Teaching SD to blind people

This forum is for discussion of topics related to System Dynamics. You must be logged in to post but anyone who has registered on the System Dynamics web portal may post. Note that there may be a delay of a working day or two after registering before you can post anything.
Post Reply
Andreas Groessler
Posts: 17
Joined: Sun Feb 22, 2009 9:59 am

Teaching SD to blind people

Post by Andreas Groessler » Wed Feb 26, 2014 3:53 am

Does anybody in this forum has experiences with or ideas about how to teach SD to a blind person? One of the strengths of our method is the clear visualization of structures--but what do you do when exactly this feature isn't helpful?

Richard Dudley
Posts: 65
Joined: Sun Jan 11, 2009 1:17 pm
Location: Etna, New York, USA
Contact:

Re: Teaching SD to blind people

Post by Richard Dudley » Wed Feb 26, 2014 8:42 am

That is a very interesting question. I can imagine that there might be a way to do this conceptually using sound... especially if the person is musical. But musical complexity probably goes beyond, or is different from, the type of complexity we think of in SD.

Richard

Robert Muetzelfeldt
Posts: 29
Joined: Sun Jan 08, 2012 7:20 pm
Location: Edinburgh, Scotland, UK
Contact:

Re: Teaching SD to blind people

Post by Robert Muetzelfeldt » Wed Feb 26, 2014 7:07 pm

Hi Andreas,

I have no experience of working with System Dynamics with visually-impaired people.

However...
I am currently developing a toolkit for building web pages for viewing, building, editing and simulating System Dynamics models, and have thought (briefly) about accessibility issues. The toolkit is based on the idea of a set of discrete widgets - e.g. one for displaying a SD diagram, another for plotting results) which can be assembled by web page developers any way they want. This widget-based approach makes it tempting to allow oneself to think of novel ways of working with models and/or simulation results, such as addressing accessibility.

I have experimented with a couple of ideas on the simulation-results side. One is an audio rendering of the value of a chosen variable, with pitch corresponding to its value. This works well: it is quite easy to get a picture of the changing value from the sound. The other is a text-based summary of the behaviour of a variable: e.g.
- starts at 10.0;
- rises to a maximum of 17.3 at time 12;
- falls to a minumum of 3.2 at time 34;
etc
It is easy to extract this from the internally-held simulation results, and of course can be rendered into Braille, or speech using text-to-speech software.

On the diagram side, I posted a request on precisely this topic on the W3C Web Accessibility Initiative (w3c-wai) mailing list about a year ago, and got a couple of interesting responses. Here's one response, from Jonathan Avila:
Hi Robert, I and another colleague created a whitepaper on this about 6 years ago. We came up with a keyboard based method to provide traversal of these diagrams. We also created a concept of exposing this information through an accessibility API for given nodes and edges using the Java Accessibility API. Our plan was to allow the user to traverse from a node to an out edge by using right arrow and left arrow to move onto an edge. Once on an edge up and down arrows would allow the user to move among edges of that type. Identity info on each node and edge would describe it such as how many in or out edges it had and for the name, state, type of the node. From an edge right arrow would move to the our node and left arrow to the in-node.

We also devised a way to allow the user to ask questions. For example, the user could mark a node and then traverse to another node and ask questions about the path from node a to b such as how many paths and how many edges between the nodes. We also indicated that a summary method to describe the system as a whole would also be necessary. At that time we were planning to use yFiles [RM: ?] to do this. The whitepaper did not result in the actual implementation of a public example of this.
I haven't tackled this yet, but I think it should be fairly straightforward to implement - as a widget in my toolkit, or as a stand-alone bit of code to handle text-based model representations (e.g. Vensim .mdl, Simile .pl, or XMILE).

Please feel free to contact me off-line if you'd like more details.

Robert

Kim Warren
Posts: 113
Joined: Mon Feb 16, 2009 5:56 am

Re: Teaching SD to blind people

Post by Kim Warren » Thu Feb 27, 2014 4:33 am

I had experience with a blind person in one class, without the benefit of these great technical solutions.

The first thing I realised was that he had an amazing ability to visualise if given only small amounts of information - I don't know if this would apply to people blind from birth [this person went blind in late childhood]. So, so long as I said out loud what was on the board or screen, he could follow OK. Then, after class, he would come to the front of class and I would move his finger over the diagrams and time-charts. I was quite amazed how, but he understood very well.

Luckily, he had helpful class-mates, and the support of a sighted reader who could not only read to him, but also write/draw any diagrams he asked for [though some I know he did himself]. With these additions, he passed my class with a good grade.

William Braun
Posts: 65
Joined: Sun Jan 11, 2009 11:18 am
Location: Cleveland Heights, OH

Re: Teaching SD to blind people

Post by William Braun » Tue Mar 04, 2014 4:26 pm

For the real basics you could use cardboard cut-outs for stocks, flows, auxiliaries, and parameters, and one for delays (that could be superimposed on flows and auxiliaries). You could build a very simple model to help them visualize essential dynamic relationships. For links use straws, toothpicks, heavy twine, or the like.

