Are you conducting mobile surveys responsibly?

Twitter card data responsibility

The use of mobile technology is a tremendous opportunity to better communicate with people in humanitarian settings. However, these advanced capabilities also involve new privacy and security risks for people in the communities where remote mobile surveys are implemented. We therefore collaborated with the International Data Responsibility Group and Leiden University’s Centre for Innovation to draft a practical guide: ‘Conducting Mobile Surveys Responsibly: A field book for WFP staff’.

The field book outlines the main risks for staff engaged in remote data collection and details best practices for data security, privacy and responsible data approaches in the very complex environments in which WFP operates.

data responsibility front page

VAM Talks: Episode 10

Logo2

Alice Clough interviews the VAM team in the Afghanistan country office about why they’ve started using mVAM’s remote food security monitoring system.

 

 

VAM Talks: Episode 9

Logo2VAM’s Arif Husain and Jean-Martin Bauer travel to New York for a hackathon run by Nielsen where the participants try to build a food security surveying chatbot.

Calling all developers: Join us at #Hackforhunger

hackathon-advert-pic

Have you ever wanted to help out the World Food Programme? Sign up for our Chatbot Hackathon!

When: Our partner Nielsen is holding a 24 hour ‘Hack for Hunger’ Chatbot Hackathon from Saturday, January 7th to on Sunday, January 8, 2017 at their global headquarters.  The Hackathon is sponsored by Nielsen, the world’s largest data and information company, and their data science subsidiary eXelate and Nielsen Marketing Cloud. You can sign up at this link

The Challenge:  Build an ‘emergency response chatbot’ to revolutionize how we get the information we need to respond during emergencies.  

The United Nations World Food Programme is the world’s largest humanitarian agency fighting hunger. When there is a crisis, we get the bags of food or cash assistance to people to make sure they don’t go hungry. We currently assist around 80 million people in 80 countries around the world. But to do this well, we need to know where people are that need the most help.

The chatbot you build will allow community members to report to WFP about food security conditions in their local area. This information can save lives after a disaster like Hurricane Matthew in Haiti where roads were destroyed and ports were closed for days. WFP can chat with community members and find out what is happening on the ground in order to get assistance to the areas that need it most.

Our Chief Economist Arif Husain will be at the hackathon to tell you more about the World Food Programme’s work in emergencies and technology’s potential to accelerate our response. For more info on  work we’ve been doing on chatbots with InStedd, read our blog: chatbot prototype

So, are you up for the challenge?

To Register: Sign up at this link – we’re looking forward to seeing you there!  Interested engineers or developers should be highly experienced in Java, Javascript, PHP, Ruby, Hadoop, SQL and development in Android, HTML, and iOS. UI/UX, Product Management, QA and HTML/CSS experience also welcome.

Got Maize? Using Free Basics to reach millions in Malawi

WFP/Gregory Barrow

We have said it before: open data is not really useful unless it’s also accessible to everyone. WFP maintains an extensive food price database that is accessed by a lot of people, but most of our visitors happen to be donors or agencies in North America and Europe. We feel that in order to achieve Zero Hunger, information needs to be accessible for people living in the most vulnerable geographies.

free-basics-app-exampleWe’ve already experimented with ways to share food price information with vulnerable communities using SMS and IVR in Somalia and DRC. Recently, an interesting opportunity came up: sharing our data through Facebook’s new internet platform Free Basics.  This initiative aims to provide the 2/3 of the world’s population who do not have internet with basic web access and is currently available in 53 countries. Certain websites with “basic” content like news, employment opportunities, health, education and local information are available for free with no data charges. We’re always interested in exploring how new technology could help in the fight against hunger, so this month, we began testing our first Free Basics website in Malawi.

 Malawi: Our first test for Free Basics.

Why did we choose Malawi? Along with other countries in southern Africa, Malawi was affected by a drought that has affected agricultural production and caused food prices to soar. In late 2015, WFP set up a phone based market monitoring system that helps us track food prices all over the country on a weekly basis. Current forecasts estimate that 6.5 million people won’t be able to meet their basic food requirements. Households at risk of food insecurity can spend anywhere from half to three quarters of their budget on food, so sharing the data we have about food prices with the population might help people make informed decisions about their food purchases.

