Contributing to Opasnet

From Testiwiki
Jump to: navigation, search


We are trying to develop Opasnet into a major website informing decision-making in the society. You can participate in several different ways to help us reach the goal. Join us! This pages gives hints and practical guidance about the most common things newcomers should know when they participate in a project.

Scope

How can people contribute to Opasnet? What are the most common things that a newcomer should know when participating in a project? Especially, what are things that are not obvious simply by looking at existing pages and mimicing what others have done?

Result

Methods to contribute
What you have What you want to get How to do it
Interesting content outside Opasnet To find a link from a relevant page in Opasnet Find a relevant page (or create one if there is none) and add a link to See also section. See Help:Editing
Important content outside Opasnet To find a description from a relevant page in Opasnet As above, but also write a description on the page
A file, important for an assessment, outside Opasnet To find the data from Opasnet Contact someone from THL and ask that the file is uploaded to Opasnet File. If the file can be published with CC-BY-SA copyright, it can be made openly available. Otherwise, it will be uploaded to Heande behind a password. Files can also be uploaded to Opasnet wiki but that is generally not recommended for files that will not be edited.
Data needed in an assessment. Make it reachable by the R-tools. Possibilities

Research question

All pages (except encyclopedia articles) have a specific research question they attempt to answer. It is important to stop for a while to think of a good research question, because everything else on the page simply tries to answer that question convincingly. Consider these things when developing a question:

  • The question is clear and unambiguous. For example, with the question "What are health effects of lead?" it is not clear whether the question is about a complete list of different endpoints caused by lead, the number of people with a disease caused by lead, or the risk of a disease after a certain lead exposure.
  • You can imagine a clear answer to the question. This is also called as the clairvoyant test.
  • Boundaries should be mentioned unless they are obvious. A wider interpretation is taken unless otherwise mentioned. The previous example about lead is apparently about humans so there is no need to mention it (if it is about some other species, that should be mentioned). However, if we are interested in children and not any humans, that should be mentioned. Also, geographical boundaries should be defined, e.g. "What is the number of children with lower intelligence quotient due to lead exposure in Finland?" Questions are about current situation and/or time trends unless otherwise specified.
  • Preferably, the answer is quantitative, i.e. it can be expressed by numbers. Another typical form of an answer is an exhaustive list of things fulfilling the conditions described in the question. On the other hand, many answers cannot be numbers or lists, like the one on this page.
  • Try to get the question approximately right from the beginning. You are allowed to change the question of a page, but typically one page is connected to several others, and these connections are easily broken, if the questions change. It is quite a task to fix the pages back into a coherent whole.


Basic structure: question, answer, rationale

The objective of each page (except encyclopedia articles) is nothing more and nothing less than to answer the question of the page. (The question is sometimes called scope, but that is essentially the same thing.)

It is therefore quite obvious that each page also has a part where the answer is found. This part comes typically after the question and is called either Answer or Result, whichever is more clear in the situation. The answer should be quite short and contain just enough explanation so that the answer can be understood without reading other parts. Preferably, a data table is used to summarise the answer.

Everything that is needed to convince a reader about the answer should not go under Answer but under a third sub-heading, namely Rationale (or Definition, as that name was previously used). Notice that Rationale may be the longest part of the page, but still it is only serving the Answer.

There are also other subheadings that are routinely used on each page. There may be a summary text about the main points of the page (possibly in a summary box) before the actual question. See also contains links to pages within Opasnet or elsewhere; they are related topics or somehow useful for a reader. Keywords are used to help search engines to find the page. References may contain manually added references, but we recommend the use of <ref> tags (see below). Related files may contain reports or other files that have been uploaded to the Opasnet File system and are relevant for the topic. The use of these subheadings is recommended on every page, even if there is nothing under them. It will guide new contributors to provide right kind of information.


Documenting all ideas

There is a major difference between the purposes of Opasnet and Wikipedia. Wikipedia is for documenting well-established information, while Opasnet is for both documenting information and developing new ideas needed for solving practical problems. This difference is also seen in scoping. In Wikipedia, all well-documented information are welcome, but discussions about ideas with little data, or original data itself, are out of scope of the website. In practice, a killer argument among Wikipedists is this: "There is no published information from reliable sources about this, so we will not write about it." In contrast, discussions among open assessors take a different path: "To solve this problem, we need information about this topic. So, however vague our information is, we collect anything there is and do our best."