In the same vein, try something with Legos. While you form each of the dynamic objects you could talk about their functionality in SD. That could increase tactile involvement and give him/her a greater sense of "constructing" a model.

Also, a felt board might work well.

Andreas Groessler
Posts: 17
Joined: Sun Feb 22, 2009 9:59 am

Re: Teaching SD to blind people

Post by Andreas Groessler » Wed Mar 05, 2014 2:52 am

First of all: many thanks to all people reacting on my question. It was very good to see that there are some ideas around that could help...

For my concrete problem, things are a bit more difficult than my original post might imply: the student is part of a huge class (ca. 150 undergraduate students), we are already into the course for a few weeks (the issue didn't become immediately apparent although in hindsight...), I must offer a "solution" to the student asap (otherwise he needs to change the course) so there's not much time to try out the fascinating technical solutions that were indicated...

Anyhow, I will have another meeting with the student today and keep you posted what we agree on.

In the meanwhile, for archiving purposes, I will post other reactions that I received per email and in another forum.

Andreas Groessler
Posts: 17
Joined: Sun Feb 22, 2009 9:59 am

Re: Teaching SD to blind people

Post by Andreas Groessler » Wed Mar 05, 2014 2:56 am

Answers I got outside this forum:

**********************
I would ask how math, especially geometry and graphs are taught to blind people. I guess they use voice-operated computers in a lecture and can easily capture basics of SD. However, graph reading will be different.... Have you tried to ask at a Montessori school and associations like DBSV.org? (Max Happach)

**********************
I remembered this paper from the Italy 2002 conference. I don't know anything about it, but sending it along to you. (Roberta Spencer)

Pfeiffer, David, "Symphony of Eigenvalues: A Prototype for Sonification in System Dynamics"

Sonification is the use of nonspeech audio to convey information and can enhance the ability of system dynamics in the analysis of relationships between structure and behavior. Today's desktop computing power allows users of software such as MAX and MSP to create a variety of sounds whose parameters are determined by the mappings of imported data. The prototype discussed used pitch, timbre and amplitude to describe the values and first and second order differences, respectively, of the state variables. A two-parameter model of Hares and Lynx demonstrated the sonic effects, while another two-parameter model of Alcohol and Cells featured the addition of sonified eigenvalues, illustrating the combined fields' potential to provide an early warning detection to changes in feedback loop dominance. The prototype's ability to express 10 parameters of interest with just 4 sounds should appeal to the fields of business, education, and the arts.

http://www.systemdynamics.org/conferenc ... EIF1_S.ZIP

Robert Muetzelfeldt
Posts: 29
Joined: Sun Jan 08, 2012 7:20 pm
Location: Edinburgh, Scotland, UK
Contact:

Re: Teaching SD to blind people

Post by Robert Muetzelfeldt » Thu Mar 06, 2014 8:46 am

Motivated by this discussion, I looked around a bit and found that there are various ways of producing "tactile graphics" for blind people (http://www.rnib.org.uk/professionals/ac ... phics.aspx, and follow the Production Methods link). They are all relevant to the production of tactile System Dynamics diagrams, but the one that is the most immediate is the use of "swell paper". This is special paper that can be run through a laser printer, then treated by a special device (a "fuser"). The fuser raises the special coating on the paper underneath black or dark printing by about 1mm or so, the height varying depending on the darkness or stippling.

I obtained some swell paper from http://www.zychem-ltd.co.uk (I have no financial interest in this company; and I assume that there are suppliers of similar products in other countries), and printed out an enlarged part of a System Dynamics diagram. Given that fusers are expensive, I experimented by holding the paper under a radiant heat source (a downlight), and it worked OK, although one would probably need quite a strong lamp held at some distance to get an even, controllable raising of the surface.

This seems to offer a promising route for addressing one aspect of the current topic, namely allowing blind students to immediately "read" a System Dynamics diagram developed in class. Note that there are guidelines on diagram preparation for tactile graphics (strong bold lines, no fine detail etc), so one might need to pre-process diagrams, e.g. by zooming in. Also, this (obviously) does not address the need to use Braille for the text, which could require customisation of the modelling software.

Robert

Andreas Groessler
Posts: 17
Joined: Sun Feb 22, 2009 9:59 am

Re: Teaching SD to blind people

Post by Andreas Groessler » Fri Mar 21, 2014 3:32 am

Thanks, Robert, again very helpful.

Although this won't be a solution to my special case (because I simply started too late considering all this), I'm trying to come up with some general lessons with the help of my student. I will report on them once available...

Post Reply

Who is online

Users browsing this forum: No registered users and 1 guest