal decisions with consent from its owner rather than nudge the voter to act according to a social media company's wishes. The recent microtargeting activities by Cambridge Analytica and AggregateIQ to use value-profile-like data proved to be very effective in influencing voting decisions. Value profiles are clearly severely underutilised as a tool to inform decisions. We are not aware of systems that would collect value profile data for actual democratic policy support between elections.

Acceptability

A major factor increasing acceptability was whether the stakeholders thought that they had been given all relevant information and whether their concerns had been heard. This emphasised the need to be open and clarify reasonings of different stakeholders. It was also found important to acknowledge disagreements. The test of shared understanding (Table 2.) appeared to be a useful tool in documenting these aspects.

Experts were often reluctant to participate in open assessments because they had concerns about the acceptability of the process. They thought that expertise is not given proper weight, if open participation is allowed. They feared that strong lobbying groups hijack the process. They feared that self-organised groups produce low-quality information or even malevolent dis-information. They often demanded the final say as the ultimate quality criteria, rather than trusted that data, reasoning, and critical discussion would do a better job. In brief, experts commonly thought that it is simply easier and more efficient to produce high-quality information in closed expert groups.

In a vaccine-related assessment (2, Table S1-1), comments and critique were received from both drug industry and vaccine citizen organisations by using active facilitation, and they were all very matter-of-fact. This was interesting, as the same topics caused outrage in social media, but this was not seen on structured assessments. This was possibly because the questions asked were specific and typically required some background knowledge of the topic. Interestingly, one of the most common objections and fears against open assessment was that citizen contributions are ill-informed and malevolent. The experience with open assessments showed that they were not.

Efficiency

Open policy practice combines information production, synthesis, and use to a single co-creation endeavour covering a whole policy process. When successful, this approach saved time and resources because of parallel work and rapid feedback and guidance. However, not all open assessments were optimally designed to maximise co-creation between decision makers and experts. Rather, efficiency was typically achieved when knowledge crystals improved structure and reuse and thus saved resources in assessment modelling.

A common solution to co-operation needs seemed to be a strict division of tasks. Detailed understanding of and contributions to other groups' work and models remained low or non-existent. This was typical in large assessment projects (assessments 4, 5, 7, Table S1-1). On the other hand, most researchers were happy in their own niche and did not expect that other experts could or should learn the details of their work. Consequently, the perceived need for shared tools or open data was often low, which hindered mutual sharing, learning, and reuse.

The implementation phase of Climate Watch, which started in December 2018, involved also citizens, decision-makers, and other municipalities. It was the largest case study so far using open policy practice. It combined existing and produced new climate emission models for municipalities. A long-term objective was to collect detailed input data ideally about the whole country and offer all models to all municipalities, thus maximising reuse.

An important skill in open policy practice was to learn to identify important pieces of relevant information (such as scientific facts, publications, discussions etc.) and to add that information into a proper place in an insight network by using open policy ontology and a reasonable amount of work. The more there was user need for a piece of information, the more time was worth producing it. An ontology helped to do this in practice so that the output was understandable for both humans and computers.

Accumulation of scientific merit was a key motivator for researchers. Policy support work typically did not result in scientific articles. When researchers evaluated the efficiency of their own work, they preferred tasks that produced articles in addition to societal benefit. The same reasoning was seen with open assessments and knowledge crystals, resulting in reluctance to participate. Win-win situations could be found, if policy processes were actively developed into containing research aspects, so that new information would be produced for decision makers but also for scientific audiences.

Categories of interaction

Assessment methods have changed remarkably in forty years. During the last decades, the trend has been from isolated to more open approaches, but all categories of interaction (Table 6) are still in use. The trend among the open assessments (Appendix S1) seemed also to go for more participatory processes. Enabling participation was not enough, as interaction required facilitation and active invitation of decision-makers, experts, and stakeholders. Although openness and participation were available in all the open assessments in theory, only a minority of them actually had enough resources for facilitation to realise good co-creation in practice. In the first open assessments in the early 2000's, people were not familiar even with the concepts of co-creation. In recent examples, especially in the Helsinki climate strategy (assessment 8, Table S1-1), co-creation and openness were insisted by decision-makers, civil servants, and experts alike. There was also political will to give resources for co-creation and facilitation. This resulted in actual shared interaction between all groups.

The example in Helsinki produced interest and enthusiasm in both climate activists and other municipalities. The activists started to self-organise evaluation and monitoring using Climate Watch and ask explanations from civil servants whose actions were delayed. Several municipalities expressed their interest to start using Climate Watch in their own climate work, thus indicating that they had adopted the principles of openness and collaboration. This implies that although the popularity of co-creation increased slowly during previous years, good experiences and awareness increase the rate of change, thus resulting in supra-linear progress in interaction.

Test of shared understanding

Shared understanding clarified complex issues and elicited implicit valuations and reasonings in the open assessments. It facilitated rational discussion about a decision and explicated values of stakeholders e.g. about vaccines (assessments 1, 2 in Table S1-1). It also created political pressure against options that were not well substantiated, e.g. about health effects of food (assessment 31, Table S1-1). Shared understanding was approached even when a stakeholder was ignorant of or even hostile to new insights, or not interested in participating, such as in trip aggregation assessment or health benefit-risk assessment of Baltic fish (assessments 11 and 16, Table S1-1). Then, there was an attempt to describe stakeholders' views based on what other people know about their values. Everyone's views are seen as important policy-relevant information that may inform decision making.

Shared understanding was a well accepted idea among many decision makers in Finland. This was observed in collaboration with Prime Minister's Office of Finland (assessment 27, Table S1-1). Many civil servants in ministries liked the idea that sometimes it is better to aim to understanding rather than consensus. They soon adopted the easy version of the term and started to use it in their own discussions and publications.

However, shared understanding was not unanimously accepted. Experts were often reluctant to start scientific discussions with citizens, especially if there were common or strong false beliefs about the topic among the public. In such cases, a typical argument was that the role of an expert is to inform and, if possible, suppress false beliefs rather than engage in producing common descriptions about differing views. The target seemed to be to convince the opponent rather than increase understanding among the audience.

The test of shared understanding was a useful tool to recognise when not all values, causal chains or decision makers' rationale were known and documented. Yet, lack of time or resources often prevented further facilitation, information collection, or expansion of the scope of an assessment.

Discussion

This article presents methods and tools designed for decision support. Many of them have already been successfully used, while others have been identified as important parts of open policy practice but have not been extensively tested.

The discussion is organised around the five principles of open policy practice: collaboration, openness, causality, criticism, and intentionality. The principles are looked at in the light of popularity, acceptance, and lessons learned from practical experience.

The five principles are not unique for open policy practice; on the contrary, they have been borrowed from various disciplines (for reviews, see ). The aim was to use solid principles to build a coherent set of methods that gives practical guidance to decision support. It is reassuring that many principles from the original collection have increased in popularity in the society. There are also studies comparing parts of open policy practice to other existing methods

The results showed that the methods connected the five principles quite well to the properties of good policy support (Table 8). Open collaboration indeed resulted in high-quality content when knowledge crystals, web-workspaces and co-creation were utilised. End-user interaction and structured discussions helped to revise scoping and content, thus improving relevance and usability. Acknowledging disagreements and producing shared understanding created acceptability. And openly shared information objects such as data and models improved availability and efficiency.

The experiences about open policy practice demonstrate that it works as expected when the participants are committed to collaborate using the methods, practices, and tools. However, there have been less participants in most open assessments than what had been hoped for. This can partly be affected by own actions, as reader and contributor numbers clearly went up with active facilitation or marketing with large media coverage and public interest. Some other reasons cannot be easily affected directly, such as inertia to change established practices or lack of scientific merit. Thus, a major long-term challenge is to build an attractive assessor community, culture, and incentives for decision support.

The GovLab in New York is an example of such activity (www.thegovlab.org). They have expert networks, training, projects, and data sources available to improve policy support. There is a need for similar tools and training designed to facilitate a change elsewhere. New practices could also be promoted by developing ways to give scientific — or political — merit and recognition more directly based on online co-creation contributions. The current publication counts and impact factors — or public votes — are very indirect measures of scientific or societal importance of the information or policies produced.

Knowledge crystals offer a collaboration forum for updating scientific understanding about a topic in a quicker and easier way than publishing scientific articles. Knowledge crystals are designed to be updated based on continuous discussion about the scientific issues (or valuations, depending on the topic) aiming to back up conclusions. In contrast, scientific articles are expected to stay permanently unchanged after publication. Articles offer little room for deliberation about the interpretation or meaning of the results after a manuscript is submitted: reviewer comments are often not published, and further discussion about an article is rare and mainly occurs only if serious problems are found. Indeed, the current scientific publishing system is poor in correcting errors via deliberation.

Shared understanding is difficult to achieve if the decision maker, media environment, or some political groups are indifferent about or even hostile against scientific knowledge or public values. For many interest groups, non-public lobbying, demonstrations and even spreading faulty information are attractive ways of influencing the outcome of a decision. These are problematic methods from the perspective of open policy practice, because they reduce the availability of important information in decision processes.

Further studies are needed on how open, information-based processes could be developed to be more tempting to groups that previously have preferred other methods. A key question is whether shared understanding is able to offer acceptable solutions to disagreeing parties and alleviate political conflict. Another question is whether currently under-represented groups have better visibility in such open processes. Also, more information is needed about how hostile contributions get handled, when they occur; fortunately, they were very rare in the open assessments.