Because of the different scope, there is also a need for different approaches. Because the vagueness of information is not a rejection criterion in Opasnet, there must be ways to deal with vague information. The most important thing is to document all ideas on a relevant page, together with any information that can be used for the evaluation of those ideas. The only important rejection criterion is relevance: Does this idea answer the question of this page? Or, does this idea help in answering the question? If not, the idea is not relevant on that page and it should be documented elsewhere.

In contrast, validity is not a rejection criteria: even if some idea is probably false, i.e. invalid, it still should be documented if someone may think that it is true. This is an important rule, and the rule is against what people are used to. Therefore, it must be discussed thoroughly. A reason for the rule is that we cannot know a priori, or beforehand, which ideas will turn out valid and which invalid. An important part of Opasnet is that we document the rationales of our answers. We need rules that we can apply before we make conclusions about which of the suggested answers we find valid and which not. The easiest way to do this is to treat all information in a similar way, even if we personally have an opinion about its validity. My personal opinion is not the same as the general agreement of a large group (although many of us tend to think that when a large group of rational people think about things carefully, they will make the same conclusions as I have already made).

There is another important reason why invalid ideas should be documented in Opasnet: Opasnet is not only for documenting correct answers for questions; it is also not only for documenting rationale for correct answers; it is also for documenting why false answers are false so that people would learn from mistakes and understand why they should abandon some misbeliefs. The more common a misbelief is, the more important it is to carefully document the rationale that concluded the idea to be invalid. This serves two purposes. First, more people find it easier to understand how things actually are. Second, more people find it easier to attack those parts of rationale that is not convincing to them, thus finding bugs in the thinking and making the rationale even stronger. Of course, bugs may be found from the prevailing answer as well, so that a today's answer may be rejected tomorrow. But even when this happens, it is important to see that even then our new answer will be based on a better shared understanding than the one we believe today. We as open assessors are allowed to defend any idea we think is valid, but even more eagerly we must defend other people's right to attack our ideas.

It is not important to document all possible false ideas, and indeed that would be impossible. If I make a mistake and then find it afterwards, there is no need to document that, just correct it. But if a mistake seems to be common, severe, or from a popular source, it is important and useful for many people to explicitly discuss it and document why it is invalid.

The previous discussion was about scientific information (how things are). But everything that was said about documenting applies to value judgements (how things should be) as well. It is important to document also extreme opinions and opinions that are against general values or against my personal values. The same reasoning applies to both facts and values. Only when values are documented, their popularity can be explicitly studied. Only then can they also be attacked e.g. by showing that these values are in contradiction with other, more profound values.

There is a common view that extreme opinions should not be allowed in public discussions, because that could increase their popularity. This Opasnet rule goes against that view. However, there are reasons to believe that the benefits of the rule are larger than the risks. Opasnet is not just one discussion forum, where discussions can be hijacked by a small active group who repeat the same extreme arguments over and over again. This is because a) in structured discussions, repeating is useless, b) all arguments must be relevant to the research question or they will be removed, and c) in a causal structure, all implications of an argument are easier to see, so that incoherent thought structures tend to fall.

Object type

When you create a new page, the first question is about page type (we also call it object type as pages are often called objects). Each object type is used for a specific purpose. Most pages are variables, but if you are uncertain about the type, you can make it an encyclopedia page. On the other hand, type can easily be changed, so don't worry too much about the choice. Different types are described in detail on the page Universal object.

The object type is defined on the top of a page by adding a template that contains the name of the object type in double braces, such as {{variable}}.

  • Assessment is a project that aims to answer some practical information needs of a particular decision-making situation. It typically has a deadline and a responsible assessor who organises the work to get it done.
  • Variables describe reality by answering questions typically like "How much is X?", "What is the list of things fulfilling X?", or "How good is X?". Variables are building blocks of assessments, but they can be used in several assessments. Therefore their development does not end even if one assessment ends.
  • Methods describe how to do things in assessments, in Opasnet, or somewhere else. For example, this page is a method.
  • Encyclopedia articles are general descriptions about a topic. They do not attempt to answer a specific research question. They can have a free page structure (unlike other pages that should have the same basic structure question-answer-rationale).
  • There are also other page types that are more rarely needed. Two are mentioned here. A study describes a research study and its results. A lecture contains a piece of information that is written specifically to a defined audience and with a defined learning objective.


