Statisfaction

Meta-analysis of Aid Programs

Posted in General by Pierre Jacob on 2 April 2012

Hey hey,

Here I’m going to advertise a project of Eva Vivalt from the World Bank, who wants to write a book on meta-analysis of aid programs.

The project is hosted there:

http://www.kickstarter.com/projects/972584134/what-works-in-development-10-meta-analyses-of-aid

It’s on kickstarter.com, which means that if (and only if) enough people pledge to give money to the project (here at least 10,000 US$), then Eva Vivalt and her colleagues will write the book and print it and everything. It’s an innovative model, which reminds me of MyMajorCompany.

More about the project itself. Meta-analysis is the art of combining multiple, heterogeneous studies and aggregate them to get an overall answer to a given problem. It’s especially popular in epidemiology: suppose that a lab has run a clinical trial on many people and that another lab has done a similar study in a different country, how does one combine both studies to get more precise results, while still taking into account the specificity of each study? Which study is contributing the most to the results? What differences are there between the two studied populations?  Meta-analysis allows to answer such questions. Statistically speaking it usually amounts to some kind of hierarchical model, for instance a mixed model.

Eva’s project is about using this statistical machinery to assess the effectiveness of aid programs. It seems meaningful to me, because there might be many studies studying each particular aid program, but for the good Samaritan the problem is really about choosing who to give money to. Meta-analysis seems able to help him out in this dilemma. Why this research project is hosted on the Internet and is not funded by a large public research agency, I’m not sure but the project description page answers that to some extent. If you have any opinion on this funding system leave a comment!

They have to collect 10,000USD during the next 27 days, so if you readers can talk about it to your data-friendly friends, it’d help them for sure. Good luck to them!

One Response

Subscribe to comments with RSS.

  1. Julyan Arbel said, on 9 April 2012 at 23:56

    Good idea. Without meta-analysis, a strategy is to diversify ones portfolio and give a bit to several programs, but it’s not optimal🙂
    I’m curious to see how she will compare programs with very different goals and areas of action.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: