Community Call 12/7: Play date with GraphRyder!

Community Call 12/7: Play date with GraphRyder! 

A one hour online demo and interactive event where participants learn to read the opencare community network graph!

  • What do participants mean when they talk about care?
  • Which aspect of care are you are interested in? See who (else) is talking about it and in which context 
  • Discover co-occurences of care concepts: i.e. which care concepts are associated with migrant care, or care regulations in our conversation. 

GraphRyder is opensource, web based, and meant to be a resource for network scientists, researchers, community managers and citizens at large - graphryder.opencare.cc

Our goal for this first session is to help people get familiar with it. Anyone interested can come, learn how to navigate graphically and conceptually our community database, and ask questions. We hope to get useful feedback on the tool to help us adjust the views that need improvements.

Who is leading?

Guy Melancon is a professor and researcher at University of Bordeaux and the uni computer science research lab (LaBRI). A mathematician and expert in network visualizations, with a knack for interdisciplinary collaboration and learning events (Masters of Networks), Guy is leading the team building GraphRyder as a compelling interface for open care community data.

I get involved as much as I can in tech and intellectual transfer action towards the industry. I dream every citizen would be able to handle, mine and visualize open data to defend their cause.

Jason Vallet. works with melancon, bpinaud at the University of Bordeaux to provide expertise in visualization for the opencare conversations.

Amelia Hassoun. A digital and medical anthropologist, and PhD student at Oxford University, Amelia has been coding over. I’m secretly hoping she’ll become Edgeryders in-house ethnographer.

Federico Monaco. Teaching at Universita degli Studi di Parma, Federico is currently opencare’s ethnographer in residence at WeMake makerspace in Milano.

When and where

This Wednesday at 18:00 CEST in the google hangout - click here for direct access (no login required).

You don’t need to prepare ahead of the call, just bring an open mind.

Interested to attend? Press the blue button “Attend” or leave a comment below so we can do a proper head count, thanks!

Date: 2017-07-12 19:00:00 - 2017-07-12 19:00:00, Europe/Brussels Time.

Notes from the call

We made some notes during the call, here they are.

Intros:

Guy - researcher working on the software & network science part of Open Care.

Winnie - curating open and citizen science theme. Involved in biohackerspace in Ghent, Open Insulin collaboration.

Alberto Rey - working on citizen science for the festival together with Winnie, and fly fishing.

Michael - new to Edgeryders, figuring it out

Gehan - curating Architectures of Love theme: researching conditions in which opencare happens, policy. Works at Galgael trust. Conditions evolve not entirely by design.

Amelia - Edgeryders’ ethnographer. I use Graph Ryder to help with the data analysis! I’m an anthropologist and my home is London

Federico - sociologist at university of Parma. Doing anthropological research in makerspace WeMake.

Rachel - invited by Winnie, met at Biofabbing convergence. Part of Hackuarium, into citizen science and choices we can make that influence our health (risks). In on citizen science track for the festival.

Some more people joined during the session: Alberto, Bernard

This session is not technical, rather a walkthrough for the GraphRyder program.

When signing up on Edgeryders, you consented that your ‘data’ (posts, comments, …) would be used for research.

Posts have tags. You can navigate the tags and then go back to the comments.

Link tags that co-occur in conversations. Look at links that appear more frequently. Slider helps to focus on strong links. This helps to analyse why these might be appearing more frequently.

This can be accessed following the url: http://graphryder.opencare.cc and a detailed view: http://164.132.58.138:9000/index.html#/dashboard/globalView

Aware of the issues in terms of navigation around the various websites. Suggested that a simple webpage is created that links to the various sites and gives an outline.

Key word: Autonomy. Program grabs all tags that autonomy is linked with. It’s linked with a lot of key words.

High score tags that link with ‘autonomy’. You can access into the original content and posts.

The tags are made by an ethnographer. We access the whole conversation at once.

How autonomy connects with strongest connection. An association made by the conversation itself. You can read the comments and posts so you can figure out why people think these two things are connected. When someone comments on a post, chances are they will make similar comments and be tagged with the same label/s. However they will often bring something new into the conversation that will create association with new tags.

The innovation is related to the thread and it denotes new tags/elements brought to the conversation… indication of how much the thread/discussion added to the conversation. If the number is high it means many new concepts were brought into the conversation. Amelia: we haven’t properly tested the innovation theory - so it’s not to say that higher “innovation” = more interesting, it was an idea we had. So don’t assume posts with lower innovation tag aren’t interesting :slight_smile:

We’re looking for views of community members as to how to use it. Post on the platform and ping …. Guy (@melancon) or Jason (@Jason_Vallet ).

Amelia: now we have the link we can experiment on our own and then come back and discuss further on the next call

Detangler view - people on the left and tags on the right. Use the example such as Alberto’s post some time ago that follows how a policy maker might use Graphryder as a tool.

Amelia and Jason will do a session at OpenVillage. This was the first in a series of 3 demos on navigating the GraphRyder software. They’ll shape the session considering the input of people in the following calls.

Some questions for further discussion:

  • Do you have feedback for us?
  • Is this tool useful to make sense of the community and conversations you're part of?
  • What would you change about it?
  • Do you see other applications for GraphRyder? New databases, communities etc?

Personal reflections

I had used the software before during a Masters of Networks event, so I was already kind of familiar with it. As an introduction I found the call a bit too technical. It went into the software and its features more than the ‘why?’ and the usefulness for non-experts. The latter would be more relevant to go into for an audience like we had at the call. The software is also clearly built for research at this point. That being said, I found the technical explanation to be very clear.

The Google Hangouts was full at some point (10 people) and some people were left out, so it’s worth considering a different platform. Many people seem to be interested.

A wiki page to explain

I felt the same, @WinniePoncelet . I have put up a quick page on the GraphRyder repo’s wiki to explain:

Is that concrete enough? It seems clear to me, but then I spend a lot of time with this stuff.

Also ping

@Amelia | @Noemi |  @Federico_Monaco | @melancon | @Jason_Vallet

As I said in Bordeaux, high time to take the documentation of software & methodology very seriously. The wiki page is an attempt to formalise the (procedural) knowledge sloshing about in the community call.

  • Is it clear enough?
  • Is the wiki the right place for it?

Playbook-readiness

About my question 2 above, ping @Costantino . I think you said the Playbook is to be written in markdown, correct? This means that wiki pages can be lifted and copied into the Playbook, so me writing in the GitHub wiki is not so bad.

But even better: do you have a Playbook chapter template that we can use for the documentation?

Hm

Thanks for putting this together @Alberto , and Winnie for note taking.

While it is clear for you, and me, I find if this is meant as a user support text  it should be more hands on and placed in a pop up in the web application. Some language might be toned down to less technical:

Example

"Which codes connect to which other codes. For example, the network could be disconnected into “islands” of codes, with no code in each of the islands ever occurring with any code in any of the other islands. This would be a strong indication that the informants have not associated concepts with each other. think there are entirely separate, mutually independent sides to the problem at hand. In a less extreme variant of the same scenario, the network could be highly modular.

Not exactly

What I had in mind was not inline help. That needs to be super-short, agree. I have in mind a kind of user manual.

Who is it intended for?

Who are you aiming to reach with the wiki?

If it’s anthropologists & network scientists the wiki is good. Then again, the explanation during the call was also fine. Experts and really interested non-experts will read you no matter the effort you put in polishing the information.

If you want to a non-expert that’s not particularly interested or aware of the possibilities, you need to present it differently. More explanations (eg. what is an ethnographic code?) and examples (eg. Winnie wants to know more about influencing policy, how is the software useful for him?) need to be added. I’d also structure the information differently, such as putting the relevance/novelty first (eg. the collective intelligence bit) or adding an element of storytelling (like you did in the policy maker example at some point). Right away there needs to be a justification for a non-expert to take the effort to understand the rest of the wiki.

Ethno/anthro

We set out to build a research method that extends the reach of ethnographers and other qualitative researchers. GraphRyder is not a finished product: it is a tool for analysis. Once researchers have interpreted the data, with the help of a lot fiddling with it, they are ready for returning the results of the whole exercise.  The non-expert-ready form of approaching the data is generally post-analysis: a report, an executive summary, an infographics, depends how much you want to editorialise it.

That said, you make a good point. When we were writing the methodological paper, I realised there are no experts on this stuff. Some people understand ethnography. Other understand networks. Almost no one understands both. Even in the research group we tend to be either-or, with maybe a vague idea of what the other camp is all about (and the vast majority of humanity, of course, understands neither, nor does it care). That leaves us with two strategies: either we write in a heavily context-dependent way (explain ethnography when submitting to netsci journals, explain graphs when submitting to anthro journals); or we assume that no one knows anything at all. In the case in point, we were submitting to something called the Internet Science Conference. Whatever the hell Internet Science is, it is unwise to assume an audience that understands networks like @melancon , or ethno like @Amelia . So, we actually introduce all concepts (not bad for a 7 pages paper!):

Ethnography is a qualitative research technique aimed at discovering how a certain group of humans perceives a set of issues […]

So: what you propose is relatively simple to implement, though we would definitely have different pages for the background information. Actually, I already added a page on key concepts. Why, the wiki is three hours old! But my question to you (and all) is:

Is this interesting for people who are not involved with ethno/anthro/qualitative research? Should we even make the effort to talk to them?

Final remark for @WinniePoncelet and the “science communication” crowd in Reagent and elsewhere. I am def no expert, but I find that you cannot dumb down the “how did you get this result” question more than so much. For example, a much simpler work like my own network analys of Horizon 2020 consortia looks like this. Going all out on it, restricting myself to only one result, spending time in crafting a nice viz, writing with Hemingway App, etc. I managed to get to this. And that’s not an analysis tool, but an actual result! What’s your experience on this issue?

A new field or a new method?

From the methodological paper I gather you consider it mainly a method at this point. Then it makes sense explaining it in terms your intended user base understands. If it’s part of a new field, it makes sense to explain it in and of itself. A field being a set of methods, people, language, … that have crossed some threshold. Then there’s less need to justify its usefulness for another field or set of users.