Moderator and his/her tasks

Each page should have a moderator. If you create a new page, the default is that you will be the moderator of the page. Moderator's task is to keep eye on the page from time to time to see that it is not attacked by vandalism or some other disruptive editing. Moderating does not bring any legal requirements or responsibility to you. On the other hand, you have no power over other contributors of the page. For example, if you want something to removed but the other contributor disagrees, you cannot use your role as the moderator to force things; you must discuss things through just like everyone.

The moderator is defined within the object type template. For example:

{{method|moderator=Jouni}}

It is a good idea to watch all pages you moderate. Then, they will show on your watchlist and you can see all edits to all your pages at once. You can start watching a page and go to your watchlist by using the buttons on the top of each page.

If you permanently stop watching the pages you have been moderating, you should discuss with someone who could start moderating the pages.


Using text from other sources

Opasnet has the same copyright as Wikipedia, IEHIAS and some other info sources, namely Creative Commons Attribution/Share-Alike License. You can freely copy text from these sources as long as you properly cite where you got the text from. However, this is usually not the case: only rarely you can directly copy text or images, but you must use your own words to describe the information. So, be careful about copyrights, or your texts may be removed.


References and citing

Always remember to cite your info source, because many people want to see the original content before they believe what you say. Within Opasnet, you can and should simply use internal links, e.g. [[Main Page]]. However, internal links also work to Wikipedia this way: [[:en:Main Page]]. the ":en:" in front tells that the page is in English Wikipedia, not in Opasnet. Links to Wikipedias with other languages work as well: :fi: for Finnish, :de: for German etc. These are the same abbreviations that are used between Wikipedias.

References follow Vancouver style, which is approximately this format:

Scientific and other articles
Authorlastname Firstname, Author2lastname Firstname: Name of the publication. Journal (Year) Issue: Volume: Pagenumbers.
Reports and books
Authorlastname Firstname, Author2lastname Firstname: Name of the publication (page numbers if relevant). Publisher, City, Year.

You should put external links within <ref> tags, so that they automatically appear in the reference list. All links within a ref tag work normally. If one reference appears on a page several times, you should give it a name. Then, they are automatically merged in the reference list. A name is given this way: <ref name="reportname">Author A. A., Writer B. B. Report on an interesting topic. Report Publisher, London, 1999.</ref>

When the reference is once written out in full, you can use an abbreviated form with the subsequent citations on the page: <ref name="reportname"/>. Note that the slash / is needed to end the reference.


Dependencies between pages

Under the Rationale (or Definition) heading, there may be a sub-heading called Dependencies (which used to be called Causality). Links under this sub-heading go to pages that affect this page. To be precise, the Answer of this page is dependent on the Answer of the other page, and if the latter changes, the former will change also. In addition to the link, the Dependencies should explain, how the result actually changes. In many cases, a thorough explanation may require complex quantitative modelling which is not described here but on page Modelling in Opasnet. However, also a rough explanation is usually very helpful, e.g. ("If the Price of oil goes up, the answer of this page will go down.") Even a plain link is often very useful, as it points readers' attention to a new dependency, which might otherwise have been unnoticed.


Data tables as the answer in a nutshell

Data table is a table whose result goes automatically into a database, from where it can be downloaded for modelling and analysis. We hope that more and more Answers on pages will be written down as data tables so that they can be easily used by computers. Of course, this requires some learning from the humans who contribute to Opasnet.

A data table is a two-dimensional table which has two kinds of columns: explanations that explain or classify the result by e.g. country, age group, sex, type etc.; and observations that contain the actual answers to the research question. For more detail about the concept, see Opasnet base structure. For technical instructions to use the data table, see Table2Base. For guidance on how to use data tables in modelling, see Modelling in Opasnet. Note that one page may only contain a single data table, otherwise the answer is not unambiguous.

All pages under the Dependencies sub-heading potentially affect the structure of the data table and the explanation columns used. Therefore, it is important to check the structure of the Answer in these pages to make sure that they are coherent with your page. If they are not, rethink your page and/or start a discussion with those who are writing those pages you are dependent on.

Ignorance of motives and organisations

In rhetoric, argumentum ad hominem means a fallacy where an argument is claimed false based on who said it or from which organisation he comes from. All too often the focus of discussants is on who is talking and what are the presumed motives. Although this may be relevant sometimes, this kind of discussion is discouraged in Opasnet. The focus should always be on the substance and not on the discussants or the organisations they represent. The rationale is that if a large self-organised group of people bring facts on the table, the information available tends to be only little biased even if most participants wanted to emphasise facts beneficial to their own interests. If the participation is active enough, the weight of a single contribution about a piece of evidence is so small that there is no need to consider who contributed and why.

A proper ignorance of motives and organisations on the level of a single discussion is important. However, it has also wider implications that are fundamental. The current political system is based on organised groups of people that form political parties, trade unions, lobbying groups, think tanks and so on. The purpose of these groups is to promote a certain agenda, which is backed up in public and private discussions with a selection of arguments that is intentionally biased towards supporting the agenda. In theory, what was just said about large participation ensuring only little bias is applicable also here. But when we look at the true participation in the current system things do not look so good after all.

In the media, there is a lot of emphasis on individual political leaders and their opinions. This is rightly so, because the opinions and views of a party are heavily dependent on those of its leader. But this unfortunately implies that the opinions of a party are not heavily dependent on those of its members. Indeed, the role of a member is mostly to support the ideas of the leader (or, move to another party if that leader is more in line with the member's thinking). The voice of a single party member is extremely small in a political discussion. The only way to increase the voice is to actively work for years to gain political merit either by getting important positions in the party, or collect a lot of votes in an election (after which it is possible to demand for important positions in the party).

This is clearly an unoptimal system. The system should increase the voice of those who have something wise to say and decrease the voice of those who don't, irrespective of their position in an organisation. The current system gives significant voices to only ten party leaders, and the relative importance of their ideas depends on the popularity of the party, which, at best, correlates with how wise their ideas are. The system ignores those ten thousand people who would have something to say and would be interested in saying it. Most of the richness of ideas, opinions, and nuances are thus forgotten.

Parties and trade unions were crucial in their time because they made it possible for citizens or employees to have any voice at all. Now, there are technical facilities like Opasnet to collect and synthesise information and values to evaluate policies. This is not to say that parties are useless and should be replaced by discussion web workspaces. The point is that these web workspaces are better equipped than parties to develop shared understanding. Therefore, the role and capabilities of parties in this respect are threatened unless they are able to adopt new ways of working. Actually, their role in developing a civil society may even be negative if they maintain the false impression that the only way to affect a policy is to join a party or vote for it and deny the role of directly contributing to an open discussion about the policy. Organisations still have their roles in engaging people and organising election campaigns or pressure groups. But what will be roles of existing organisations on the one hand and new web-based methods on the other hand in having societal discussions and making conclusions, that will only be learned when large enough groups realise that new methods exist and start to use them.

In conclusion, the personality, formal training, or political connections of a participant should typically be ignored in Opasnet. There are only rare cases where this kind of information may actually affect the conclusion of a discussion without resulting in a fallacy.

Rationale

Based on experience on things that easily cause trouble.

See also

Advice for using Opasnet
Basics of Opasnet: What is Opasnet · Welcome to Opasnet · Opasnet policies · Open assessment · What is improved by Opasnet and open assessment? · FAQ
How to participate?: Contributing to Opasnet · Discussions in Opasnet · Watching pages · Open assessment method
How to edit pages?: Basic editing · More advanced editing · Quick reference for wiki editing · Wikipedia cheatsheet · Templates
Help for more advanced participation: Copyright · Archiving pages · Copying from Wikipedia · ImageMap · SQL-queries · Analytica conventions · Developing variables · Extended causal diagram · GIS tool · Risk assessment · M-files · Stakeholders · Heande · Todo · Text from PDFs and pictures · Word2MediaWiki · Glossary terms · Formulae

Keywords

Opasnet, open assessment, collaborative work, mass collaboration, web-workspace

References


Related files

<mfanonymousfilelist></mfanonymousfilelist>