Introducing ‘Za Pamsika’

We’ve been working with the Praekelt Foundation incubator to set up a Free Basics website that shares this weekly price data for all of the districts we get the data from. We’ve called it ‘Za Pamsika’ literally ‘things you can find in the market’ in Chichewa, the main language in Malawi. It’s a really simple website. You click on your region and district to find out food prices in your area. You can also compare prices at nearby markets if you’re in an area with many market options.

fb-home-screenfb-district-screenfb-market-screen
The great thing about Free Basics is that you don’t need a smartphone to access our data for free – just an internet-enabled one from a participating MNO. The project therefore has the potential to provide food price information to all Malawians who have mobile phones with internet browsing capabilities. It’s also not even necessary to have a Facebook account! While we know that by only contacting people who have internet-enabled phones, we may be missing the most vulnerable households.  But it will still provide useful information for a large section of the population. Essentially, we see it as a step in the right direction toward making our data accessible to everyone.

The site has now been live for 10 days, and we’ve started seeing some results coming in, both the number of visitors and the demographics. As we’d expected, most of the people who’ve visited so far have been male and under 25.

It’s great to see that we already have some users, but we still have to make sure more people are aware of the site and it’s useful for them. We’ll be heading to the markets covered in the site with the WFP Country Office team, to speak to people first hand about the site and learn how we can improve it. Ultimately, the feedback we receive from people on the ground will help us to evaluate Free Basics as a tool to share data about food with the communities we serve.

Is the road to hell paved with donated smartphones?

WFP/mVAM

WFP/mVAM

From time to time, people interested in mVAM will suggest we distribute cell phones to people who live in the vulnerable and food insecure communities where we work. It is usually a well-meaning idea, originating from specialists as they scope out a mobile-based data collection activity or a donor trying to help. After all, while phone
ownership has expanded exponentially, many poor people still do not have phones. It’s tempting to distribute devices to them as a solution to this disparity. However, our experience at mVAM and elsewhere has shown that we’re often better off going with the phones that people already have. This blog post explains why.

A checklist from UNICEF innovation

Others working with technology in developing countries have faced the same debates. We found that UNICEF innovation has come up with a useful list of questions to think about before distributing handsets or other hardware.

  1. Do you really need to distribute these devices?
  2. Who are they going to?
  3. How many of those people already have a smartphone? How about a dumb phone?
  4. How will you account for any loss of devices? Is that planned for?
  5. How will they be fixed when they break?
  6. How will you make sure devices are charged?
  7. What infrastructure do they connect to for information flow? Internet?
  8. Who trains people in how to use them?
  9. How does this affect the local economy? What is the market distortion on local hardware sellers? How many vendors might get put out of business?
  10. How do you make sure you don’t create a false incentive for the future?

An addendum from mVAM

On the basis of our experience, we’d like to add a few additional considerations to this list. mVAM’s own experience with cell phone distribution has been mixed. In 2012, we provided 400 cell phones to IDPs in eastern DR Congo, an idea we picked up from the literature (e.g. Listening to Dar). We tried to do the right thing by consulting with the community first. We offered low-end feature phones. We set up a solar charging station in the camp where people could recharge their phones for free. We also ensured that people received training in using the new devices.

On the one hand, we were pleased to see that we obtained a good response rate to our surveys from the camp and that providing access to cell phone technology empowered people (because the cell phones we provided allowed displaced people to call home, use mobile money for remittances, and obtain information). On the other hand, there unfortunately was also theft (42 phones were reported stolen!) and even cases of people being attacked for their devices. These findings are captured in the independent review of the mVAM activity that was published in 2015. Due to these concerns, we have not provided cell phones to people in other settings as mVAM has expanded.