There is no data about open policy practice usage in a hostile environment. Yet, open policy practice can be collaboratively used even without support from a decision maker or an important stakeholder. Although their objectives values are important for an assessment, these may be either deduced indirectly from their actions, or even directly replaced by the objectives of the society at large. Thus, open policy practice is arguably a robust set of methods that can be used to bypass non-democratic power structures and focus on the needs of the public even in a non-optimal collaboration environment.

There is still a lot to learn about using co-created information in decision making. Experiences so far have demonstrated that decision making can be more evidence-informed than what it typically is, and several tools promoting this change are available.

Openness in science is a guiding principle and current megatrend, and its importance has been accepted much more widely during recent years. Yet, the practices in research are changing slowly, and many current practices are actually in conflict with openness. For example, it is common to hide expert work until it has been finalised and published, to publish in journals where content is not freely available, and to not open the data used.

A demand to produce assessments openly and describe all reasoning and data already from the beginning was often seen as an unreasonable requirement and made experts reluctant to participate. This observation raised two opposite conclusions: either that openness should be incentivised and promoted actively in all research and expert work, including decision support; or that openness as an objective hinders expert work and should be rejected. The latter conclusion was strong among experts in the early open assessments, but the former one has gained popularity.

There are several initiatives to open scientific processes, such as Open Science Foundation (www.osf.io). These are likely to promote change in science at large and indirectly also in scientific support of decision making.

Among experts, causality was seen as a backbone of impact modelling. In political arenas, causal discourse was not as prominent, as it was often noticed that there was actually little solid information about the most policy-relevant causal chains, and therefore values dominated policy discussions. Climate Watch was the most ambitious endeavour in the study material to quantify all major causal connections of a climate action plan. The approach was supported by the city administration and stakeholders alike. Causal quantification created an additional resource need that was not originally budgeted. It is not yet known, how Helsinki, other cities, and research institutes will distribute the resources and tasks of causal modelling and information produced. Yet, actions in the national energy and climate plans total 260 billion euro per year in EU. So, even minor improvements in the efficiency or effectiveness of climate actions would make causal assessments worthwhile.

Criticism has a central role in the scientific method. It is applied in practical situations, because rejecting poor statements is easier and more efficient than trying to prove statements true. Most critique in open assessments was verbal or written discussion between participants, focussing on particular, often detailed topics. Useful information structures have been found for criticism, notably structured discussions that can target any part of an assessment (scope, data, premises, analyses, structure, results etc).

The current practices of open criticism in research are far from optimal, as criticism rarely happens. Pre-publishing peer review is almost the only time when scientific work is criticised by people outside the research group, and those are typically not open. A minute fraction of published works are criticised openly in journals; a poor work is simply not cited and subsequently forgotten. Interestingly, some administrative processes follow scientific principles better than many research processes do: for example, environmental impact assessment has a compulsory process for open criticism at both design and result phases.

Intentionality requires that the objectives and values of stakeholders in general and decision makers in particular are understood. In the studied assessments, some values were always identified and documented. But it was not common to systematically describe all relevant values, or even ensure that the assessed objectives were actually the most important ones for the decision maker. There is clearly a need prioritise facilitation and interaction about values.

In shared understanding, some claims were found unsubstantiated or clearly false. On the societal level, open policy practice aimed to increase political pressure against decisions based on poor ideas by explicating the problems and informing the public about them. The purpose was not to pressure individuals to reject their unsubstantiated thoughts. Personal beliefs were understood rather than threatened, because the aim was to build acceptance and facilitated contributions. However, it is not known what happens with very sensitive personal topics, because there were no such issues in the studied assessments.

Politics in western democracies is typically based on a premise that ultimately the citizens decide about things by voting. Therefore, in a sense, people can not vote "wrong". In contrast, open policy practice is based on a premise that the objectives of the citizens are the ultimate guiding principle, and it is a matter of discussion, assessment, and other information work to suggest which paths should or should not be taken to reach these objectives. This thinking is close to James Madison's ideas about democracy in Federalist 63 from 1788.. In this context, people vote wrong if they vote for an option that is incapable of delivering the outcomes that they want.

If people are well-informed and have time and capability of considering different alternatives, the two premises lead to similar outcomes. However, recent policy research has shown that this prerequisite is often not met, and people can be and increasingly are being mislead, especially with modern microtargeting tools. The need for protecting people and decision making from misleading information has been recognised.

Public institutions such as independent justice system, free press, and honest civil servants provide protection against misleading activities and disruptive policies. These democratic institutions have deteriorated globally and in some countries particularly, even in places with good track record.

Destructive policies may be an effective way to inform stakeholders in a grim societal environment. Open policy practice may not be very effective in choosing the best alternative among good ones, but it is probably more effective in identifying and rejecting poor alternatives, i.e. destructive policies, which is often more important. This is expected to result in more stable and predictable policies. It is possible to focus on disseminating information about what actions especially should not be taken, why, and how it is known. In such discourse, the message can be practical, short, clear, and rationale is available for anyone interested. Practical experiments are needed to tell, whether this could reduce the support of destructive policies among the public.

Further research is also needed to study other aspects of destructive policies: Can such policies be unambiguously recognised? Is shared understanding about them convincing enough among decision makers to change policies? Does it cause objections about science being biased and partisan? Does open policy practice prevent destructive policies from gaining political support?

Conclusions

In conclusion, open policy practice works technically as expected. Open assessments can be performed openly online. They do not fail due to reasons many people think they will, namely low quality contributions, malevolent attacks or chaos caused by too many uninformed participants; these phenomena are very rare. Shared understanding has proved to be a useful concept that guides policy processes toward more collaborative approach, whose purpose is wider understanding rather than winning.

However, open policy practice has not been adopted in expert work or decision support as expected. A key hindrance has been that the initial cost of learning and adopting new tools and practices has been higher than what an expert is willing to pay for participation in a single assessment, even if its impacts on the overall process are positive. The increased availability, acceptability, and inter-assessment efficiency have not yet been fully recognised by the scientific or policy community.

Active facilitation, community building and improving the user-friendliness of the tools were identified as key solutions in improving usability of the method in the future.

List of abbreviations

• THL: Finnish Institute for Health and Welfare (government research institute in Finland)
• IEHIAS: Integrated Environmental Health Impact Assessment System (a website)
• RDF: resource description framework

Declarations

• Ethics approval and consent to participate: Not applicable
• Consent for publication: Not applicable
• Availability of data and materials: The datasets generated and/or analysed during the current study are available at the Opasnet repository, http://en.opasnet.org/w/Open_policy_practice
• Competing interests: The authors declare that they have no competing interests.
• Funding: This work resulted from the BONUS GOHERR project (Integrated governance of Baltic herring and salmon stocks involving stakeholders, 2015-2018) that was supported by BONUS (Art 185), funded jointly by the EU, the Academy of Finland and and the Swedish Research Council for Environment, Agricultural Sciences and Spatial Planning. Previous funders of the work: Centre of Excellence for Environmental Health Risk Analysis 2002-2007 (Academy of Finland), Beneris 2006-2009 (EU FP6 Food-CT-2006-022936), Intarese 2005-2011 (EU FP6 Integrated project in Global Change and Ecosystems, project number 018385), Heimtsa 2007-2011 EU FP6 (Global Change and Ecosystems project number GOCE-CT-2006-036913-2), Plantlibra 2010-2014 (EU FP7-KBBE-2009-3 project 245199), Urgenche 2011-2014 (EU FP7 Call FP7-ENV-2010 Project ID 265114), Finmerac 2006-2008 (Finnish Funding Agency for Innovation TEKES), Minera 2010-2013 (European Regional Development Fund), Scud 2005-2010 (Academy of Finland, grant 108571), Bioher 2008-2011 (Academy of Finland, grant 124306), Claih 2009-2012 (Academy of Finland, grant 129341), Yhtäköyttä 2015-2016 (Prime Minister's Office, Finland), Ympäristöterveysindikaattori 2018 (Ministry of Social Affairs and Health, Finland). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
• Authors' contributions: JT and MP jointly developed the open assessment method and open policy practice. JT launched Opasnet web-workspace and supervised its development. TR developed OpasnetUtils software package from an original idea by JT and implemented several assessment models. All authors participated in several assessments and discussions about methods. JT wrote the first manuscript draft based on materials from MP and TR. All authors read and approved the final manuscript.
• Acknowledgements: We thank Einari Happonen and Juha Villman for their work on developing Opasnet, and Juha Yrjölä and Tero Tikkanen for developing Climate Watch; and Arja Asikainen, John S. Evans, Alexanda Gens, Patrycja Gradowska, Päivi Haapasaari, Sonja-Maria Ignatius, Suvi Ignatius, Matti Jantunen, Anne Knol, Sami Majaniemi, Päivi Meriläinen, Kaisa Mäkelä, Raimo Muurinen, Jussi Nissilä, Juha Pekkanen, Mia Pihlajamäki, Teemu Ropponen, Kalle Ruokolainen, Simo Sarkki, Marko Tainio, Peter Tattersall, Hanna Tuomisto, Jouko Tuomisto, Matleena Tuomisto, and Pieta Tuomisto for crucial and inspiring discussions about methods and their implementation, and promoting these ideas on several forums.

