Project of the Month, November 2011: The Number Race

The Number Race

The Number Race is an “adaptive game” intervention designed for remediation of dyscalculia (mathematics learning disabilities or difficulties) in children aged 4-8. It may also be useful for prevention of dyscalculia, or to teach number sense in kindergarten children without learning difficulties. The software was made by experts in the field of mathematical cognition and cognitive neuroscience, and is based on our current knowledge of the brain circuits underlying numerical cognition. Unlike most educational software it is open-source and distributed free on SourceForge. It has been translated into multiple languages, and has been evaluated in several efficacy studies.

(See previous Project Of The Month winners)

Why and how did you get started?

The original ideas which lead to the development of this software were generated
through numerous discussions between Stanislas Dehaene, Bruce McCandliss, and Michael
Posner, in the context of discussions on learning and the brain fostered by the OECD (Organization for Economic Cooperation and Development). The project started in earnest in 2003 with a Fyssen Foundation grant to Anna Wilson. The Finnish team joined the project in January 2005, and have been making ongoing improvements.

Who is the software’s intended audience?

Parents, teachers and researchers are welcome to download and use the software with children. It is designed for children to use with a minimum of supervision.

What are a couple of notable examples of how people are using your software?

There are two main groups of users. First, the game is publicly available in several languages, therefore teachers and daycare teachers use it. Secondly, has been used by researchers to analyse factors related to the development of children’s number skills. In Finland, thousands of teachers and preschool teachers have downloaded our game via the Lukimat.fi web service, a nation-wide learning disabilities information website.

What are the system requirements for your software, and what do people need to know about getting it set up and running?

Windows/Linux/OSX computer required. Java 1.5 (or later). Ability to download and run jar files.
The forthcoming version 3 will have a Windows specific installer/launcher .exe file.

What gave you an indication that your project was becoming successful?

Once the software was developed, we launched collaborative studies with schools, educators, and other researchers, to test whether it was efficacious (Räsänen, Salminen, Wilson, Aunio, & Dehaene, 2009; Wilson, Dehaene, Dubois, & Fayol, 2009; Wilson, et al., 2006). The results were encouraging: the children’s arithmetic skills were clearly improving as they were playing. Although the effects were not huge, they provided a platform from which we could think of further improvements.

There are few controlled studies evaluating computerized interventions for children’s number skills. This game is one of the few with several studies already published.
Publishing these results in scientific journals encouraged many of our colleagues to undertake studies of their own. Our software has now been translated into 9 different languages (Dutch, English, Finnish, French, German, Greek, Italian, Spanish and Swedish).

What has been your biggest surprise?

The extent of ongoing and global interest in the software.

What has been your biggest challenge?

Limited resources (time and money), as always.

Why do you think your project has been so well received?

There is a real hunger for research-based intervention for dyscalculia.

What advice would you give to a project that’s just starting out?

Realise that software production is resource intensive, and plan accordingly. Seek out advice from people who have completed similar projects. For educational software, if possible, find ways to trial your core instructional factors prior to starting development.

Where do you see your project going?

In November 2011, the Niilo Mäki Institute team will release a totally new revised version (3.0), which implements recent research findings on how children learn number. We will also release a GUI utility which will provide a simple way to translate the game into new languages.

Planned further developments include:

* Make language packs self installable.

* Improve game data storage.

* Make an XO-1 and/or mobile version.

* Add learning assessment tasks into the game.

What’s on your project wish list?

To be able to continue developing the game to follow ideas from the latest research findings. We would like to find the most important features of children’s early number skills development and learn how to change that knowledge to a motivating game for learning. Therefore the number one thing in our wish list is intervention studies using our game versions.

What are you most proud of?

The multidimensional adaptive algorithm developed for the software by Stan Dehaene which models the child’s current knowledge state. It pushes new learning by delivering problems on the boundaries of this knowledge, but at the same time maintains motivation by ensuring that these problems are not too difficult. This algorithm can be used for teaching any learning tasks.

Secondly we are proud that we not only developed a game but were also able to conduct research on that game. We hope that this model encourages others to follow.

If you could change something about the project, what would it be?

There has been some small scale research projects in which the game engine has been used. It might be a good idea to put all those different versions and the results from those experiments to one place to encourage people to develop new of versions of the game based on different theoretical approaches about learning.

How do you coordinate the project?

For most of the development we had only one primary programmer (Anna Wilson, then Alex Maslov), so this has not been a problem. Bugs assigned to lead developer (Maslov). When “spare” resources like temporary workforce available task reassigned if needed.

How many hours a month do you and/or your team devote to the project?

For most of the development we had only one primary programmer (Anna Wilson, then Alex Maslov), so this has not been a problem. Bugs assigned to lead developer (Maslov). When “spare” resources like temporary workforce available task reassigned if needed.

What is your development environment like?

The software is written in Java. For the first version we used Eclipse as a development platform. Now MacBookPro, OSX 10.6.4, Java, Eclipse 3.7.1, YourKit JavaProfiler 8, Maven 2&3, Mercurial over Subversion.

Milestones:

Version / Date Milestone
2003 Project began
2004 Testing/debugging; pilot study with 5 children
2005 First research study in Paris, using version 1
2006 Second research study in Paris
2006 Version 2 released on www.unicog.org
2007 Version 2.21 released on SourceForge, including multiple translations
2009 Version 2.3.5 released on SourceForge

How can others contribute?

1. Translations into new languages.

2. Further development of source code is welcome.

3. Transcoding the game into new programming languages (e.g. pygame) to get it running easier on new platforms (E.g. the OLPC “100$” XO computer, tablets, mobile phones).

4. Easier access for children in developing countries.

5. Teams willing/interested in running intervention research.


More projects of the month
https://sourceforge.net/projects/numberrace/

Project name: The Number Race

Date founded: 2003

Project page: https://sourceforge.net/projects/numberrace/

Project Leader


Prof. Stanislas Dehaene

Prof. Stanislas Dehaene

Occupation:Professor at the Collège de France, Director of the INSERM-CEA Cognitive Neuroimaging Unit, France
Location:Paris, France; www.unicog.org
Education:Ecole Normale Supérieure, Paris; MS (Paris VI, Mathematics), PhD (EHESS, Cognitive science)


Pekka Räsänen

Pekka Räsänen

Occupation:Executive vice director (research and communications), Niilo Mäki Institute, University of Jyväskylä
Location:Jyvaskyla, Finland
Education:Clinical neuropsychologist

Key Contributor


Dr. Anna Wilson

Dr. Anna Wilson

Occupation:Lecturer, ESHD, College of Education, University of Canterbury, New Zealand
Location:Christchurch, New Zealand
Education:PhD (Oregon, Psychology)


Alexander Maslov

Alexander Maslov

Occupation:Software Designer
Location:Jyvaskyla, Finland
Education:M.Sc. (subj. Computer-aided Systems Software), Kharkiv National University of Radio and Electronics, Ukraine

Why did you place the project on SourceForge.net?

We wanted the project to be open-source, and it was recommended as a good hosting site for such a project.

How has SourceForge.net helped your project succeed?

I think it has provided more visibility – and also supported a greater volume of downloads.

The number one benefit of using SourceForge.net is:

Exposure to a wide audience. One disadvantage is that the user interface can be a bit difficult for novices. For instance users need to download two files for our software to run, but often don’t realise to look under “browse all files”.