So, we want to add another set of considerations to the UNICEF list :

  1. Are we putting people at risk ? mVAM surveys are sometimes carried out in very vulnerable conflict-affected communities. By receiving phones, people can be put at risk.
  2. Are there specific risks to women, the elderly, and the disabled? These groups are at specific risk of physical abuse.
  3. How would you replace a lost or stolen phone?
  4. Is providing phones sustainable? Providing and replacing phones in the long term can quickly become a financial burden, especially if people are in remote areas. When the population is concentrated (e.g. a camp or city), costs are lower.
  5. How could local regulations on cell phones affect your project? SIM cards are becoming highly regulated with requirements to provide an ID to authorities. This can be a barrier for some groups such as refugees. Some SIM cards will be cancelled if they are not ‘registered’ after a few weeks.
  6. What are the alternatives to providing phones? In some communities, poor and vulnerable households already have phones or have access to shared phones. It’s also important to remember that mobile data collection is not appropriate in all settings: sometimes, conventional tools such as face to face surveys are a better choice.

As we have gained more experience working in different contexts, we have concluded that we’re better off working with the phones people already have. Sometimes, that means designing our data collection strategy around the information that can be credibly collected when phone ownership or network coverage is limited. For instance, in places where cell phone penetration is low (~20%), we have not attempted to run a representative survey of households but have focused instead on obtaining information from a set of key informants, as we did in Afghanistan or the Central African Republic.

WFP Yemen M&E: Reaching Beneficiaries During Widespread Conflict

As mVAM has been expanding we’ve started to see the remote technology used in other areas of WFP’s work. This week’s blog is from Katy Huang who works for the Yemen Country Office. She shares how the M&E unit is using remote live calls to get feedback from beneficiaries about WFP’s assistance. 

yemen-blog-pic-1

WFP/-Asmaa Waguih

Before joining WFP Yemen’s Monitoring and Evaluation (M&E) unit 8 month ago, I worked as a researcher for the New York City Health Department. As I love the creative process of collecting, analyzing and reporting data, I was excited for the opportunity to manage our unit’s “remote M&E” (rM&E) system. Currently, Yemen’s Emergency Operation assists about 3 million beneficiaries a month. Our rM&E system uses an third country call center to conduct phone surveys with beneficiaries post-distribution to hear about their experiences receiving and using the assistance. The center completes about 2000-2400 surveys per month.

Before establishing the rM&E system in September 2015, we learned in a previous post-distribution monitoring survey that a large majority of our emergency in-kind food beneficiaries owned a mobile phone or had access to a friend or neighbor’s mobile. We also found out that a large majority of mobile owners were able to charge their mobiles on a regular basis. This information meant that conducting mobile surveys proved to be ideal within the context of Yemen’s ongoing and widespread conflict as it allowed us to reach large numbers of beneficiaries without compromising the safety of field monitors. Other benefits of using rM&E include it’s relative low cost and being we can reach beneficiaries in all the governorates where we offer assistance. Also, in the 15-20 minutes it takes to complete a survey, we have been able to collect all the key process and food security outcome indicators that we also collect in our longer bi-annual face-to-face post-distribution monitoring surveys. Ultimately, rM&E complements other M&E systems (i.e., on-site distribution monitoring and beneficiary hotline) to triangulate and confirm findings.

yemen-blog-pic-2

WFP/Asmaa Waguih

Although there are many benefits to using rM&E, challenges do exist. Bias is the main issue as data collected by rM&E tends to be more biased than data collected face-to-face. Some of the biases we face relate to the following:

  • Sampling frame bias: We don’t have the entire list of mobile numbers of beneficiaries for random calling. The amount of mobile numbers we receive depends on what cooperating partners collect from beneficiaries at the time of food distribution. We have had to regularly remind cooperating partners about the importance of sending us these mobile lists. In addition, some beneficiaries don’t own a mobile phone and they may have different characteristics, such as being more poor or vulnerable, than those that do own mobiles.
  • Gender bias: The frequency of female respondents for rM&E (about 5 percent) are lower than that of face-to-face (about 10 percent). This may be due to more males than females owning mobile phones. To try to address this, the call center recently hired more female enumerators to engage female beneficiaries to respond.

Despite these biases, the amount and quality of data we have been able to collect on a monthly basis have been invaluable. The large sample size has allowed us to report nationally representative data and to disaggregate data by activity type (i.e., in-kind, voucher) or demographics (i.e., displacement, gender). With regular monitoring, we are able to see trends and compare results over months and quarters. To see how we used this data for reporting, please see our Yemen M&E Quarter 1 2016 report.