Endnotes

^a This paper has its foundations on environmental health, but the idea of decision support necessarily looks at aspects seen relevant from the point of view of the decision maker, not from that of an expert in a particular field. Therefore, this article and also the method described are deliberately taking a wide view and covering all areas of expertise. However, all practical case studies have their main expertise needs in public health, and often specifically in environmental health. ^b Whenever this article presents a term in italic (e.g. open assessment), it indicates that there is a page at the Opasnet web-workspace describing that term and that it can be accessed using a respective link (e.g. http://en.opasnet.org/w/Open_assessment). ^c Insight network was originally called pyrkilo (and at some point also extended causal diagram). The word and concept pyrkilo was coined in 1997. In Finnish, pyrkilö means "an object or process that tends to produce or aims at producing certain kinds of products." The reasoning for using the word was that pyrkilo diagrams and related structured information such as models tend to improve understanding and thus decisions. The first wiki website was also called Pyrkilo, but the name was soon changed to Opasnet.

Figures and tables

Move them here for submission.

Appendix S1: Open assessments performed

A number of open assessments have been performed in several research projects (see the funding declaration) and health assessments since 2004. Some assessments have also been done on international Kuopio Risk Assessment Workshops for doctoral students in 2007, 2008, and 2009 and on a Master's course Decision Analysis and Risk Management (6 credit points), organised by the University of Eastern Finland (previously University of Kuopio) in 2011, 2013, 2015, and 2017.

More assessments can be found at Opasnet page Category:Assessments.

Table S1-1. Some environmental health assessments performed using open assessment. References give links to both an assessment page and a scientific publication as applicable.

Topic	#	Assessment	Year	Project
Vaccine effectiveness and safety	1	Assessment of the health impacts of H1N1 vaccination	2011	In-house, collaboration with Decision Analysis and Risk Management course
	2	Tendering process for pneumococcal conjugate vaccine	2014	In-house, collaboration with the National Vaccination Expert Group
Energy production, air pollution and climate change	3	Helsinki energy decision	2015	In-house, collaboration with city of Helsinki
	4	Climate change policies and health in Kuopio	2014	Urgenche, collaboration with city of Kuopio
	5	Climate change policies in Basel	2015	Urgenche, collaboration with city of Basel
	6	Availability of raw material for biodiesel production	2012	Jatropha, collaboration with Neste Oil
	7	Health impacts of small scale wood burning	2011	Bioher, Claih
	8	Climate strategy of Helsinki: Carbon neutral Helsinki 2035 action plan	2018	In-house, collaboration with city of Helsinki
	9	Climate mitigation of the social affairs and health sector in Finland	2020	In-house, commissioned by the Prime Minister
Health, climate, and economic effects of traffic	10	Gasbus - health impacts of Helsinki bus traffic	2004	Collaboration with Helsinki Metropolitan Area
	11	Cost-benefit assessment on composite traffic in Helsinki	2005	In-house
Risks and benefits of fish consumption	12	Benefit-risk assessment of Baltic herring in Finland	2015	Collaboration with Finnish Food Safety Authority
	13	Benefit-risk assessment of methyl mercury and omega-3 fatty acids in fish	2009	Beneris
	14	Benefit-risk assessment of fish consumption for Beneris	2008	Beneris
	15	Benefit-risk assessment on farmed salmon	2004	In-house
	16	Benefit-risk assessment of Baltic herring and salmon intake	2018	BONUS GOHERR
Dioxins, fine particles	17	TCDD: A challenge to mechanistic toxicology	1999	EC ENV4-CT96-0336
	18	Comparative risk assessment of dioxin and fine particles	2007	Beneris
Plant-based food supplements	19	Compound intake estimator	2014	Plantlibra
Health and ecological risks of mining	20	Paakkila asbestos mine	1999	In-house
	21	Model for site-specific health and ecological assessments in mines	2013	Minera
	22	Risks of water from mine areas	2018	Kaveri
Water safety	23	Water Guide for assessing health risks of drinking water contamination	2013	Conpat
	24	Bathing Water Guide for assessing health risks of bathing water contamination	2019	Water Guide update
Organisational assessments	25	Analysis and discussion about research strategies or organisational changes within THL	2017	In-house
	26	Transport and communication strategy in digital Finland	2014	Collaboration with the Ministry of Transport and Communications of Finland
Information use in government or municipality decision support	27	Case studies: Assessment of immigrants' added value; Real-time co-editing, Fact-checking, Information design	2016	Yhtäköyttä, collaboration with Prime Minister's Office
	28	Evaluation of forest strategy process for Puijo, Kuopio	2012	In-house
Indicator development	29	Environmental health indicators in Finland	2018	Ympäristöterveysindikaattori
Structuring discussions	30	Developing and testing tools and practices for structured argumentation	2019	Citizen Crystal
Food safety and diet	31	Health risks of chemical and microbial contaminants and dietary factors in food in Finland	2019	Ruori, collaboration with e.g. Ministry of Agriculture and Prime Minister's Office

Appendix S2: Examples of insight networks

• 

  Figure S2-1. Insight network about dioxins in Baltic fish. The focus is on reasoning and value judgements and their connections to causal chains about dioxins and health.

• 

  Figure S2-2. Legend for main object types used in insight networks. The actual colours and formatting depend on the capabilities of the software used.

• 

  Figure S2-3. Items in open policy ontology shown in an insight network format. All relations are of types 'has subclass' or 'has part'.

• 

  Figure S2-4. Relations in open policy ontology shown in an insight network format. In this graph, relations are shown as nodes, and the arcs between relations are of types 'has subclass', 'has part', or 'inverse'.

• 

  Figure S2-5. Structured discussion about risks of open governance. The original discussion was held in Finnish and can be found from http://fi.opasnet.org/fi/Keskustelu:Jaettu_ymm%C3%A4rrys. Trapezoids are statements. Light blue is an opening fact statement ('Openness causes serious problems to the quality of policy making.'), and blue is closing fact statement that is updated based on the discussion ('Openness causes serious problems to the quality of policy making, if it is too tightly connected to impulsive thinking in social media.'). Orange arguments are true and gray are false. Red arrow is an attack, green is a defense, and gray is irrelevant (Accessed 1 Feb 2020).

• 

  Figure S2-6. Structured discussion at an argumentation tool (https://dev.tietokide.fi/?Q10. Accessed 1 Feb 2020)

• 

  Figure S2-6. Insight network about the assessment model for Helsinki energy decision 2015.

Appendix S3: Open policy ontology

Shared understanding aims at producing a description of different views, opinions, and facts related to a specific topic such as a decision process. The open policy ontology describes the information structures that are needed to document shared understanding of a complex decision situation. The purpose of the structure is to help people identify hidden premises, beliefs, and values and explicate possible discrepancies. This is expected to produce better understanding among participants.

The basic structure of a shared understanding is a network of items and relations between them. This network uses Resource description framework, which is an ontology standard used to describe many Internet contents. Items and relations (aka properties) are collectively called resources. Each item is typically of one of the types mentioned below. This information is documented using property instance of (e.g. Goherr assessment is instance of assessment).

Items are written descriptions of the actual things (people, tasks, publications, or phenomena), and on this page these descriptions rather than the actual things are discussed. Different item types have different levels of standardisation and internal structure. For example, knowledge crystals are web pages that always have headings question, answer and rationale, and the information is organised under those headings. Some other items describe e.g. statements that are free-text descriptions about how a particular thing is or should be (according to a participant), and yet some others are metadata about publications. A common feature is that all items contain information that is relevant for a decision.

In the open policy ontology, each item may have lengthy texts, graphs, analyses or even models inside them. However, the focus here is on how the items are related to each other. The actual content is often referred to as one key sentence only (description). Each item also has a unique identifier URI that is used for automatic handling of data.

The most important items are knowledge crystals and they are described here.

• Assessment describes a particular decision situation and focuses on estimating impacts of different options. Its purpose is to support the making of that decision. Unlike other knowledge crystals, assessments typically have a defined start and end dates and they are closed after the decision is made. They also have contextually and situationally defined goals`to be able to better serve the needs of the decision makers of the decision.
• Variable answers a particular factual or ethical question that is typically needed in one or more assessments. The answer of a variable is continually updated as new information arises, but its question remains constant in time. Variable is the basic building block of assessments. In R, variables are typically implemented using ovariable objects from OpasnetUtils package.
• Method tells how to systematically implement a particular information task. Method is the basic building block for describing the assessment work (not reality, like variables). In practice, methods are "how-to-do" descriptions about how information should be produced, collected, analysed, or synthesised in an assessment. Typically, methods contain a software code or another algorithm to actually perform the method easily. In R, methods are typically ovariables that require some context-specific upstream information about dependencies before it can be calculated.

There are also other important classes of items:

• Publication is any documentation that contains useful information related to a decision. Publications that are commonly used at Opasnet include encyclopedia article, lecture, nugget, and study. Other publications at Opasnet are typically uploaded as files.
  • Encyclopedia article is an object that describes a topic like in Wikipedia rather than answers a specific research question. They do not have a predefined attribute structure.
  • Lecture: Lecture contains a piece of information that is to be mediated to a defined audience and with a defined learning objective. It can also be description of a process during which the audience learns, instead of being a passive recipient of information.
  • Nugget is an object that is not editable by other people than a dedicated author (group) and is not expected to be updated once finalised. They do not have a predefined attribute structure.
  • Study describes a research study and its answers, i.e. observational or other data obtained in the study. The research questions are described as the question of the information object, and the study methods are described as the rationale of the object. Unlike in an article, introduction or discussion may be missing, and unlike in a variable, the answer and rationale of the study are more or less fixed after the work is done; this is because the interpretations of the results typically happen elsewhere, e.g. in variables for which a study contains useful information.
• Discussion is a hierarchically structured documentation of a discussion about a defined statement or statements.
• Stakeholder page is used to describe a person or group that is relevant for a decision or decision process; they may be an actor that has an active role in decision making or is a target of impacts. Contributors of Opasnet are described on their own user pages; other stakeholders may have their page on the main namespace.
• Process describes elements of a decision process.
• Action describes what, who and when should act to e.g. perform an assessment, make a decision, or implement policies.

