Tags: Data capture and analysis

Good data is essential to ensure well-targeted, effective development policies and programmes. Today, innovation in how we gather, analyse and use data is creating unparalleled opportunities to gain insights and information in support of development and poverty reduction. UN Development Operations Coordination Office has been supporting UN country teams around the world to begin to harness the power of these new ways of working with data.

Why data innovation?

Traditional ways of gathering development data have focused on surveys and evaluations. These are essential, but are expensive and often infrequent. To drive real-time decision-making, data gathering needs to be more frequent and responsive to changing needs.

New forms of data can generate insights into areas that have been traditionally difficult to measure, and population groups for whom little official data exists. They can support planning by highlighting new trends and changes ahead of the curve. And they can help the UN move towards real-time decision making to target resources where they are needed most – for example during fast-moving emergency situations. Open data can contribute to a culture of greater transparency, and enables many different organizations to generate analysis and insight.

What we’ve done

UN teams have been putting in place more frequent data collection to help programmes to respond more rapidly to relevant events and changes. The data allows the UN and governments to quickly identify bottlenecks and take corrective action as required. In Mauritania, for example, the UN team has created a “real-time” system for health workers to upload data about maternal deaths and their drug inventory using their smartphones. This will help to identify patterns in maternal mortality, and to quickly target responses and resources accordingly.

Another area of innovation is the mining of new sources of data. This can include data from populations and channels that are under-represented in traditional data collection. In Uganda, the UN team has been mining data from local radio call-in shows as an untapped source of data about local people’s experiences and opinions of current events.

Bringing together and opening up data from different sources can create powerful insights into development problems. In Albania, the team is bringing together administrative data from the police department, the tax office and the education system with survey data on violence against women and children. They are blending data sources to create a fuller picture to make Tirana a safer city.

‘A-ha’ moments

  • It is essential to maintain a clear focus on the ‘why’ of your data innovation project. The range of technical capabilities available can be a distraction, leading people to try to do too many different things instead of focusing on “what can we do now?”, and “what will we do differently once we get the new data?”
  • Visualizing data is key to building confidence and buy-in. Enabling users to see and manipulate the outputs of data projects builds their engagement and understanding of why data innovation is essential.
  • We still need traditional statistics for a complete picture. Not all interventions can be monitored in real time, and real time monitoring with new sources of data cannot replace traditional statistical methods. Taken together, old and new forms of data provide the fullest picture.

What’s next?

Experiences from joint UN data innovations illustrate that collecting and analysing data is easier than getting people to use it for policy decisions. Simply having better data does not necessarily ensure that it is being used effectively: the biggest challenge remains understanding how we can use data to influence decision making and improve development outcomes.

Likewise, UN teams around the world need to ground data innovations in public accountability and build partnerships with civil society. Our goal with data innovation should not just be about extracting data to inform decisions. It must also focus on engaging citizens to use and interpret data as an instrument for accountability.

The collection of new data on an unprecedented scale brings risks as well as potential rewards. A commitment to human rights, data protection and privacy principles forms a key element of our approach to data innovation.