For more information on mVAM’s work in Yemen, please visit the mVAM Yemen site.

VAM Talks: Episode 8

Logo2Alice Clough interviews Gideon Shimshon from Leiden University’s Centre of Innovation about collaborating with WFP and how using Big Data analysis can help promote food security.

From the Rift Valley to Silicon Valley

screen-shot-2016-09-30-at-13-18-47

Private sector sponsors have played a large part in helping us use mobile technology to reach out to people in the world’s most vulnerable communities. Since mVAM began in 2012, we have received support from The Nielsen Corporation, Cisco CSR, and Google.org. We’re also collaborating with Tableau Foundation and Praekelt Foundation/Facebook. These types of partnerships are pretty new at WFP so we think it’s important to explain how working with the private sector differs from the usual way of doing things.

Some of our sponsors have helped us by writing a check, providing us with the vital financial resources we needed to achieve our goals. This was the case of Google.org when the Ebola epidemic occurred in 2014-2015.  Their support helped us scale up SMS data collection at the height of a major emergency, an essential proof-of-concept  that allowed us to scale up the approach in other complex settings such as Iraq, Yemen and Syria. Cisco CSR financed the development of the open-source SMS and IVR tools that we use to communicate with communities in Somalia, DRC and other places.

Other partners have provided us with their skills rather than cash. For instance, Nielsen allowed us to spend some time with their data science staff, and we greatly benefited from their expertise in survey methodology.  It turned out that lessons and insights from media monitoring in the US are entirely relevant to food security monitoring in humanitarian contexts. Nielsen’s knowledge of incentives, retention and engagements helped us keep our respondents engaged. Importantly, the collaboration was win-win, and both WFP and Nielsen’s staff were able to learn about the new techniques that are revolutionising data collection.

Similarly, Tableau’s specialists have worked with us to to design data viz tools that help people better interact with our data. This is a very different level of engagement than the relationships we have with ‘traditional’ donors to the UN – usually governments – who tend not to get involved hands-on with our data.

screen-shot-2016-09-30-at-13-22-52

We can hear you sigh – are these not examples we hear about in the media of ill-intentioned private companies using UN data for their own ends?  Our experience has been that they genuinely want to help us better assist people in need. In some cases, private partners have approached us first when a disaster occurs to offer their help. Though, of course, it seems that it’s much easier for companies to offer assistance for a natural disaster (Ebola in 2014, Nepal earthquakes in 2015) rather than in conflict and complex emergency settings.

So, how do they compare to government donors?

As you can imagine, working with private companies is different than with traditional donors. Private partners tend to require less paperwork than government donors commonly require. Proposals are shorter, to the point, and decisions are made rapidly. What private donors look for is a compelling use case and a good match with the skills they offer, such as analytics or connectivity.

But what’s great about these private sector partnerships is that they have allowed us to grow from ‘startup’ to a fully fledged mainstream project. Since last year traditional government donors, including Belgium and the Netherlands, have come on board for substantial multi-year support for mVAM. This suggests private and public funding are complementary — with the private sector helping an idea emerge before it is taken to scale with more traditional resources.

praekelt-logo

Our latest private sector collaboration is with Praekelt Foundation to use Facebook’s Free Basics internet service. This is going to be a great addition to our two-way communication system as we’re going to be able to share the price data we collect via mobile phones through their free internet.The partnership is currently in its early stages, so watch out for dedicated entries once we have our test version!

Hacking Hunger Episode 11: pinpointing hunger with mobile phones

In the latest of their ‘hacking hunger’ podcast series, WFP USA’s M.J. Altman talks to Jean-Martin Bauer about  how mobile phones in the hardest-to-reach corners of our world are changing how we understand and fight hunger.

In the podcast Bauer discusses how his 12 years as a humanitarian worker stationed in West Africa inspired him to think about using mobile phones to gather food security data. He touches on using mVAM methods in emergencies like Iraq and tackling gender disparities in mobile phone use, and also discusses how mVAM shares the data it collects with the communities it serves as well as wider humanitarian community.

Check out the podcast below, or subscribe to “Hacking Hunger” on iTunes, Stitcher, Soundcloud, and TuneIn Radio.