Relations show different kinds of connections between items.

• Causal link tells that the subject may change the object (e.g. affects, increases, decreases, prevents).
• Participatory link describes a stakeholder's particular role related to the object (participates, negotiates, decides).
• Operational link tells that the subject has some kind of practical relation to the object (executes, offers, tells).
• Evaluative link tells that the subject shows preference or relevance about the object (has truthlikeness, value, popularity, finds important).
• Referential link tells that the object is used as a reference of a kind for the subject (makes relevant; associates to; has reference, has tag, has category).
• Argumentative link occurs between statements that defend or attack each other (attack, defend, comment).
• Property link connects an evaluative (acceptability, usability), a logical (opposite, inverse) or set theory (has subclass, has part) property to the subject.

Item types

This ontology is specifically about decision making, and therefore actions (and decisions to act) are handled explicitly. However, any natural, social, ethical or other phenomena may relate to a decision and therefore the vocabulary has to be very generic.

Table S3-1. Item types used in open policy ontology.

Class	English name	Finnish name	Description
	resource	resurssi	All items and relations are resources
resource	item	asia	Relevant pieces of information related policy making. Sometimes also refers to the real-life things that the information is about. Items are shown as nodes in insight networks.
resource	relation	relaatio	Information about how items are connected to each other. Relations are shown as edges in insight networks.
item	substance	ilmiö	Items about a substantive topic or phenomenon itself: What issues relate to a decision? What causal connections exist between issues? What scientific knowledge exist about the issues? What actions can be chosen? What are the impacts of these actions? What are the objectives and how can they be reached? What values and preferences exist?
item	stakeholder	sidosryhmä	Items about people or organisations who have a particular role in a policy process, either as actors or targets of impacts: Who participates in a policy process? Who should participate? Who has necessary skills for contributing? Who has the authority to decide? Who is affected by a decision?
item	process	prosessi	Items about doing or happening in relation with a topic, especially information about how a decision will be made): What will be decided? When will it be decided? How is the decision prepared? What political realities and restrictions exist?
item	action	toiminta	Items about organising decision support (impact assessment, decision making, implementation, and evaluation): What tasks are needed to collect and organise necessary information? How is information work organised? How and when are decisions implemented? Actions are also important afterwards to distribute merit and evaluate the process: Who did what? How did information evolve and by whom?
item	information object	tieto-olio	A specified structure containing information about substance, stakeholders, processes, methods, or actions.
information object	knowledge crystal	tietokide	information object with a standardised structure and contribution rules
knowledge crystal	assessment	arviointi	Describes a decision situation and typically provides relevant information to decision makers before the decision is made (or sometimes after the decision about its implementation or success). It is mostly about the knowledge work, i.e. tasks for decision support.
knowledge crystal	variable	muuttuja	Describes a real-world topic that is relevant for the decision situation. It is about the substance of the topic.
knowledge crystal	method	metodi	Describes how information should be managed or analysed so that it answers the policy-relevant questions asked. How to perform information work? What methods are available for a task? How to participate in a decision process? How to use statistical and other methods and tools? How to motivate participation? How to measure merit of contributions?
information object	discussion part	keskustelun osa	Information object that is used to organise discussions into a specified structure. The purpose of the structure is to help validation of statements and facilitate machine learning.
information object	discussion	keskustelu	Discussion, or structured argumentation, describes arguments about a particular statement and a synthesis about an acceptable statement. In a way, discussion is (a documentation of) a process of analysing the validity of a statement.
discussion	fact discussion	faktakeskustelu	Discussion that can be resolved based on scientific knowledge.
discussion	value discussion	arvokeskustelu	Discussion that can be resolved based on ethical knowledge.
discussion part	statement	väite	Proposition claiming that something is true or ethically good. A statement may be developed in a discussion by adding and organising related argumentation (according to pragma-dialectics), or by organising premises and inference rules (according to Perelman).
statement	value statement	arvoväite	Proposition claiming that something is ethically good, better than something else, prioritised over something, or how things should be.
statement	fact statement	faktaväite	Proposition claiming how things are or that something is true.
value statement	true value statement	tosi arvoväite	A statement that has not been successfully invalidated.
value statement	false value statement	epätosi arvoväite	A statement that has been successfully invalidated.
fact statement	true fact statement	tosi faktaväite
fact statement	false fact statement	epätosi faktaväite
statement	true statement	tosi väite
statement	false statement	epätosi väite
statement	opening statement	avausväite	A statement that is the basis for a structured discussion, a priori statement.
statement	closing statement	lopetusväite	A statement that is the resolution of a structured discussion, a posteriori statement. Closing statement becomes an opening statement when the discussion is opened again.
opening statement	fact opening statement	avausfaktaväite
closing statement	fact closing statement	lopetusfaktaväite
opening statement	value opening statement	avausarvoväite
closing statement	value closing stetement	lopetusarvoväite
discussion part	argument	argumentti	A statement that has also contains a relation to its target as an integral part. Due to this relation, arguments appear inside discussions and target directly or indirectly the opening statement.
discussion part	argumentation	väittely	Hierarchical list of arguments related to a particular statement.
information object	knowledge crystal part	tietokideosa	This is shown separately to illustrate that the objects are actually linked by has part rather than has subclass relation.
knowledge crystal part	question	kysymys	A research question asked in a knowledge crystal. The purpose of a knowledge crystal is to answer the question.
knowledge crystal part	answer	vastaus	An answer or set of answers to the question of a knowledge crystal, based on any relevant information and inference rules.
knowledge crystal part	rationale	perustelut	Any data, discussions, calculations or other information needed to convince a critical rational reader that the answer of a knowledge crystal is good.
knowledge crystal part	answer part	vastausosa	This is shown separately to illustrate that the objects are actually linked by has part rather than has subclass relation.
answer part	result	tulos	The actual, often numerical result to the question, conditional on relevant indices.
answer part	index	indeksi	A list of possible values for a descriptor. Typically used in describing the result of an ovariable.
answer part	conclusion	päätelmä	In an assessment, a textual interpretation of the result. Typically a conclusion is about what decision options should or should not be rejected and why based on the result.
knowledge crystal part	ovariable	ovariable	A practical implementation of a knowledge crystal in modelling code. Ovariable takes in relevant information about data and dependencies and calculates the result. Typically implemented in R using OpasnetUtils package and ovariable object type.
ovariable	key ovariable	avainovariable	An ovariable that is shown on an insight network even if some parts are hidden due to practical reasons.
information object	publication	julkaisu	Any published report, book, web page or similar permanent piece of information that can be unambiguously referenced.
publication	nugget	tiedomuru	An object that is not editable by other people than a dedicated author (group).
substance	topic	aihe	A description of an area of interest. It defines boundaries of a content rather than defines the content itself, which is done by statements. When the information structure is improved, a topic often develops into a question of a knowledge crystal, while a statement develops into an answer of a variable.
priority	objective	tavoite	A desired outcome of a decision. In shared understanding description, it is a topic (or variable) that has value statements attached to it.
substance	risk factor	riskitekijä
substance	indicator	indikaattori	Piece of information that describes a particular substantive item in a practical and often standard way.
indicator	risk indicator	riski-indikaattori	Indicator about (health) risk or outcome
information object	data	tietoaineisto
information object	graph	kuvaaja	Graphical representation of a piece of information. Typically is related to an information object with describes relation.
work	data work	tietotyö
work	data use	tiedon käyttö
substance	priority	prioriteetti
substance	expense	kustannus
substance	health impact	terveysvaikutus
stakeholder	decision maker	päättäjä
stakeholder	public officer	virkamies
stakeholder	assessor	arvioija
stakeholder	expert	asiantuntija
stakeholder	citizen	kansalainen
stakeholder	agent	toimija
action	task	toimenpide	action to be taken when the option has been selected
action	decision	päätös	action to be taken when the option is yet to be selected. Describes a particular event where a decision maker chooses among defined alternatives. This may also be a part of an assessment under heading Decisions and scenarios.
action	work	työ	continuous actions of the same kind and typically independent of the decision at hand. If the decision changes work routines, the action to make this change happen is called task.
work	prevention	ennaltaehkäisy	trying to prevent something
work	treatment	hoito	trying to fix something when something has already happened
work	support	tuki	work that aids in the completion of the selected option, in whatever way
method	open policy practice	avoin päätöksentekokäytäntö	framework for planning, making, and implementing decisions
method	open assessment	avoin arviointi	method answering this question: How can factual and value information be organised for supporting societal decision making when open participation is allowed?
method	analysis	analyysi
method	reporting	raportointi
method	measurement	mittaus
publication	study	tutkimus
publication	encyclopedia article	ensyklopedia-artikkeli	An object that describes a topic rather than answers a specific research question.
publication	lecture	luento	Contains a piece of information that is to be mediated to a defined audience and with a defined learning objective.
method	procedure	toimintamalli
method	principle	periaate	a short generic guidance for information work to ensure that the work is done properly. They especially apply to the execution phase.
principle	intentionality	tavoitteellisuus	See Table 3 for explanations.
principle	causality	syysuhteiden kuvaus
principle	criticism	kritiikki
principle	permanent resource locations	kohteellisuus
principle	openness	avoimuus
principle	reuse	uusiokäyttö
principle	use of knowledge crystals	tietokiteiden käyttö
principle	grouping	ryhmäytyminen	Facilitation methods are used to promote the participants' feeling of being an important member of a group that has a meaningful purpose.
principle	respect	arvostus	Contributions are systematically documented and their merit evaluated so that each participant receives the respect they deserve based on their contributions.
objective	expense objective	kustannustavoite
process	step	jakso	one of sequential time intervals when a particular kind of work is done in decision support. In the next step, the nature of the work changes.
step	impact assessment	vaikutusarviointi	the first step in a decision process. Helps in collecting necessary information for making a decision.
step	decision making	päätöksenteko	the second step in a decision process. When the decision maker actually chooses between options.
step	implementation	toimeenpano	the third step in a decision process. When the chosen option is put in action.
step	evaluation	evaluointi	the fourth step in a decision process. When the outcomes of the implementation are evaluated.
process	phase	vaihe	one part of a decision work process where focus is on particular issues or methods. Typically phases overlap temporally.
phase	shared understanding	jaettu ymmärrys	documenting of all relevant views, facts, values, and opinions about a decision situation in such a way that agreements and disagreements can be understood
phase	execution	toteutus	production of necessary information for a decision at hand
phase	evaluation and management	seuranta ja ohjaus	ensuring that all work related to a decision will be, is, and has been done properly
phase	co-creation	yhteiskehittäminen	helping people to participate, contribute, and become motivated about the decision work

