Details
HealthBridge has a team in Togo now carrying out the first of three surveys on bednet useage for malaria control. It will be a challenging series of surveys for many technical and logistical reasons. How to properly draw a random sample from an uncensused population and how to get survey teams to remote villages over poor roads are two of the problems we wrestle with for hours. But those are issues we know we can figure out, and it is not what keeps me up at night. What really worries me is this: When we finally get to the randomly selected, remote households, do we know what questions to ask and how to ask them? We know what we want to learn through our surveys, but do we know how to learn it?
TED is a wonderful online resource that houses hundreds of talks by global leaders in Technology, Entertainment and Design. A talk by behavioral economist Dan Ariely, presents examples of how the design of questions influences the answers that are given to those questions. When individuals are given multiple choices -including choices that are never selected - the never-selected choices influence which of the selected choices are chosen.
So if there is a choice to buy (1) A for $10, (2) B for $20 or (3) A and B for $20, we know that nobody would select option 2. Yet its inclusion in the list of choices increases the proportion who will select option 3.
Ariely was focused more on marketing and economics, but what scares me is how it may play out in our surveys. In our surveys we include questions that are directly analogous to Ariely's example, in which we read all the possible options and ask the respondent which examples apply to them. For example, we may ask which of the following vitamin A rich foods they eat: mango, papaya, liver, pumpkin, squash. The general rule of Ariely is that never selected options will increase the frequency of selection of the option most similar to it. So perhaps they never eat pumpkin - but including it as an option may increase the number of people who report eating squash. And are there other, less analogous, questions where such influences may exist? We don't really know and it is enough to keep me up at night.
Maybe I haven't explained it well. Watch the ted video online and see if you share my fears. And while you are there, also watch this 15 minute presentation by Richard Wilkinson, in which he sums up the evidence presented in his book, The Spirit Level, on the negative health effects of social inequity.