By Darren Cambridge & Tom de Boor
Darren: Education is awash in data as never before. Increasingly, the challenge for education decision-makers is literally seeing what all this data means and communicating it to other stakeholders. A collaborative data visualization service like IBM’s Many Eyes, optimized for the needs of educators, could help address this problem by allowing educational administrators to easily share data sets and use the service’s platform to generate and share creative data visualizations from any data set that’s been uploaded.
Needless to say, access to sensitive data sets would need to be limited (education officials, associations could host and vet participants) though ultimately the goal would be to share the resulting visualizations widely. It would probably help administrators if visualization experts were regularly invited to participate so as to generate a steady stream of creative visualizations to model and inspire. But what other features would make this type of service particularly relevant and useful to educational decision-makers?
Tom: I’ll kick this conversation off with a couple of thoughts, and will have more depending what others have to say. First of all, I think it’s very important that a service of this type support the creation and sharing of visualization *sequences*, not just single visualizations. One of the big trends in data visualization in the last year or so has been using visualizations to tell stories, and it seems to me like this could be particularly important in education, where data is supposed to be used to make the case for change, and making arguments for anything typically requires narrative, not just single snapshots.
Another thing I think an educational version of service like this has to do is recognize the multi-level nature of education. Specifically it needs to provide the built-in capability to easily aggregate data sets that have been submitted, either to help decision-makers project results to higher levels of organization (e.g. combining data submitted from districts to get a sense of state-wide results, which can then be visualized) or to compare results (e.g. comparing results from different states in visual form). To this end (and others), ideally the service would also have built-in statistical functionality to help decision-makers determine how real any differences they are seeing really are (and show this, where necessary, in the resulting visualizations). Some decision-makers may not want their data combined and/or compared, of course, and the system would need to provide them with the option to “opt out” of those kinds of use of their data.
Another thing, albeit somewhat more peripheral: despite the amount of data already available, I could see decision-makers at all levels finding there’s data they want/need that they don’t have. It would be great if the system allowed participants to broadcast data requests to anyone in the community at a specific level and/or with a specific job title. Even better if the system aggregated and organized these requests, so that researchers and high-level decision-makers could get a snapshot sense of what kinds of data are most wanted that don’t seem to be getting collected today, and appropriate actions could be taken.