Relation types

Relations are edges between items (or nodes). A relation I is said to be an inverse of relation R, iff, for all items subject and object, claim "subject R object" is always equal to claim "object I subject".

Table S3-2. Relation types used in open policy ontology.

Class	English name	Finnish name	English inverse	Finnish inverse	Description
relation	participatory link	osallisuuslinkki			The subject is a stakeholder that has a particular role related to an object
relation	operational link	toimintolinkki			The subject has some kind of practical relation to the object (a fairly wide class)
relation	evaluative link	arvostuslinkki			The subject shows preference or relevance about the object
relation	referential link	viitelinkki			The subject is used as a reference of a kind for the object
relation	argumentative link	argumentaatiolinkki			The subject is used as an argument to criticise the object.
relation	causal link	syylinkki			The subject has causal effect on the object (or vice versa in the case of an inverse relation)
relation	property link	ominaisuuslinkki			The object describes a defined property of the subject.
causal link	negative causal link	negatiivinen syylinkki			The subject reduces or diminishes the object.
causal link	positive causal link	positiivinen syylinkki			The subject increases or enhances the object.
negative causal link	decreases	vähentää	is decreased by	vähentyy
positive causal link	increases	lisää	is increased by	lisääntyy
negative causal link	worsens	huonontaa	is worsened by	huonontuu
positive causal link	improves	parantaa	is improved by	parantuu
negative causal link	prevents	estää	is prevented by	estyy
positive causal link	enhances	edistää	is enhanced by	edistyy
negative causal link	impairs	heikentää	is impaired by	heikentyy
positive causal link	sustains	ylläpitää	is sustained by	ylläpitäytyy
causal link	affects	vaikuttaa	is affected by	vaikuttuu
causal link	indirectly affects	vaikuttaa epäsuorasti	indirectly affected by	vaikuttuu epäsuorasti
causal link	cause of	syy	caused by	johtuu	Wikidata property P1542
causal link	immediate cause of	välitön syy	immediately caused by	johtuu välittömästi	Wikidata property P1536
causal link	contributing factor of	vaikuttava tekijä			Wikidata property P1537
participatory link	performs	toteuttaa	performer	toteuttajana	who does a task?
participatory link	decides	päättää	decider	päätäjänä
participatory link	asks	kysyy	asker	kysyjänä
participatory link	participates	osallistuu	participant	osallistujana
participatory link	accepts	hyväksyy	accepted by	hyväksyjänä
participatory link	develops	kehittää	developed by	kehittäjänä
participatory link	proposes	ehdottaa	proposed by	ehdottajana
participatory link	answers	vastaa	answered by	vastaajana
participatory link	responsible for	vastuussa	responsibility of	vastuullisena
participatory link	negotiates	neuvottelee	negotiated by	neuvottelijana
participatory link	recommends	suosittelee	recommended by	suosittelijana
participatory link	controls	kontrolloi	controlled by	kontrolloijana
participatory link	claims	väittää	claimed by	väittäjänä
participatory link	owns	omistaa	owned by	omistajana
participatory link	does	tekee	done by	tekijänä
participatory link	maintains	ylläpitää	maintained by	ylläpitäjänä
participatory link	oversees	valvoo	overseen by	valvojana
operational link	has option	omistaa vaihtoehdon	option for	vaihtoehtona
operational link	has index	omistaa indeksin	index for	indeksinä
operational link	tells	kertoo	told by	kertojana
operational link	describes	kuvaa	described by	kuvaajana
operational link	maps	kartoittaa	mapped by	kartjoittajana
operational link	contains data	sisältää dataa	data contained in	data sisältyy
operational link	data for	on datana	gets data from	saa datansa
operational link	uses	käyttää	is used by	on käytettävänä	an input (object) for a process (subject)
operational link	produces	tuottaa	is produced by	tuottajana	Object is an output of a process produced by a stakeholder (subject)
operational link	provides	varustaa	is provided by	varustajana
operational link	about	aiheesta			a task is about a topic. This overlaps with has topic; merge them?
property link	logical link	looginen linkki			Relations based on logic
property link	set theory link	joukko-oppilinkki			Relations based on set theory
set theory link	part of	osana	has part	sisältää osan	is a part of a bigger entity, e.g. Venus is part of Solar System. Wikidata property P361 (part of) & P527 (has part). Previously there were relations about a decision: substance of, decision process of, stakeholder of, method of, task of, irrelevant to. But these are depreciated and replaced by has part, because the class of the object makes specific relations redundant.
set theory link	context for	kontekstina	has context	omistaa kontekstin
set theory link	has subclass	omistaa alajoukon	subclass of	alajoukkona	Wikidata property P279
set theory link	has instance	omistaa instanssin	instance of	instanssina	Object belongs to a set defined by the subject and inherits the properties of the set. Sysnonym for has item, which is depreciated. Wikidata property P31
logical link	opposite	vastakohta			subject is opposite of object, e.g. black is opposite of white. Wikidata property P461; it is its own inverse
logical link	inverse	toisinpäin			a sentence is equal to another sentence where subject and object switch places and has the inverse relation. This is typically needed in preprocessing of insight networks, and it rarely is explicitly shown of graphs. Wikidata property P1696; it is its own inverse
logical link	if - then	jos - niin	if not - then not	jos ei - niin ei	If subject is true, then object is true. Also the negation is possible: if - then not. This links to logical operators and, or, not, equal, exists, for all; but it is not clear how they should be used in an insight network.
operational link	prepares	valmistelee	prepared by	valmistelijana
operational link	pays	kustantaa	paid by	kustantajana
operational link	rationale for	perustelee	has rationale	perusteltuu
operational link	offers	tarjoaa	offered by	tarjoajana
operational link	executes	suorittaa	executed by	suorittajana
operational link	irrelevant to	epärelevantti asiassa			If there is no identified relation (or chain of relations) between a subject and an object, it implies that the subject is irrelevant to the object. However, sometimes people may (falsely) think that it is relevant, and this relation is used to explicate the irrelevance.
evaluative link	finds important	kokee tärkeäksi	is found important	tärkeäksi kokijana
evaluative link	makes relevant	tekee relevantiksi	is made relevant	relevantiksi tekijänä	if the subject is valid in the given context, then the object is relevant. This typically goes between arguments, from a variable to value statement or from a value statement to a fact statement. This is a synonym of 'valid defend of type relevance'.
evaluative link	makes irrelevant	tekee epärelevantiksi	is made irrelevant	epärelevantiksi tekijänä	Opposite of 'makes relevant'. Synonym of 'valid attack of type relevance'.
evaluative link	makes redundant	tekee turhaksi	is made redundant	turhaksi tekijänä	Everything that is said in the object is already said in the subject. This depreciates the object because it brings no added value. However, it is kept for archival reasons and to demonstrate that the statement was heard.
evaluative link	has opinion	on mieltä			Subject (typically a stakeholder) supports the object (typically a value or fact statement). This is preferred over 'values' and 'finds important' because it is more generic without loss of meaning.
evaluative link	values	arvostaa	valued by	arvostajana	A stakeholder (subject) gives value or finds an object important. Object may be a topic or statement. Depreciated, use 'has opinion' instead.
evaluative link	has truthlikeness	on totuudellinen			A subjective probability that subject is true. Object is a numeric value between 0 and 1. Typically this has a qualifier 'according to X' where X is the person or archetype who has assigned the probability.
evaluative link	has preference	mieltymys	preference of	mieltymyksenä	Subject is better than object in a moral sense.
evaluative link	has popularity	on suosiossa			A measure based on likes given by users.
evaluative link	has objective	omaa tavoitteen	objective of	tavoitteena
argumentative link	agrees	samaa mieltä
argumentative link	disagrees	eri mieltä
argumentative link	comments	kommentoi	commented by	kommentoijana
argumentative link	defends	puolustaa	defended by	puolustajana
argumentative link	attacks	hyökkää	attacked by	hyökkääjänä
argumentative link	relevant argument	relevantti argumentti			Argument is relevant in its context.
argumentative link	irrelevant argument	epärelevantti argumentti			Argument is irrelevant in its context.
argumentative link	joke about	vitsi aiheesta	provokes joke	kirvoittaa vitsin	This relation is used to describe that the subject should not be taken as information, even though it may be relevant. Jokes are allowed because they may help in creating new ideas and perspectives to an issue.
referential link	topic of	aiheena	has topic	aiheesta	This is used when the object is a publication and the subject is a (broad) topic rather than a statement. In such situations, it is not meaningful to back up the subject with references. Useful in describing the contents of a publication, or identifying relevant literature for a topic.
referential link	discussed in	kerrotaan	discusses	kertoo
referential link	reference for	viitteenä	has reference	viite	Subject is a reference that backs up statements presented in the object. Used in the same way as references in scientific literature are used.
referential link	states	väittää	stated in	väitetään kohteessa	Describes the source of a statement; may also refer to a person.
referential link	tag for	täginä	has tag	omistaa tägin	Subject is a keyword, type, or class for object. Used in classifications.
referential link	category for	kategoriana	has category	kuuluu kategoriaan
referential link	associates with	liittyy			Subject is associated with object in some undefined way. This is a weak relation and does not affect the outcomes of inferences, but it may be useful to remind users that an association exists and it should be clarified more precisely. This is its own inverse.
referential link	answers question	vastaa kysymykseen	has answer	vastaus	Used between a statement (answer) and a topic (question). In knowledge crystals, the relation is embedded in the object structure.
irrelevant argument	irrelevant comment	epärelevantti kommentti			Inverses are not needed, because the relation is always tied with an argument (the subject).
irrelevant argument	irrelevant attack	epärelevantti hyökkäys
irrelevant argument	irrelevant defense	epärelevantti puolustus
relevant argument	relevant comment	relevantti kommentti
relevant argument	relevant attack	relevantti hyökkäys
relevant argument	relevant defense	relevantti puolustus
property link	evaluative property	arviointiominaisuus			characteristic of a product or work that tells whether it is fit for its purpose. Especially used for assessments and assessment work.
evaluative property	property of decision support	päätöstuen ominaisuus			What makes an assessment or decision support process fit for its purpose?
evaluative property	setting of assessment	arvioinnin kattavuus			See Table 5.
setting of assessment	impacts	vaikutukset
setting of assessment	causes	syyt
setting of assessment	problem owner	asianomistaja
setting of assessment	target users	kohderyhmä
setting of assessment	interaction	vuorovaikutus
interaction	dimension of openness	avoimuuden ulottuvuus			See Table 6.
dimension of openness	scope of participation	osallistumisen avoimuus
dimension of openness	access to information	tiedon avoimuus
dimension of openness	timing of openness	osallistumisen ajoitus
dimension of openness	scope of contribution	osallistumisen kattavuus
dimension of openness	impact of contribution	osallistumisen vaikutus
interaction	category of interaction	vuorovaikutuksen luokka			See Table 2. How does assessment interact with the intended use of its results? Possible values: isolated (eristetty), informing (tiedottava), participatory (osallistava), joint (yhteistyöhakuinen), shared (jaettu).
property of decision support	quality of content	sisällön laatu			See Table 4.
quality of content	informativeness	tarkkuus
quality of content	calibration	harhattomuus
quality of content	coherence	sisäinen yhdenmukaisuus
property of decision support	applicability	sovellettavuus
applicability	relevance	merkityksellisyys
applicability	availability	saatavuus
applicability	usability	käytettävyys
applicability	acceptability	hyväksyttävyys
property of decision support	efficiency	tehokkuus
efficiency	intra-assessment efficiency	sisäinen tehokkuus
efficiency	inter-assessment efficiency	ulkoinen tehokkuus