The line is blurry because it’s interdisciplinary research: you’re building new tools. Thinking about it more deeply, it makes sense to me to place what you’re doing in a spectrum from monodisciplinary research to transdisciplinary research. This excerpt from my workshop on interdisciplinary collaboration & communication explains a little. It seems to me you’re doing the interdisciplinary thing, but not yet to the extent it can be considered a new field?

For me this classification in terms of mono, inter, trans is tied to communication efforts. In the mushroom material research, Elise and I went from multidisciplinary to interdisciplinary and ultimately transdisciplinary. Now we’re at the point where we actively involve society in the research through workshops & outreach. Initially our communication with each other had to improve, then to other professionals (eg. designers, biologists) and ultimately to the public (hobbyists, children, …). It took progressively more effort in terms of communication and practical things: translating experiments into digestible, accessible activities for people to interact with). I feel like the payoff is there (we get ideas, connections, an audience, …), but going that far with communication does take a lot of time. We can do it because I’m not involved in the lab research anymore, only as project manager for outreach and missions.

There is of course also the argument that what has been researched using tax money should be made accessible and understandable apart from there being a payoff for your research by involving society.

Making complex information understandable takes time, so the most economical solution is to define your audience and write for them, rather than for any audience. The latter means taking into account the group that has the hardest time understanding, without compromising in content or form too much for any group. That is hard and you always end up losing detail or nuance. You wouldn’t write so that an 8 year old could understand you, even if you were perfectly capable, unless your plan is to go specifically to them with your post.

A technical person is okay with a separate glossary page. For engaging a non-technical person, who perhaps has no experience with technical texts and their structure, having to check words in a glossary may break the story and their attention. Your post on your own blog about the results of the Horizon 2020 research is pretty good overall. I see the distinction between result (good potential to be relevant for a non-expert) and analysis tool (less potential to be relevant for a non-expert), yet it comes down to your own ambitions, judgment and what you want to say to whom.

So all in all, I understand the communication aspect in this context in terms of where you’re at with the research (new field or not? Multi, inter, transdisciplinary?), where you want to go (Multi, inter, transdisciplinary?) and how you want to do it (who to involve and why?).

Maybe @NiekD or @Scigrades can pitch in as well.

New fields are for giants

… not for the likes of me. More humbly, I see a market opportunity. Consulting companies grow by “owning” a unique methodology; without it, you are just cannon fodder, competing on price. My role in ER is to build a niche where we can claim to be number one in the world. Even if it is small, it is probably enough to ensure interesting work and a measure of prosperity.

Thanks for pointing to the text you wrote. The distinction does not really apply here, because one of the ingredients of our method, ethnography, is, by your definition, itself trans. I guess that makes SSNA “meta”, or something! But the way I think of it is rather inter: let ethno people do their thing as normal, let network people also do their thing as normal, and then build a new layer on top of it that enhances both and brings about a new synthesis. Problem is, to interpret the new layer (SSNA itself) you need to speak both languages. It speaks clearly to me, but I need some effort to make the ithers see. IMHO the best SSN analyst is someone like @Noemi or @Amelia : social scientists with an inclination towards formal reasoning. The network science is, as of now, pretty basic, so it’s much easier for them to grasp it than it would be for a network mathematician to teach herself the anthro and socio. Maybe later (if there is a later) the need for mathematical sophistication will increase.

In a way, we are all children of Jacob Moreno and the sociometrists of the 1950s. Not by chance, these were all socio people, and Moreno himself was a psychiatrist (originally from Romania, like Noemi!).

I would be super-interested in the opinions of @NiekD and @Scigrades , of course!

New fields for ants?

Interesting thread I was new to. Personally, even though I’m no giant, I love contributing to these really new trans- or inter-disciplinary areas. I was deeply impressed, around 1975, reading “Scientific Knowledge and Its Social Problems” by Jerome R. Ravetz, originally from 1971. Rom Harré used to recommend it in his lectures in Oxford.  It had a great chapter headed “Immature and ineffective fields of inquiry” (you can get some results from searching for the phrase in quotes) which set out some of the dangers, and excitements, of these areas.

It’s easy to imagine this is all about giants, and maybe it used to be, but I don’t think that is so any more. The world (of knowledge) has grown so vast, no one is a giant anymore. If instead we constitute ourselves as collaborative, cooperative intelligent ants, though, I think we can make progress. Ants (real ones) do so much with so little individual intelligence. What could we do, if we were to coordinate with the power of ants? This links so well across our discussions … it is the culture of collaboration that is key, and that in turn rests on how we interact day to day – “micro” level, you could say. To me, it involves deep listening to each other, just as care more generally does. It involves co-creating, and co-managing, a culture in which the commons is held in higher regard than the individual, and where the commons wisdom knows who to ask about particular issues; where we all help to redirect, rather than individualistically claiming intellectual territory for ourselves. In the network, some people are valuable more as nodes, others more as links.

My personal interest in the thread might start off by asking about the nature, the provenance and the governance of tags themselves.