Experimental Research

Can we implement a small set of laboratory controlled experimental studies that will allow us to better understand the power and constraints of public access ICT facilities? By laboratory experimental study we mean a controlled examination of the effects of variation along some detailed independent variable on some detailed dependent variable that describes an outcome of interest. For example does shared use of a PC as opposed to private individual use (the independent variable) enhance learning outcomes on a memorization task (the dependent variable). A grand ambition to these research activities would be to establish a set of cases when public access is most valued (and those cases when it is least effective).

Studies could focus on interest areas such as:

• Environment. Are there environmental conditions that can enhance some aspects or some people’s participation in public access ICT facilities? For instance cybercafés that offer privacy to users can enhance some experiences while cafes that explicitly turn computer monitors towards the public can enhance others. (In one experiment in India making the computer monitors visible to passerbies on the street was shown to increase female usage.)
• Technology. What new computer hardware or software designs might enhance the public access use models? What co-present shareware technologies would make a big difference?
• Application. Are there specific applications or types of activities that are particularly well served in public access facilities? Alternatively, are there particular application designs that can enhance their effectiveness within a public access context?
• Geography and context. Findings that can be applied broadly across countries and regions have obvious appeal. Yet we know that the success of public access facilities can be highly contingent upon culture, context, and geography. On the one hand what context is most effective? On the other hand, can we develop innovations that support broader decontextualizations? Or is that a bad idea?
• User profiles. Related to context and culture is the question of what users are well served by what sort of public access settings and systems? For instance what variations change user demographics as it relates to education levels, gender, economic status, age, community, and so forth?

This is, admittedly, an overly broad set of questions and the key to success in this type of laboratory experimental research is to take small sharp steps – ultimately that will build up to our broad conclusion.

Methods and Implementation Plan

We propose a small set of laboratory experiments that will attempt to illuminate when public access to ICT’s may be preferred to private access and what conditions create the best public access outcomes. A useful collaborative model is to partner research groups from the RWG with country teams. Through this partnership an experimental research design will be established, a protocol refined (perhaps initially in a highly controlled environment, e.g. at a US research university) and then experiments will be conducted in-country executed by country team members.

For Seattle

I, personally, do not feel that I have a good grasp on related research in this area. Some attention should be placed on a literature review. Additionally, and most critically, specific experimental questions should be proposed. In other words, what are the independent variables we would most like to examine? Moreover, it would be wonderful if we could develop a comparative research design by establishing a single dependent variable (or small set of variables) that we could apply across our experimental perturbations. In other words does there exist some specific computer activity (like the learning task mentioned above) that would be a suitable surrogate for more general user success. Or do we need a suite of such activities measuring learning outcomes, time to solution, user satisfaction, etc.?