Appendix S4: Workspace tools: OpasnetUtils package and Opasnet Base

Ovariable

Ovariable is an object class that is used in R to operationalise knowledge crystals. In essence, impact assessment models are built using ovariables as the main tool to organise, analyse, and synthesise data and causal relations between items. The purpose of ovariables is to offer a standardised, generalised, and modular solution to modelling. Standardised means that all ovariables have the same overall structure, and this makes it possible to develop generalised functions and processes to manipulate them. Modular structure of a model makes it possible to change pieces within the model without braking the overall structure of functionality. For example, it is possible to take an existing health impact model, replace the ovariable that estimates the exposure of the target population with a new one, and produce results that are otherwise comparable to the previous results but differ based on exposure.

What is the structure of an ovariable such that

• it complies with the requirements of variable and
• it is able to implement probabilistic descriptions of multidimensional variables and
• it is able to implement different scenarios?

An ovariable contains the current best answer in a machine-readable format (including uncertainties when relevant) to the question asked by the respective knowledge crystal. In addition, it contains the information needed to derive the current best answer. The respective knowledge crystal typically has an own page at Opasnet, and the code to produce the ovariable is located on that page under subheading Calculations.

It is useful to clarify terms here. Answer is the overall answer to the question asked (including an evaluated ovariable), and it is the reason for producing the knowledge crystal page in the first place. Answer is typically located near the top of the page to emphasise its importance. An answer may contain text, tables, or graphs on the web page. It typically also contains an R code for evaluating the respective ovariable. Output is the key part (technically a slot) of the answer within an ovariable and contains the details of what the reader wants to know about the answer. All other parts of the ovariable are needed to produce the output or understand its meaning. Finally, Result is the key column of the Output table (technically a data frame) and contains the actual numerical values for the answer.

Slots

The ovariable is a class S4 object defined by OpasnetUtils in R software system. An ovariable has the following separate slots that can be accessed using X@slot (where X is the name of the ovariable):

@name

• Name of <self> (the ovariable object) is useful since R's S4 classes doesn't support self reference. It is used to identify relevant data structures as well as to set up hooks for modifiers such as scenario adjustments.

@output

• The current best answer to the question asked.
• A single data frame (a 2D table type in R)
• Not defined until <self> is evaluated.
• Possible types of columns:
  • Result is the column that contains the actual values of the answer to the question of the respective knowledge crystal. There is always a result column, but its name may vary; it is of type <self>Result.
  • Indices are columns that define or restrict the Result in some way. For example, the Result can be given separately for males and females, and this is expressed by an index column Sex, which contains locations Male and Female. So, the Result contains (at least) one row for males and one for females. If there are several indices, the number of rows is typically the product of numbers of locations in each index. Consequently, the output may become very large with several indices.
  • Iter is a special kind of index used in Monte Carlo simulations. Iter contains the number of the iteration. In Monte Carlo, the model is typically run 1000 or 10000 times.
  • Unit contains the unit of the Result. It may be the same for all rows, but it may also vary from one row to another. Unit is not an index.
  • Other, non-index columns can exist. Typically, they are information that were used for some purpose during the evolution of the ovariable, but they may be unimportant in the current ovariable if they have been inherited from parent ovariables. Due to these other columns, the output may sometimes be rather wide.

@data

• A single data frame that defines <self> as such.
• data slot contains data about direct measurements or estimates of the output. Typically, when data is used, the output can be directly derived from the information given, with possibly some manipulations such as dropping out unnecessary rows or interpreting given ranges or textual expressions as probability distributions.
• Probability distributions are interpreted by OpasnetUtils/Interpret.

@marginal

• A logical vector that indicates full marginal indices (and not parts of joint distributions, result columns, or units or other row-specific descriptions) of output.

@formula

• A function that defines <self> using objects from dependencies as inputs.
• Returns either a data frame or an ovariable, which is then used as the output of the ovariable.
• Formula and dependencies slots are always used together. They estimate the answer indirectly in cases when there is knowledge about how this variable depends on the results of other variables (called parents). The slot dependencies is a table of parent variables and their identifiers, and formula is a function that takes the outputs of those parents, applies the defined code to them, and in this way produces the output for this variable.

@dependencies

• A data frame that contains names and tokens or identifiers for model runs of variables required for <self> evaluation (list of causal parents). The following columns may be used:
  • Name: name of an ovariable or a constant found in the global environment (.GlobalEnv).
  • Key: the run key (typically a 16-character alphanumeric string) of a model run that is stored to Opasnet server. Key to be used in objects.get() function to fetch the dependent object.
  • Ident: Page identifier and rcode name to be used in objects.latest() function where the newest run contains the dependent object. Syntax: "Op_en6007/answer".
  • Also other columns are allowed (e.g. Description), and they may contain additional information about parents.
• Dependencies is a way of enabling references in ovariables by using function OpasnetUtils/ComputeDependencies. It creates variables in .GlobalEnv environment so that they are available to expressions in formula.
• Dependent ovariables are fetched and evaluated (only once by default) upon <self> evaluation.

@ddata

• A string containing an Opasnet identifier e.g. "Op_en1000". May also contain a subset specification e.g. "Op_en1000/dataset".
• This identifier is used to download data from the Opasnet database for the data slot (by default, only if empty) upon <self> evaluation.
• By default, the data defined by ddata is downloaded when an ovariable is created. However, it is also possible to create and save an ovariable in such a way that the data is downloaded only when the ovariable is evaluated.

