Measurement in Haiti

by admin on March 12, 2010

(posted by Erica Gralla, in Port-au-Prince, Haiti)

I came down to Haiti as part of Team Boston, to get a better sense of what’s going on and inform our data analysis. It’s hard to analyze data without understanding the context, so I’ll try to do just that while I’m here. I’m not an experienced “blogger”, but I’ll do my best to give you a sense of my impressions and thoughts.

I arrived yesterday (Thursday) with Mischa Shattuck and Marc Zissman from Lincoln Labs. I knew immediately I was in the tropics from the heat on the jetbridge. After a long line at passport control, we joined the scrum of people around big doors through which workers were tossing our bags. The key was to watch as they were thrown in, because once they were piled on the floor it was extremely hard to find them. We watched for our bags as the room emptied out. Marc’s bag arrived, then mine, but Mischa’s never came. We sighed, filed a claim, and hoped this was the biggest problem we’d have. Matt Kercher picked us up in the even greater chaos outside the airport, and we headed in a rented SUV to the US Embassy. My overwhelming impression was that I was back in Africa, with all the people and the noise and the bumpy streets. Of course, Haiti is different, but I wasn’t processing it that way at the time, until I started seeing the occasional crumbled building along the side of the road. But the drive to the embassy is short, and soon enough we pulled up outside a fancy building and piled out of the SUV. New badges in hand, we found ourselves in a different world, all manicured lawns (now covered with army-khaki tents) and air-conditioned buildings. Marc and Matt introduced us around, we set up cots and mosquito nets in our sweltering tents, and had our first taste of the Army’s MREs. The first two days were something of a blur, but my overall impressions of the work are pretty clear.

In some way, the first two days here in Haiti have revolved around measurement. Of course, we are here to support a needs assessment, so it makes sense that we would need to measure something. But what to measure, and how to measure it, are not so straightforward. I arrived in the midst of the ramp-up to our first collection of data. The pilot is on Saturday (tomorrow), the survey instrument is still being finalized, and we are working with the clusters to support the pilot, the principal investigator to sign off on the survey, and the data collection contractors to upload the final survey into their tools. This late in the game some confusion still remains, because everyone wants to find the best way to measure needs and progress in Haiti relief. Of course, it’s hard to know the best way before you’ve tried it. And that’s why we need the pilot tomorrow.

Our goal is to measure “demand,” meaning the needs of the Haitian people, in a broad set of locations, and over time. The Army (who is paying for the assessment) hopes to understand the progress of the relief effort over time, and the gap remaining for them to fill. The assessment is also intended to support the international and non-governmental organizations (who are organized into functional “clusters” in areas like health, shelter, food, and water/sanitation), helping them to understand where and how to focus their efforts.

Measuring needs in Haiti takes all three types of people — those who will use the data (the military, the clusters, and the international aid community), those who design survey questions and methodology, and those who actually carry out the data collection. As an MIT student studying humanitarian relief, I have a vague academic understanding of survey design and methodology and of data analysis. The part I hadn’t really thought about was the collection of the data itself. Yesterday I met the contractors who will manage the data collection. They are a company called Global Relief Technologies [link to], based in New Hampshire. They have a system for collectiong data in emergencies using handheld PDAs that provide the survey questions, camera, GPS, bar code reader, and communication with a central server. They also provide a web-based platform for accessing and doing basic manipulation of the data, including a google maps interface. It’s pretty impressive stuff. And I certainly hadn’t realized how difficult the data collection piece was until I spoke with them.

Finally, in the midst of setting up this survey, it is easy to fall into a narrow focus on the specifics of the data collection — getting the survey questions worded correctly, changing the ordering, etc. However, as an Army Colonel reminded me, it is important to think about the larger context. What else do we need to know in order to get the most out of this measurement effort? We also need to understand the activities taking place; otherwise, our measurements tell us less. Things got better, but why? Did we do something right? Did the local people do something right? Which of our efforts are helping, and which are not? As we go forward with this effort, it will be important to keep in mind the larger picture, so that our careful measurements mean as much as possible.

Leave a Comment

Previous post:

Next post: