Feed Subscriptions
Rolling Links
- Open Code Blog - New York Times A blog about open source technology at The New York Times, written by and primarily for developers.
- Postgenomic Postgenomic collects posts from hundreds of science blogs and then does useful and interesting things with that data.
- hyperexperience for a better sense of reality
- Treehugger Well known green blog.
- Mashable The Social Networking Blog
As we near the target deadline for releasing our prototype recycled research project, it makes sense to talk a little bit about what we mean. Here is an overview of the project, what we’re doing and why we’re doing it.
The recycled research project deals explicitly with 1) software tools that achieve some empirical experimental goal and 2) the open sharing and dissemination of data that results from these tools. Recycled research projects are intended to be sharable, embeddable and (ideally) useful outside of their explicit research goals.
The prototype recycled research tool is a theme for popular blogging software Wordpress. The theme includes mouse-tracking software that monitors user behavior in a blog using the theme as well as computer interpretation of blog content. Options within the theme allow different visual layouts. One goal of this tool is to understand how different layouts for different kinds of content effect usability. This is an exemplary recycled research tool, packaging a useful software tool with a method of inquiry and experimental treatments.
This kind of example showcases several technical goals necessary for the recycled research project:
- Viral Experimentation: The model constructed in this approach is a viral experimentation model. One created piece of software is offered for use. From here, several users can opt to use this software on their own web site and projects. Any settings within the software serve as experimental treatments. Each deployed instance is, in turn, used by a number of the deploying user’s web visitors. This is the data for the experiment. We turn each deployer of the software into an assistant experimenter, with the protocol coded into the software to avoid mistakes. Any of their visitors could also become deploying users if they like the software and choose to install it for themselves.
- Explicit Research Licensing: Software developed for recycled research is not exactly free software. There is no monetary cost, but there is a cost in terms of computing and community. The software license associated with each project, the license to learn, requires that in exchange for use of the software the user allows the software to perform observation, analysis, and experimentation. In some cases this may be completely passive, such as observations of user interaction with the system by observing mouse tracking and system status. In other cases this may be more active - actually altering presentation or prompting users of the software with questions and stimuli. This ‘component for data collection and experimentation’ is explained in detail to possible users of recycled research software.
The idea is that they are able use the software freely and in exchange they agree to support the kind of research questions that can be answered by the software. This explores user reaction and adjustment to an unusual kind of software license so we can measure the response to simple exchanges. Here the exchange is participation in an open research project for using a useful piece of software.
- Open Data Sharing and Analysis: All of the data collected by these experiments is offered openly in a central repository. This repository allows sorting by conditions (software options) but takes efforts to remove identifying deployer information. From here the data can be manipulated, downloaded, graphed, and modeled. Use of this data can accomplish several things, including: informing the design of future iterations of the software; constructing new research hypothesis; and answering existing research questions. In some sense the public sharing of this data releases the perception that software that gathers data is in someway malicious. Here the data can be freely observed by anyone.
- Pseudo Genetic Adjustment: An interesting consequence of the way recycled research projects are constructed is the potential for increasing the role of the computer in facilitating experimental iterations. A given project has both a tangible software artifact and (usually) a mechanism for evaluating that artifact. This capability for evaluation can be levered to construct a basic ‘evolutionary’ mechanism that allows changes to experimental conditions as a result of successful or unsuccessful evaluation results. For example, consider a fictitious experiment that explores the use of color, size, typography, and spatial positioning as representation tools in an introductory presentation of the lambda calculus. A highly simplistic ‘genetic’ algorithm could be implemented to vary experimental treatment based on correct and incorrect subject inputs. Although not a complex example, structuring this kind of approach within recycled research projects encourages broader, more flexible experiment design through computer augmentation.
As we release Recycled Canvas, our wordpress theme, we hope that people try it out. We’ll use the data we collect as part of an iterative design process and usability report on some of the issues with the theme, and with blogging in general.