@meta

• A list of descriptive information of the object. Typical information include date created, username of the creator, page identifier for the Opasnet page with the ovariable code, and identifier of the model run where the object was created.
• Other meta information can be added manually.

OpasnetUtils and operations with ovariables

OpasnetUtils is an R package found in CRAN repository (cran.r-project.org). It contains tools for open assessment and modelling at Opasnet, especially for utilising ovariables as modelled representations of knowledge crystals. Typically, ovariables are defined at Opasnet pages, and their data and evaluated output are stored to Opasnet server. There are also special user interface tools to enable user inputs before an R code is run on an Opasnet page; for further instructions, see http://en.opasnet.org/w/R-tools. However, ovariables can be used independently for building modular assessment models without any connection to Opasnet.

The example code shows some of the most important functionalities. Each operation is followed by an explanatory comment after # character.

install.packages("OpasnetUtils") # Install the package OpasnetUtils. This is done only once per computer.

library(OpasnetUtils) # Open the package. This is done once per R session.

objects.latest("Op_en4004", code_name="conc_mehg") # Fetch ovariables stored by code conc_mehg at Opasnet page Mercury concentrations in fish in Finland (with identifier 4004)

conc_mehg <- EvalOutput(conc_mehg) # Evaluate the output of ovariable conc_mehg (methyl mercury concentrations in fish) that was just fetched.

dat <- opbase.data("Op_en4004", subset="Kerty database") # Download data from Kerty database on the same page and put that to data.frame dat

a <- Ovariable("a", data=data.frame(Fish=c("Herring","Salmon"), Result=c(1,3))) # Define ovariable for scaling salmon results with factor 3.

mehg_scaled <- conc_mehg * a # Multiply methyl mercury concentrations by the scaling factor.

An ovariable is well defined when there is enough data, code or links to evaluate the output. Ovariables often have upstream dependencies whose output affect the output of the ovariable at hand. Therefore, ovariables are usually stored in a well defined but unevaluated format (i.e. without output). This makes it possible to use the same ovariable in different contexts, and the output varies depending on the upstream dependencies. On the other hand, it is possible to store all evaluated ovariables of a whole assessment model. This makes it possible to archive all details of a certain model version for future scrutiny.

Ovariables have an efficient index handling, which makes it possible to do arithmetic operations such as sums and products in a very simple way with ovariables. The basic idea is that if the outputs of two ovariables have two columns by the same name, they are automatically merged (or joined, using the SQL vocabulary) so that rows are merged iff they have the same location values in those two columns. The same principle applies to all pairs of columns by the same name. After the merge, the arithmetic operation is performed, row by row, to the Result columns of each ovariable. This results in an intuitive handling of outputs using a short and straightforward code.

Recursion is another important property of ovariables. When an ovariable is evaluated, a code checks whether it has upstream dependencies. If it does, those ovariables are fetched and evaluated first, and recursively the dependencies of those ovariables are fetched also, until all dependencies have been evaluated. Case-specific adjustments can be done to this recursion by fetching some upstream ovariables before the first ovariable is evaluated; if an upstream ovariable exists already in the global environment, the existing object is used and the respective stored object is not fetched (dependencies are only fetched if they do not already exist; this is to avoid unnecessary computation).

Decisions and other upstream commands

The general idea of ovariables is such that their code should not be modified to match a specific model but rather define the knowledge crystal in question as extensively as possible under it's scope. In other words, it should answer its question in a reusable way so that the question and answer would be useful in many different situations. (Of course, this should be kept in mind already when the question is defined.) To match the scope of specific models, ovariables can be modified without changing the ovariable code by supplying commands upstream. A typical decision command is to make a new decision index with two scenarios, "business as usual" and "policy" and use the original ovariable result for business as usual and adjust the result for the policy e.g. by adding or multiplying it by a constant reflecting the impact of the policy on the ovariable. Such adjustments can be done on the assessment level without a need to change the ovariable definition in any way.

Evaluating a latent ovariable triggers first the evaluation of its unevaluated parent ovariables (listed in dependencies) since their results are needed to evaluate the child. This chain of evaluation calls forms a recursion tree in which each upstream variable is evaluated exactly once (cyclical dependencies are not allowed). Decision commands about upstream variables are checked and applied upon their evaluation and then propagated downstream to the first variable being evaluated. For example, decisions in decision analysis can be supplied this way:

1. pick an endpoint ovariable
2. make decision variables for any upstream ovariables (this means that you create new scenarios with particular deviations from the actual or business-as-usual answer of that ovariable)
3. evaluate endpoint ovariable
4. optimize between options defined in decisions.

Other commands include: collapse of marginal columns by sums, means or sampling to reduce data size; and passing input from model level without redefining a whole ovariable.

Opasnet Base

Opasnet Base is a storage database for all kinds of data needed in open assessments. It may contain parameter values for models, which are typically shown as small tables on knowledge crystal pages, from which they are automatically stored to the database. It may also contain large dataset such as research datasets or population datasets of thousands or even millions of rows, and they are uploaded to the database using an importer interface. Each table has its own structure and may or may not share column names with other tables; however, if a table is directly used as data slot for an ovariable, it must have a Result column.

Technically, Opasnet Base is a noSQL database using MongoDB software. Metadata of the tables is stored in a MySQL database. This structure offers the speed, searchability, and structural flexibility that a large amount of non-standard data requires. The database also offers version control, as old versions of a data table are kept in the database when new data is uploaded.

The database also contains data about model runs that have been performed at Opasnet, if objects were stored during that model run. This makes it possible to fetch objects produced by a particular code on a particular knowledge crystal page. Typically the newest version is fetched, but information about the old versions are kept as well. The objects stored are not located in MongoDB but on server files that can be accessed with a key. It is also possible to save objects in a non-public way so that the key is not stored in the database and is only given to the person who ran the code. Due to disc storage reasons, Opasnet does not guarantee that stored objects will be kept permanently; therefore, it is a good practice to store final assessment runs with all objects to another location for permanent archival.

There are several ways to access database content.

• If the data is on an Opasnet page, simply go to that page, e.g. http://en.opasnet.org/w/Mercury_concentrations_in_fish_in_Finland#Data
• Use a link to the Opasnet Base interface, e.g. http://en.opasnet.org/w/Special:Opasnet_Base?id=op_en4004.mercury_in_baltic_herring
• Use a function in R: dat <- opbase.data("Op_en4004", subset="Mercury in Baltic herring")
• Use a function in R for stored objects: objects.latest("Op_en4004", code_name="conc_mehg")

For further instructions, see http://en.opasnet.org/w/Opasnet_Base_UI for user interface and http://en.opasnet.org/w/Table2Base for the wiki interface of small tables.

Appendix S5: Tools to help in shared understanding

There are lots of software and platforms to support decision making. Some of them have been listed here. The focus is on open source software solutions when available. Many examples come from Finland, as we have practical experience about them. The list aims to cover different functionalities and show examples rather than give an exhaustive list of all possibilities; such lists may be found from Wikipedia, e.g. https://en.wikipedia.org/wiki/Comparison_of_project_management_software. All links were accessed 1 Feb 2020.

Table S5-1. Useful functionalities and software in open policy practice.

Item	Functionality or process phase	Tool or software
Decision process	Information-based decision support	There is no single tool covering the whole decision process. Development work is needed. An interesting pilot software is being developed by the city of Helsinki for comprehensively managing and evaluating their ambitious Climate Watch and its impacts.
	Initiative	Several websites for launching, editing, and signing citizen initiatives at municipality or national level: Kansalaisaloite (Citizen Initiative), Nuortenideat (Ideas of the Young), Kuntalaisaloite (Municipality Initiatives). Similar tools could be used also for initiatives launched by Members of Parliament or the Government.
Substance	Content management	Diary systems, file and content management systems. Lots of individual solutions, mostly proprietary. VAHVA project by the Finnish Government will provide knowledge and tools for content management.
	Research data and analyses	AVAA, IDA, Fairdata and other data management tools help in managing research data from an original study to archival. Avoin data (open data in Finland), platform for publishing open data. Findicator: indicators from all sectors of the society. Datahub for open data sharing. Tools for separate analysis tasks are numerous, e.g. QGIS for geographical data. Several research fields have their own research and article databases, such as ArXiv.org (articles about physics, mathematics and other fields). Several biological databases.
	Public discussion, argumentation, statements	Otakantaa, Facebook, Twitter, blogs, and other social media forums for discussion. Websites for fact checking: Factbar, Fullfact, Need to know project for fact checking. Agoravoting is an open voting system. Lausuntopalvelu collects statements from the public and organisations related to planned legislation and Government programs in Finland. Swarm AI for collective intelligence
	News	News feeds (open source) CommaFeed, Tiny Tiny RSS. Semantic, automated information searches, e.g. Leiki.
	Description and assessment of decision situations and relevant causal connections	Opasnet for performing Open assessments and impact assessments. Knowledge crystals as integral parts of models and assessments. Simantics System Dynamics in semantic models. Jupyter notebooks for collaborative model development. Wikidata, Wikipedia as storages of structured data and information.
	Laws and regulations	Semantic Finlex contains the whole Finnish legislation and e.g. the decisions by the Supreme Court in a semantic structure.
Methods	Preparation of documents, co-creation, real-time co-editing	Several co-editing tools, e.g. Hackpad, MS Office365, Google Docs, Etherpad, Dropbox Paper, MediaWiki and Git. These tools enable the opening of the planning and writing phase of a decision. E.g. the climate strategy of Helsinki was co-created online with Google Docs and Sheets in 2018.
	Development and spreading good practices	InnoVillage helps to develop practices faster, when everyone's guidance is available online and can be commented.
	Organising systems for information and discussions	Decentralised social networking protocol Activitypub Tools: Full Fact automated fact checking Compendium. Vocabularies and semantic tools: Resourse Description Framework (RDF), Finto (Finnish Thesaurus and Ontology Service), AIF-RDF Ontology using Conceptual Graphics User Interface COGUI. These act as a basis for organising, condensing and spreading knowledge.
	Information design, visualisations	Interactive and static visualisations from complex data. Shiny, Diagrammer, Gapminder, Lucify Plotly Cytoscape
Work	Work processes in decision making, research etc: follow-up, documentation	Ahjo decision repository and Openahjo interface document and retrieve decisions that have been done in the city of Helsinki. Git enables reporting of both research and decision processes. There are several new platforms for improving science, such as Open Science Framework for facilitating open collaboration in research. Omidyar Network is a philantropic investment firm supporting e.g. governance and citizen engagement.
	Co-creation, experiments, crowdsourcing	Kokeilun paikka promotes experiments when applicable information is needed but not available. Sociocracy 3.0 provides learning material and principles for open collaboration in organisations of any size.
	Project management	There are lots of project management software, mainly targeted for enterprise use but somewhat applicable in decision making or research. Some examples: OpenProject, Project Management Body of Knowledge, Comparison of project management software, Fingertip.
Stakeholders	Expert services	ResearchGate Solved and other expert networks.

