Probably a bit late in to the discussion but I was wondering if someone can share their experience (if any) with application of Randomize Control Trial (RCT) in evaluating ICT projects. Basically I was hoping to get answers to the following questions:

1. The time and cost involvement from inception to results.

2.  The sample size that was required (control +treatment group) and the unit of sample (individual, class, household etc)

3. The size of the project (in terms of their geographic scope and number of interventions, other than the one tested, that they were working on)

4. The challenges faced.

The reason for asking this is that from what I have heard and read, it seems RCT has been applied quite successfully in projects which have few interventions, with very precise crisp measurable goals (literacy, health indicators, reduction in absenteeism etc) and significant access to resource (time and money). I am bit apprehensive when it comes to applying such methodologies in  projects with 20-30+ interventions, working on light touch approach, private sector driven and focusing on more complex indicators like competitiveness, sustainability, systemic change etc.