Related Blogs and Country stories

Silo Fighters Blog

Crowdsourcing the campfire: how our data visualization contest opened doors

BY Abigail Taylor-Jones | November 14, 2018

“Visualizations act as a campfire around which we gather to tell stories.” - Al Shalloway, founder and CEO of Net Objectives. Telling our story well is key to ensuring we can influence policy and other key decision-making processes. In order to do so, it is important to get new insights from the evidence we generate from the data we collect. To give a sense of the scale, we collect data from 130 UN Country Teams, serving 165 countries. The types of data we collect ranges from operational data, socio-economic data, financial data, data on coordination and results. Sitting behind the walls of the UN can sometimes be lonely ploughing through all this data (other times it is quite daunting). So, we have to think of creative ways to gather new insights to tell a good and compelling story. The UN is known as an organization that brings people together globally to participate in various ways, for example working towards realizing the goals set for 2030 Agenda. For us, being open and inclusive about the UN’s work is always at the forefront of our minds, even when it comes to data. We started thinking about ways to include others from outside the UN in our analysis and data visualization process. As the Secretariat to the UN Sustainable Development Group (UNSDG) we have access to a wide range of data, so we thought, why not launch our first ever UNSDG data visualization contest and find out what others can see in our data? So, my colleague Kana Kudo and I did just that. In collaboration with Tableau, we launched the contest and invited data scientists and anyone interested in data visualization to use our data from the UNSDG portal, which pulls UN specific data, published by several agencies, using the International Aid Transparency Initiative (IATI) standards to report how the UN is contributing to the global development agenda. Data is powerful but we don’t always know how to tell a story with it After launching the contest, we realized there were blind spots that we failed to see. For example, some of the submissions did make use of the IATI data sets, while others did not. The guidelines we provided were clear, however the research questions were a little unclear. We ended up receiving several stunning visualizations, but they were not exactly what we were looking for. We learned that when it comes to data, it’s best to be specific. Another learning was that data scientists wanted the option to work with other data visualization tools and not be limited to Tableau; so we had to broaden the scope of tools for the contest. We brought a selection panel together to assess the submissions, and we selected two winners. The first winner crafted “Visualizing Malaria: The Killer Disease Killing Africa,” an impactful visualization that analyses malaria deaths in the world, how they have changed, and how funding has evolved over the years, particularly in Africa. The contestant explained that she had been inspired by the experience of a dear friend who had been infected with malaria. We also liked this visualization on malaria because it focused on both the positive and negative aspects of the fight against this diseases. Whilst lives are been saved through the use of mosquito nets, there’s also a downward trend in other aspects, which means more still needs to be done. [caption id="attachment_10399" align="alignnone" width="542"] Visualization by Rosebud Anwuri[/caption] The second data visualization titled “Leave no one Behind”, included the UN’s spending on each Sustainable Development Goal (SDG) per country, looking at the financial distribution among the SDGs. The underlying calculations were just as impressive as the visualization itself! We liked this visual and we were interested in how the participant highlighted the leaving no one behind aspect, which is the central promise of the 2030 Agenda; and an overarching programming principle. Looking at how we are doing from a financial expenditure perspective is key to assessing the UN’s contribution to the SDGs. Behind the scenes, our team in Headquarters was tinkering with developing UN Info, a tool that integrates the UN contributions to the SDGs and the 2030 Agenda. This is an important aspect because it keeps us accountable and helps UN Country Teams with programme management. From this contest, it was clear to us that data is obviously powerful but we don’t always know how to tell a story out of it. We were very impressed with the contestants’ interpretations and the visualizations. As a bonus, we also gained unexpected and useful insights that helped us refine our UN IATI data set.   [caption id="attachment_10400" align="alignnone" width="570"] Visualization by Pedro Fontoura[/caption] One of the things that we also discovered, is that data scientists like to get involved. Chloe Tseng, founder of Viz for Social Good contacted us to find out how she could collaborate with us. Although she didn’t participate in the contest, we were keen to work with Chloe and her team of volunteers just as she was to work with us. Goal 17 of the SDGs relates to partnerships and we know how important it is work with others to realize our goals. We gave Viz for Social Good a particular set of data related to the partnerships that the Country Teams have beyond the UN. If you haven’t read Viz for Social Good’s journey working with us, and the beautiful visualizations that came out of our partnership, check it out here. Our data was too fat! The contest was a great learning opportunity for us. From our collaboration with Chloe and the Viz For Social Good network of over 2000 data visualization experts, we learned that our data is good but we need to look at ways of improving the way data is parsed through our systems and ensure that it is formatted in a manageable and easy way for data scientists to work with it. Chloe also gave us feedback on moving from larger chunks of data to smaller chunks. We took these recommendations very seriously and have made significant changes in our data systems for optimum use by data scientists. We trimmed down our data in smaller chunks that requires little time for data cleaning which allows for quicker analysis. This experience was definitely an eye opener in terms of telling a more powerful and compelling story than we will ever be able to do if we stick to large sets of data in an excel format. The campfire is still with us Collaborating with Viz For Social Good and with the contest participants inspired our team to adapt our digital strategy work.  Seeing the way these artists take data and communicate with it opened our eyes. Our taste has changed and boy have our standards gotten higher. We are designing dashboards for future projects and seeing the artistry has upped our game for the long run.   Photo: Wenni Zhou

Silo Fighters Blog

Mining alternative data: What national health insurance data reveals about diabetes in the Maldives