See also

• From open assessment to shared understanding
• Shared information objects in policy support, a previous version of this manuscript
• heande:User talk:Jouni#Fingetrip notes
• 7 of the best and worst ways to argue
• Shared understanding
• Open policy ontology
• Open assessment
• Open policy practice
• Properties of good assessment
• op_fi:Yhtäköyttä-hankkeen loppuraportti
• op_fi:Valassafari.
• op_fi:Hiilineutraali Helsinki 2035
• Insight network
• Open policy practice
• Shared understanding
• Discussion
• Evaluating performance of environmental health assessments
• Benefit-risk assessment of Baltic herring and salmon intake
• op_fi:Jaetun ymmärryksen menetelmä
• op_fi:Yhtäköyttä-hankkeen loppuraportti
• en:Graph theory
• Helsinki climate program 2035
  • op_fi:Helsingin ohjelmalliset energiatehokkuus- ja ilmastotoimenpiteet ja -tavoitteet
  • op_fi:Helsingin strategiset energiatehokkuus- ja ilmastotavoitteet
  • OpenAhjo
  • Hel Dev people
• Daniel K. Walco, Jane L. Risen. The Empirical Case for Acquiescing to Intuition [2]
• [www.opasnet.org/testiwiki/index.php/Item:Q18 Testi-DRF:action]
• Study: Wikipedia as accurate as Britannica. The journal Nature says the open-access encyclopedia is about as accurate as the old standby.[3]
• Tutkijat pelkäävät jakaa dataansa, vaikka avoimuus voisi lievittää pelkoa [4]
  • Laine, H., (2017). Afraid of Scooping – Case Study on Researcher Strategies against Fear of Scooping in the Context of Open Science. Data Science Journal. 16, p.29. http://doi.org/10.5334/dsj-2017-029
• MILDRED. The research data infrastructure for open data in the Univesity of Helsinki [5]
• The Availability of Research Data Declines Rapidly with Article Age [6]
• This is Not a Boundary Object: Reflections on the Origin of a Concept. Susan Leigh Star [7]
• There's no such thing as 'sound science' [8]
• Metayliopisto [9]
• File:Use of risk assessment in the society.ppt
• The politics of evidence revisited [10]

Parts not used

• Political extremism is supported by an illusion of understanding
• en:Political polarization
• U.S. media polarization and the 2020 election: a nation divided [11].
• WIRED: Psychological microtargeting could actually save politics [12]
• PLOS blogs: Future of open series [13]

None of the websites and tools described in this article offer a complete environment for open topic-wise scientific information production and discussion that would also support decision making. Opasnet works well for online assessments, but it is not optimised for documenting policy discussions or scientific work in real time. Climate Watch was designed to implement open policy practice in a specific context of municipality climate action plans. There are plans to generalise the functionalities for wider use base. This could be achieved by merging the functionalities of e.g. Opasnet, Open Science Framework, open data repositories, and discussion forums. Even if different tasks would happen at separate websites, they could form an integrated system (by using e.g. standard interfaces and permanent resource locations) to be used by decision makers, experts, stakeholders, and machines. Resource description framework and ontologies could be helpful in organising such a complex system.

Boundary object is a concept for managing information work within a heterogeneous group of participants. As people come from different disciplines, they see things differently and use different words to describe things. Boundary objects are common words or concepts that are similar enough across disciplines so that they help understanding but allow specific interpretations within disciplines or by individuals. Several dioxin-related knowledge crystals were successfully used as boundary objects in BONUS GOHERR project (Table S1-1) to produce shared understanding among authorities, fishers, and researchers from public health, marine biology, and social sciences.

Shared understanding aims to bring different views together. This is something that is needed especially during this time of polarisation. The open assessments performed have identified more agreements even about heated topics that what seems to be case based on social media. The pneumococcus case is an example of this.

Presenting also controversial and unpopular ideas is a prerequisite for a complete shared understanding. Thus, a community producing shared understanding should cherish and respect such activity and promote contributions even if they are incompatible with scientific or another paradigms. This is challenging to both someone presenting such claims and someone else personally against the presented idea. It helps if all parties have faith in the process and its capability to produce fair conclusions. Therefore, the society should promote the acceptability of open decision processes, open participation, and diverse contributions. Such attitude prevails in the climate strategy of Helsinki, but it was present already five years earlier in Transport and communication strategy in digital Finland (Table S1-1).

Openness does not mean that any kind of organisation or individual is equally prone or capable of using assessment information. Such equity issues are considered as a separate question and are not dealt with in this generic examination.

Openness is crucial because a priori it is impossible to know who may have important factual information or value judgements about the topic.

Open platforms for deliberation of decisions are available (otakantaa.fi, kansalaisaloite.fi), and sharing of code is routinely done via large platforms (ubuntu.com, cran.r-project.org). Also generic online tools such as Google Drive (drive.google.com), Slack (slack.com), and others have familiarised people with online collaboration and idea that information is accessible from anywhere.

ArXiv.org is a famous example of preprint servers offering a place for publishing and discussing manuscripts before peer review. Such websites, as well as open access journals, have increased during recent years as the importance of availability of scientific information has been understood. Using open data storages (ida.fairdata.fi) for research results are often required by research funders.

Aumann's agreement theorem shows that rational Bayesians demonstrates that rational agents with common knowledge of each other's beliefs cannot agree to disagree, because they necessarily end up updating their posterior with that of the other. In this thinking, shared understanding can be seen as an intermediate phase where the disagreements have been identified but the posteriors have not yet been updated to reflect the data that is possessed by the other person.

Acquiescence, i.e. situations where people know that their choice is irrational but they choose it anyway

There is a new political movement (Liike Nyt https://liikenyt.fi/) in Finland that claims that their member of parliament will vote whatever a public online discussion concludes. This approach is potentially close to co-created policy recommendations based on shared understanding. However, they are - at least so far - not using novel information tools or concepts to synthesise public discussions. Instead, they use social media groups and online polls. VIIDEN TÄHDEN LIIKE?

Also, we hypothesise that only a few major paradigms will emerge, and those are ones whose applicability is wide and independent of the discipline. Scientific paradigm is expected to be one of them, and it will be interesting to see what else emerges. People commonly reason against some unintuitive rules of the scientific method (e.g. they try to prove a hypothesis right rather than wrong) but it is not clear whether this will cause a need to develop a paradigm for an alternative approach. It is even not clear whether people are willing to accept the idea that there could be different, competing rules for reasoning in a single assessment or decision process.

Indeed, only 7 % of people contributing to Wikipedia do it for professional reasons.

Omidyar Network is an organisation that gives grants to non-profit organisations and also invest in startups that promote e.g. governance and citizen engagement. As an example, they support tools to improve discussion online with annotations, an objective similar to with structured discussions.

Additional references to pragma-dialectics.

Some experts and politicians seem to see criticism as a threat that should be pre-emptively avoided by only publishing finalised products. In contrast, agile processes publish their draft products as soon as possible and use criticism as a source of useful and relevant information.

Open Science Framework is a project that aims to increase reproducibility in science by developing structured protocols for reproducing research studies, documenting study designs and results online, and producing open source software and preprint services to support this. The Framework maintains a web-workspace for documenting research as it unfolds rather than only afterwards in articles.

Our own experience is the same, and we have not seen hijacking, malevolent behaviour or low-quality junk contributions. However, some robots produce unrelated advertisement material at Opasnet pages, but that is easy to identify and remove, and it has not become a problem.

TO DO

• A specific link should be available to mean the object itself rather than its description. http://en.opasnet.org/entity/...?
• All terms and principles should be described at Opasnet at their own pages. Use italics to refer to these pages.
• Upload Tuomisto 1999 thesis to Julkari. And Paakkila 1999

Suggested editors: Daniel Angus, Özlem Uzuner, Sergio Villamayor Tomás, Frédéric Mertens