Very interesting post from David Mckenzie about the (mis)use of pictures as complementary data for the evaluation of grants to microentrepeneurs in Ghana.
Not quite the same, given we did not have a real experiment, but for the 2008 Economic and Social Progress Report of the IADB, where the topic was social exclusion, we integrated videos as case studies. The main goal was to have cases coming directly from the people that suffer the exclusions, so we did an open micro-documentary contest ("The faces of exclusion") in Latin-America, and received around 120 films from all Latin-american and Caribbean countries. The winners can be watched here.
Since then, I have always been thinking about how to integrate audio-visual in impact evaluations, particularly how to really have videos or pictures as an unbiased tool for evaluation, instead of a (deeply biased) "soap opera".
I wonder if having short videos, with pre-defined characteristics will work better than pictures in the setting described by Mckenzie, in terms of given a more general overview of the situation.
It would be great to hear from someone that had used videos not just as a complement to show the setting of a project, but as part of the data for the evaluation itself.
PS: When I left the IADB, and was ready to go for my PhD in Geneva, people from the World Bank approached me to repeat the micro-documentary contest for a study about the effects of global warming. I had doubts about following a career as a film producer or a doctoral student. I chose the latter, maybe wrongly....
At the WB they finally implemented the contest, with same format as we did in the IADB. Here the results: