By Amelia Pittman
Mathematician and statistician John Tukey observed, “The greatest value of a picture is when it forces us to notice what we never expected to see.” Data visualizations can help us discover innovative solutions, support new initiatives, and raise awareness of issues few were aware of. What can we see in the data measuring the well-being of women and girls around the world? There may be gaps in the data, but enough data on women and girls exist for us to start creating visualizations to confirm problems we’ve suspected for years and to see new sides of challenges.
This blog post shares our initial experience in creating interactive online visualizations with Tableau software. Not only have we been able to use it to explore the variation of health indicators for women and girls, but we have also been able to better see the gaps remaining in the data.
The Ready to Measure (R2M) project identifies and curates data for 20 gender indicators that are ready to monitor the sustainable development goals (SDGs). These data, from many different sources, are available online through a data query tool. A quick glance reveals gaps, but these gaps just mean we need to work even harder to make each data point count. Data visualizations like the one below can make information and insights accessible when they might have otherwise been lost in the columns and rows of spreadsheets or tables.
Explore our interactive Tableau dashboard below showing insights on R2M health indicators for women and girls:
What we appreciate about the dashboard’s capabilities is the interactive visualization that summarizes the data and invites users to explore the elements most relevant to their interests. Users can adjust the dashboard to show data for R2M indicators on stunting prevalence, anemia prevalence, maternal mortality, and under-five mortality. Hovering the cursor over a country reveals a tooltip showing the country’s name, year of data, and the indicator value. Users can also analyze how an indicator’s values differ across country income groups using the bar chart to the right. The map showing the income group of each country allows users to compare the bar chart and the map showing the indicator values. Beneath this international overview of a selected indicator, users can examine the data for specific countries more closely. Selecting a country produces a bar chart showing data available between 2000 and 2016 for the previously selected indicator. The map reveals countries with no available data for any year, but the bar chart section allows users to explore the issue of data scarcity across all years within each country.
Take Ghana as an example. Like many countries, data availability by indicators. The bar charts below show the year-by-year values for three health indicators. In Ghana, the maternal mortality ratio has data for only 2007. This single year provides a helpful glimpse of the risk mothers face, but it’s a static picture that tells us nothing about whether things are getting better or not. Other estimates of maternal mortality in Ghana are available, but only the 2007 value has been accepted into the international databases used by the R2M project. Clearly more data – not just modeled estimates – are needed.
Data for girls’ under-five mortality rate are available for 2000, 2010, and 2015. These three years show us that the country is improving by reducing the rate over time. The prevalence of stunting (low height for age) in girls under 5 years of age is available for five years. It shows that from 2003 to 2014, the prevalence of stunting has dropped by half, from 32.7 to 16.8 percent. But if we had only found data for 2006 and 2008, we would have seen the prevalence of stunting rise slightly from 26.2 to 26.7 percent and possibly thought that stunting was slowly getting worse in Ghana. This shows that we need to be especially careful when working with sparse data as the full story is not being told.
The data gaps created some difficulties that we needed to tackle to create this dashboard. Tableau has online training videos and guides for all users of their Tableau Desktop software. In our case, we also received help from Hannah Jackson and Tammy Glazer, volunteers through Tableau Service Corps. This program is a volunteer network of Tableau employees who help non-profits around the world bring their data to life through visualization. Although we could use the online resources as a guide to many features in Tableau, they helped us quickly learn how to resolve one of our trickier issues: it’s simple to create a map showing data for 2016, but that left out countries with less recent data. To create a map showing data for as many countries as possible, we needed to show data for the “most recent year.” Our volunteers showed us how to create the calculated fields needed for this. And if anyone else has a similar problem with needing to show data for the “most recent year” in a data visualization, we would be happy to share the method developed by our volunteers.
Dashboards like the one above can serve as a diagnostic tool for countries to assess and plan for the modernization of their national statistical systems. They can also help policy-makers and donors identify gaps in our knowledge and evidence that must be filled. And as more users discover and explore these data, review observations that are available, and more importantly, those that are missing, the international open data community and national governments can work together to fill the remaining data gaps.
Developing capacity to create interactive online data visualizations allows us to communicate stories about the well-being of women and girls using R2M and other indicators from the SDGs. We can also create visualizations that explore the data gaps across indicators and within countries more directly. And beyond R2M, we hope to use such dashboards to invite more users to explore our Open Data Inventory (ODIN) which has information on the availability and openness of official statistics in 180 countries.