BY Yuko Oaku | November 7, 2018

An island nation consisting of 1,190 small islands, the Maldives is clustered around 26 ring-like atolls spread across 90,000 square kilometers. For many centuries, the Maldivian economy was entirely based on fishing. Tuna is one of the essential ingredients in the traditional dishes of the archipelago. But between 1980 and 2013, the GDP per capita increased from $275 to $6,666 due to the success of the high-end tourism sector. With the rapid economic growth and a wave of globalization, there have also been changes in the dietary preferences and lifestyles of Maldivians. A staggering 30 percent of the Maldivians are overweight due to unhealthy diets and lack of physical activity, according to data from the Global Health Observatory.   Consuming sugary beverages is also a big problem among Maldivian youth and young adults. According to a study by the World Health Organization, in 2015, 4.7 million litres of energy drinks were imported to the Maldives, which is a very high volume for such a small population (around 410,000 people live in the Maldives). These unhealthy habits are drivers for the increase in non-communicable diseases, such as cardiovascular, cerebrovascular and hypertensive disease.  These diseases are the main causes of death among Maldivians. According to the National Health Statistics from 2014, diabetes is ranked as the ninth overall cause of death in the Maldives. [caption id="attachment_10393" align="alignnone" width="450"] "Drinking energy drinks is not cool" Health Protection Agency Maldives[/caption] Analyzing the prevalence of Type II diabetes with Insurance Data All Maldivian nationals are covered under the Government’s universal health insurance plan called “Aasandha”. Since it began its services in 2012, the plan gives full coverage to all health services from most health care providers and up to a certain amount for some of the private health care providers. The plan also covers care in affiliated hospitals in neighboring India and Sri Lanka in case the treatment is not available in the Maldives. Aasandha data provides personal data records and insurance data for all Maldivians. Since the usual data source for non-communicable diseases is the Demographic and Health Surveys, which is carried out every 6 years (most recently in 2015 and before that in 2009), we thought we could get more up-to-date data on diabetes if we looked directly at the health insurance data. Our team assumed that analyzing this data would serve as proxy indicators for the SDG indicators 3.8.1: Coverage of essential health services. Initially, this indicator was labeled as Tier 3 indicator, meaning that no internationally established methodology or standards were yet available for the indicator. As of 11 May 2018, however, 3.8.1 has been upgraded to Tier 2 indicator, which means that the indicator is conceptually clear, has an internationally established methodology and standards are available, but data are not regularly produced by countries. Our idea was to have an anonymized look at the data from the universal health insurance plan to see what else we could learn about non-communicable diseases. We at the UN Country Team in the Maldives, UNDP and WHO, partnered with the Maldives National University (MNU) research team and with the National Social Protection Agency (NSPA), the custodian of Aasandha service in the Maldives. What we found out about Type II diabetes in the Maldives: We dug into the anonymized health care records for 2016, including information about: 1) what diseases the Aasandha coverage is used for 2) the cost 3) where the medical procedures take place Together with the research team, we decided to focus on Type II diabetes for the scope of this study. We found some interesting facts about the prevalence of Type II diabetes in the Maldives: More than 3 out of every 5 people who have diabetes are women. The mean age of patients with Type II diabetes is 57, while the youngest age is 13. Females get diagnosed with Type II diabetes at a younger age compared to men and there is a relationship with gestational diabetes. Of those seeking care, 79 percent of the people go to private health care providers, whereas only 21 percent seek services from public health care providers. We also discovered that the Aasandha data was also incomplete. For instance, there were missing records from some of the largest regional hospitals in most populated atolls in the country. This may suggest that data from government hospitals are not entered into the system because patients don’t need to make a claim for the payment, whereas in private hospitals, the data is needed to allow patients to make a claim for their payment. It could be that more people are using public health care providers, but since the data is not entered into the Aasandha system,this information is unavailable to us. [caption id="attachment_10395" align="alignnone" width="393"] WHO Maldives[/caption] Next frontiers in proof of concept for alternative data With this pilot study we found some interesting facts about the prevalence of Type II diabetes in the Maldives as well as some possible data gaps in the Aasandha insurance data. We will be sharing our findings and challenges of using Aasandha data with the members of the UN Country Team as well as relevant ministries and agencies, including the Ministry of Health and the National Social Protection Agency. Reflecting on this pilot study, we will continue to support the country to explore alternative sources of data that will enable us to track more SDG indicators in the Maldives. According to an internal assessment done on data availability for all SDG indicators by the National Bureau of Statistics, there’s currently no mechanism for data generation for 56 indicators and for another 51 indicators, additional efforts will be required to make the data available. With all this data missing, we’ll need to tap into additional resources to make the data available because if we don’t know where the Maldives stands on Sustainable Development indicators, it’ll be hard to plan to achieve them. There is definitely a need for new data sources and having this data gap in mind, we have another pilot project in the works that’s going to use call detail records data to track population mobility to the urban centers of Male. Stay tuned for more in our work mining alternative data sources for the Maldives!

Related Resources