Lessons learned from Vital Signs Engagement with the Data Science for Social Good Program

  • October 18, 2017
  • Posted by: Tabby Njunge


Vital Signs successfully participated in the summer 2017 Data Science for Social Good program at the eScience Institute at the University of Washington.  During the program, Vital Signs project leads worked with two data scientists and four project fellows to answer several questions related to sustainable agricultural development.  Since the beginning of the Vital Signs Program, a large dataset had been amassed which has led to significant insights and publications around themes like agricultural intensification, yields resilience, natural resource gathering, and soil nutrient mapping (Hengl et al., 2017; Kanter et al., 2016; Nijbroek and Andelman, 2016). However, many other subsections of the Vital Signs rich dataset remained under-utilized and under-explored. Themes like water availability, nutrition, crop commercialization, and the role of education have significant data available for study with limited insights generated so far.

During the 10 weeks program the Vital Signs team sought to answer policy relevant questions pertinent to decision makers in the Vital Signs countries. The questions were further refined to ensure they could be answered using the data available, reflected fellows’ skill set and could be achieved within existing temporal constraints.

 In the end, DSSG helped us answer the following questions:

1. Are the effects of extension services on crop productivity moderated by farmers’

educational attainment?

2. How does intensification of agricultural practices relate to landscape-level distribution

of resources and equitable outcomes?

3. What is the effect of household natural food collection on household food


4. How does crop commodification affect food security and childhood nutrition?

5. How does household use of different sources of water relate to the odds of

experiencing water insecurity events?

6. What are the households’ per capita water use in the dry season compared to

WHO-recommended standards, and how do they vary by country, water source, and

gender of the household head?

Many of these questions have had more detailed analyses already posted on the vital signs blog.  In brief, many insights and lessons learned included the fact that: the effects of extension services on crop productivity are not moderated by farmers’ education attainment; households that collected more natural food make fewer food expenditures; crop commodification improves food security but has no direct effect on nutrition outcomes; and households in the four countries use significantly less water than the WHO recommends.  These results have since been shared within Vital Signs partners and generated some potential policy implications as detailed in the blog here

Aside from blog posts, these results have been compiled into a policy brief (download your copy here)for policymakers to read and have their policy informed.  Overall, it was a very useful endeavor, both for the DSSG fellows who volunteered their time and gained valuable experience and for Vital Signs and it’s stakeholders, who now have more valid analyses derived from data collected in the field. Similar programs in collaboration with local universities would be useful in building in country data science capacity and advancing the data driven decision making agenda in each of the countries. Vital Signs is thus seeking and working to establish these partnerships.



photo credits:Robin.E.Brooks,E-science Insitute,University of Washington

Back to blog listing