Q&A: Using interactive radio to inform development programming

GhettoFM in Nairobi, Kenya, participating in an interactive radio program through Africa's Voices. Photo by: Deborah Sambu / Africa’s Voices

NAIROBI — Gathering research to better inform development programming in conflict zones, as well as other challenging settings, can be difficult. Poor infrastructure, security concerns, and high costs make many traditional research methods prohibitive. Africa's Voices Foundation, a nonprofit research organization with offices in the United Kingdom and Kenya, is working to overcome this through interactive radio.

Africa’s Voices works with development partners to create radio shows that then solicit a response from the audience, through text messages that also include geographic and demographic information about the respondent. The responses are then collected, categorized, and analyzed and can be used by development organizations to improve their programs.

Devex sat down with Sharath Srinivasan, co-founder and director of Africa's Voices Foundation to discuss this method of data collection.

Our conversation has been edited for length and clarity.

How do Africa’s Voices interactive radio programs work?

On the one hand, it convenes inclusive spaces for citizen engagement and discussion, that are public in nature and have a large listening audience, not just among the people who participate. Then, on the other hand, we see this as an opportunity to develop and deploy new methods for analyzing citizen voices toward robust evidence and actionable insights that are relevant to decision-makers.

We start with the evidence needs of a partner that is making decisions that impact a target group’s lives. That could be particular beneficiary groups or the citizenry at large. We then design the media engagement around that evidence need. The idea is to not be extractive in the process of gathering citizens voices. With a lot of traditional methods, the questions are predetermined and often the answers are predetermined as well, for example in a survey. But we are more interested in an open discussion that allows audiences to answer a question that is meaningful and relevant to them, and to be able to express themselves fully to explain why they think what they think.

The real challenge, and opportunity, that digital technologies and new media present, is that there is suddenly a chance for citizen engagement to be a source of social insight. You don’t have to do artificial research techniques, or you can supplement or complement them with these new techniques. By artificial, I mean things like a focus group, or a survey, which is itself a sort of research instrument. It’s not an actual activity in the world, even though it has effects in the world. There is a live reality that we are curating and shaping, but it also provides us with this source of data that we can analyze.

You get many different answers to a question, and so one of the first things we do is develop a coding framework on the themes that emerge from the data. We build it up, from the data, as well as any hypotheses going in that we might have from literature, etc. We develop a coding framework that would have different types of themes that these answers led to, in response to a particular question. For example, we might ask what does the community think about whether female genital mutilation or female genital cutting is right or wrong. People might say it is right for religious reasons or cultural reasons or they might say it is wrong for religious or human rights reasons. We develop a coding framework with the different kinds of reasons people give for whether something is right or wrong. Then we might have a map of the beliefs, where we show the beliefs that dominate, and we give relative importance to those beliefs.

In programs in Somalia, where we broadcast on 26 different radio stations weekly, in a particular program, success means that you can hear from upwards of 6,000 to 7,000 Somalis each week, with many messages. Tens of thousands of messages come in. You build the coding frame, understand the themes, and then you label all of those messages to understand which ones are more prominent and why. For that, we’ve developed a mixture of very efficient methods to support “human expert researchers” who are usually native speakers and aware of the social and cultural context. We have an efficient interface that enables them to label messages quickly.

How is that type of information useful to development organizations? Are the participants self-selecting and not representative of the whole?

I think there are two things that matter most here: What is being said that is the insight and what is not being said. The rigor with which the researchers are very clear about what can be said and what can’t be said from this information, I think is crucial. Once that hurdle is overcome, very useful things can be said, but they have to be understood clearly.

I think the second is that we should always be looking at these methods in the context of other methods and data sources. It’s the complementariness of these methods, with others, that can matter a lot. For development and humanitarian actors, they have very rich contextual knowledge in the programming teams and the people on the ground, in implementing partners who can check for relevance and consistency of some of the findings with their existing data. Or they can pursue further corroboration and validation as well.

“There are no ‘silver bullets’ in this space, but there are opportunities to combine different data sources and different analysis to arrive at a stronger and more confident position of social evidence.”

— Sharath Srinivasan, co-founder and director of Africa's Voices Foundation

There are no “silver bullets” in this space, but there are opportunities to combine different data sources and different analysis to arrive at a stronger and more confident position of social evidence. This is why we are very careful, with the radio shows, to think of these as much more akin to a focus group-style analysis, than a survey.

What do you present to a development organization after you’ve analyzed the information?

We started as a research project at a university and that’s still part of our DNA. Our original approach has been to ask: “What is the research question?” Let’s do the interactive radio shows, bring back the data, analyze it, and write a research report. But, over time, we’ve realized there’s a lot to be gained through co-design, and a more collaborative exploration of data, and deriving insights in a collaborative way.

We are more interested now in thinking about what are the decisions that need to be made in a program, and when, for which this is going to be useful and valuable evidence. Then we design around that in terms of time frame, in terms of priority and evidence needs. Right from the onset, design it around what would be important to know to change a particular programming decision or policy decision. We involve our partners throughout the design of the radio show, the scripts, and the checking of the questions. They have a strong sense of how it is working and what it will do.

And that means that when we get to the stage where we’ve got the data, and develop the coding frames, we then have collaborative workshops where we explore the data together.

What is the significance of radio in the East African region?  Why did you choose this medium?

It began because of a phenomenon that interested me as a social researcher and a political scientist. Thirty years ago, in many parts of sub-Saharan Africa, as a general rule, landline telephone penetration was about 1 percent across the continent. In a lot of countries, the dominant broadcaster was the state broadcasting corporation, broadcasting in the official languages of the state. It usually put out quite a strong message, telling citizens how to think and behave, to put it crudely. That’s a legacy of colonial broadcasting. Then in the 1990s, liberalization occurred, in different degrees in terms of governance and democracy, but certainly in a lot of countries the media sector liberalized a lot. The affordances of broadcasting were much higher because the costs had gone down. There was a proliferation of FM radio stations and private, community, church radio stations across many countries. This meant that stations would tailor to particular localities, regions, and languages. The issues discussed were more local and meaningful to audiences.

In this region, what interested me was the new spaces of voice in public life and public discussion and the new quality they had about them. They were not all necessarily good, it was not all necessarily positive. This is also the world of vernacular radio that was implicated in post-election violence in Kenya. The growth of hate speech, of “shock jocks.” This is not all a rosy story. But it’s a story that matters in terms of how new spaces of public discussion are emerging and being shaped, and shaped by authority and dominant power.

Radio is a medium that is accessed, at some point or another, by 95 percent of the population. In most studies, it’s still very popular in spite of the growth of social media but especially for the kinds of inclusive governance questions that we are interested in. It remains a very relevant medium in large swaths of the continent. Not just rural, but urban contexts and periurban as well. It’s this power of interactive radio as a space of public discussion, not the one public and everyone all inclusive, but publics, that particularly interests me.

You have 2 free articles left
Log in or sign-up to unlock all of the free news on Devex.

About the author

  • Headshot sarajerving

    Sara Jerving

    Sara Jerving is Devex's East Africa Correspondent based in Nairobi. She is a reporter and producer, whose work has appeared in The Wall Street Journal, the Los Angeles Times, Vice News, Bloomberg Businessweek, The Nation magazine, among others. Sara holds a master's degree in business and economic reporting from Columbia University Graduate School of Journalism where she was a Lorana Sullivan fellow.