Wikidata talk:Strategy 2017/Cycle 2/Engaging in the Knowledge Ecosystem

What impact would we have on the world if we follow this theme?

Charles Matthews

The word "education" is mentioned just once on the Cycle 2 page, here. That is disappointing, certainly as one of the obstacles Wikimedia is facing is a lack of clear understanding of the difference between "reference material" and "instructional material". So my answer is: less impact than would be achieved with a less slanted view of the said "ecosystem". The implication that the folk involved are largely autodidacts and/or researchers sounds like systemic bias generated by Wikimedians. Charles Matthews (talk) 20:49, 14 May 2017 (UTC)

@Charles Matthews: I'm sorry, but it's difficult for me to get your point. What makes said ecosystem slanted? What would be a less slanted ecosystem? Why the issue you describe in the last sentence is a bias? SGrabarczuk (WMF) (talk) 22:10, 16 May 2017 (UTC)

@SGrabarczuk (WMF): Firstly, "systemic bias" is a standard Wikipedia term, defined more fully at en:w:Wikipedia:Systemic bias as "systemic bias created by the shared social and cultural characteristics of most editors". The word editors is significant here. We hear much, rightly, about male systemic bias. We hear little, which is why I think this is unfamiliar to you, about the shared views of editors on education. I believe, and I was once a university lecturer, Wikipedians tend to think in terms of how they learn themselves, which is why I mention en:w:autodidacticism. I think they are very fluent in Google search, which is why I also mention "researchers". I think that, therefore, they underestimate the need in education for teachers, and for material actually designed to instruct.

So that unpacks those comments: I'm saying that the WMF should do more to contest those views, by leading more explicitly on "education as it appears to people in general". Not everyone learns easily by reading long, scrolling web pages. Most folk don't.

The comments below by User:Snipre about databases are specifically wrong, because of the Comprende! tool recently developed by Magnus Manske, which uses WikiBase. I have a personal stake in this direction, because that blog link gives some details on my experiences with en:w:Moodle for WMUK, which I discussed also with WMF people in the past. Educational material, in my view, does not look like a wiki page, but can look like a rendered form of underlying data, which may include wiki page sections. (See the Wikisoba link in the blogpost, for an earlier form of this idea). In particular, reusable educational material may require data in a convenient form.

I was trying to keep my comments quite generic. I do think the WMF should be involved in edutech, and now Wikidata/WikiBase are in the Wikimedia stable, it is possible to see how. Charles Matthews (talk) 04:04, 17 May 2017 (UTC)

@Charles Matthews: I am not sure if you read my comment until the end or if you didn't understood (my explanation was short and my English not very good): your example precisely shows an example of Wikidata as support, by providing data, but not as actor: somebody has to create a quizz, has to select some questions and more important has to choose the correct answer. Here we reach the limitation of Wikidata: how can WD handle a question where several different answers are possible ? Take the simple example of 2 different statements having different values with both some references ?

Your experience shows one thing: you select the questions, you choose the corresponding good answers, you define the evaluation scale, not WD. WD is not able to select the critical part in the knowledge which is necessary to know, to define the good answer when several are available and finally to define what are the minimal requirements to pass the evaluation step.

If WD is going to play an active role in education, this will lead to the selection of a specific set of data (the ones which is useful for education), this will lead to the preference of an unique value each time several ones exist,... This is not the goal of WD. Snipre (talk) 22:00, 18 May 2017 (UTC)

Snipre - what is at issue here with Comprende! is a WikiBase site hosted on Wikimedia Tool Labs: though Wikidata could host such information, in a "question namespace" (I have argued that recently). Once there is a wiki-style site where people can work collaboratively on question content, many things are possible. So my vision for 2030 is: Wikimedia should have realised the importance of such data; question data should be freely mixed with videos, for example, to provide a style of "course management system" (this is not so hard, really); and besides courses there should be diagnostic tests (with conditional branching), not just tests where the goal is evaluation. Now that Wikidata has an array datatype, the underlying structures would not be hard to set up and re-use - one only needs to bitmap a directed graph to be able to store complex navigational structures. So, in fact, I reject the limitations you are discussing. Charles Matthews (talk) 05:10, 19 May 2017 (UTC)

Snipre

As database we can't play a significant role in education: education is related to explanation and WD doesn't provide any space for explanations. We provide facts and we can play an important role as data support in education but not as main actors. Snipre (talk) 14:33, 16 May 2017 (UTC)

@Snipre: what about the entire movement instead of Wikidata? SGrabarczuk (WMF) (talk) 22:12, 16 May 2017 (UTC)

SGrabarczuk (WMF) WP has an advantage over WD: different values can be presented with explanations for the difference, different points of view can be described. But again the problem with the education objective is to simply a subject in order to allow someone who didn't know anything to learn the basics, to select, to organize, to choose priorities. Finally we can say that teaching leads to selection because you can't teach complexity to someone who doesn't know the basics.

But WD can play a role in education by proposing data, information, be provinding an unique framework where differences can be put together, where contradictions can exist. But someone outside of WD has to select the data because selection is not the piliar of WD, the truth is not the priority, only facts matters. Snipre (talk) 22:00, 18 May 2017 (UTC)

Lymantria

I think Wikimedia now is mainly consisting of three "pillars of free knowledge". We have a growing and improving set of facts/data, Wikidata. Also an interesting encyclopedia in several languages, consisting of a lot of good texts, Wikipedia. Thirdly an immense collection of free media, Commons. And of course there are several interesting, but minor, branches of free knowledge. All three sets are huge, need quality improvement of cours, but will lead to a growing position of Wikimedia in the Knowledge Ecosystem. Those are great achievements and of course education benefits of the availability of it all. But, what Wikimedia does not have is a view on didactics, let alone that materials have been develloped upon such a view. In line with the comments by Charles Matthews, what I see is an environment that suits people who learn in an investigating way, finding their own path and not in need of much support. Typicly skills you will find with autodidacts. For a lot of people the gold mine of Wikimedia facts, data and images will remain closed, simply because it is not that easy to handle. We do not have truly educational material helping users to understand things better. Let alone that Wikimedia is helpful for people to, for instance, master a new language, or a mathematical technique, or improve technical or artistic skills. And without view on didactics, view on how to serve different persons of different ages and different intelligences, I don't think we will be able to create materials in a way that has deep impact on the Knowledge Ecosystem more widely, especially the educational part.

In addition to that perhaps the Wikimedia-sites are in a way too static. Our volunteers mainly create and improve static elements: data, texts, media. Interaction with users, for instance in person help, may be a very important element that is missing. An example of how that could be an addition is a project like Ask Dr. Math (in fact older than Wikipedia): people ask questions, volunteers answer them, good answers are published in a searchable library of part of the answers and even in books. Lymantria (talk) 16:22, 3 June 2017 (UTC)

How important is this theme relative to the other 4 themes? Why?

Charles Matthews

Addressing the issues crudely described as "knowledge monopolies" and "social mobility" is of key importance in creating a more meritocratic world. As I commented above, I do find the formulation flawed, and so I find the question hard to answer as posed. Charles Matthews (talk) 20:54, 14 May 2017 (UTC)

Snipre

4/5. We can't be an open community and be a reference at the same time. To be a reference for education, we have to provide a coherent knowledge and not a collection of facts without references or contradictions. As wikidata is not able to choose and to explain the choice when facing opposing statements, we can't provide any educational content where explanations are important. So we have to act as support for educational content but not as active actor. Snipre (talk) 14:23, 16 May 2017 (UTC)

Focus requires tradeoffs. If we increase our effort in this area in the next 15 years, is there anything we’re doing today that we would need to stop doing?

Charles Matthews

My feeling is this area is potentially more important to Wikimedia than the "Augmented Age", where for-profit companies are likely to cover much of the ground (excepting translation to languages outside the top ten). Charles Matthews (talk) 20:56, 14 May 2017 (UTC)

Snipre

We should limit the access for IPs and limit the bot operations at large scale without a tight control by the community or some authorities in side the community. To work as educational actor implies to become a reference and then to control in a more strict way the WD content and the modifications of WD content. We can't be an authority with a open database where every single data piece can be modified by anyone. Snipre (talk) 14:28, 16 May 2017 (UTC)

What else is important to add to this theme to make it stronger?

Charles Matthews

Software support for open and reusable educational and training material. Again some discrimination is required, given that "OER" is too non-specific to be helpful to Wikimedia. Teachers have always used mashups! Instructional design, diagnostics, data structures underlying the navigational needs of those: Wikimedia needs a strong presence in those to really make education open. Charles Matthews (talk) 21:01, 14 May 2017 (UTC)

Snipre

More control of content modifications and data import and strong rules for using data for recognized reference work. As educational actor we have to provide the best set of data and we are responsible of offering the correct information. Snipre (talk) 14:30, 16 May 2017 (UTC)

Who else will be working in this area and how might we partner with them?

Charles Matthews

There are large OER communities, but there is a serious social issue in the "siloisation" of tools for edutech. Charles Matthews (talk) 21:03, 14 May 2017 (UTC)