Real time journey of quality monitoring
By Pratik More, LGT Impact Fellow for Educate Girls, India
Data for Development
Synonyms for data are facts, figures, statistics, details, particulars, specifics, features, information, evidence, intelligence, material, background, input, proof and so many (1). All synonyms appropriately define a data and data analyst is a person who gathers, cleans and studies the data set to solve the problem (1). These problems are real world problems which affect the quality of life of individuals as well as the overall ecosystem. In the case of the social/ development sector we can consider example of challenges mentioned in 17 Sustainable Development Goals (SDG’s). Use of data will help to more agile, efficient and evidence-based decision making to solve these problems (2).
Journey Towards Fellowship
My journey in the development sector started in 2017 from my alma mater Tata Institute of Social Sciences, Mumbai. During my masters I spent significant time in urban slums, understanding the double burden of disease in slum dwellers. All my observations of study I presented in a graphical format, back then I was unaware of the different analysis and visualization techniques. When a data company came for our placement I felt this is the thing I would like to do in my life and I started my data analyst career with that digital health company.
While working in a data company in public health domain, I realized that I have to get more knowledge on the technical skills starting with data collection to data visualization. As I don’t have any background knowledge of coding, I was able to explore limited possibilities in this field. I wanted to get that technical knowledge and wish to learn the M & E framework. The opportunity came for me in the form of LGT Impact Fellowship. I was selected as a data analyst in the organisation Educate Girls.
About Educate Girls (EG)
It is a non-profit organisation that focuses on community mobilization for girl’s education in India’s rural and educationally backward areas. With different interventions EG helps to identify, enroll and retain out of school girls and to improve foundation skills in literacy and numeracy for all children (3). EG currently operates in 20,000+villages in UP, MP and Rajasthan. EG Impact team works for tracking programme implementation, measures progress on key outcomes and provides actionable insights on ways of improvement. My roles in EG are dashboard building, automation of data collection, stakeholder reporting, identifying different ways to maximize the impact, documentation of process and activities, increasing efficiency and quality of data analysis.
Growing With EG
Within a month of starting my fellowship I got the opportunity to go to a field in Udaipur. Data is very powerful to understand and solve the problem, but the human touch is needed to understand the data better. When the whole world is running behind AI/ML my heart always shouts out to me and says human intelligence and relations can change the dynamics of artificial intelligence. In the realm of data, it is important to understand the demography to analyze the data. My field visit experience is helping me day by day to understand the kind of work EG is doing on field and to understand the quality measures we take to keep our data accurate, complete, consistent, timely, valid and unique.
I already worked in the real-time monitoring domain but I’m satisfied with the work EG does on the field. With use of appropriate tools and technology EG is able to overcome the issues with data syncing, system generated data duplicity, avoided human errors with appropriate validation, easy to access and timely MIS reports. With some of the freeware tools like Google Data Studio I can directly create a dashboard without depending on coding skills. As an organisation we are trying to build our strength by developing machine learning models internally to interpret our data in a better manner. As a part of it we have developed a Fuzzy Matching algorithm to identify phonic errors to get matching names in different surveys. In door to door surveys most of the time people in India use different honorifics to call a person and when different types of names are listed for the same person in different surveys those records become duplicate records. To omit these duplicate records our quality monitoring team has to do a laborious task, with our python based Fuzzy Matching algorithm within some minutes we can remove unwanted records from millions of records.
My fellowship experience till now is very enriching. I’m trying to get command of multiple tools like SurveyCTO for data collection, Python and advanced excel for data analysis, Google data studio for data visualization, M&E tools and techniques. I’m also working on the data management strategy for EG where we use our data in an efficient manner. Outcome of work with my colleague is that we are able to successfully build a COVID research tool for data collection using SurveyCTO. This tool is used to identify the pandemic effect on the life of a girl child in rural India and how it may affect her return to school. I was able to create a Praveshotsav (school welcoming celebration) dashboard for the government stakeholders which is being used for behavioural, social and economic transformation of girls to access quality education through quality monitoring of different interventions in different geographies. By documenting EG potentials and prospects for data management strategy we are able to identify some of the gaps in the processes and technology part, we will be working more to bridge these gaps. I have reached mid-term of my fellowship, many more things I want to contribute to my team and to EG.
Summary
In the context of an educational NGO, if data is used appropriately it will help to understand the problems better and will help to give practical solutions to solve the problems. Social, behavioural, literacy related problems can be solved with community interventions with a staff and community volunteers. For efficient programme delivery data is used to train the staff and to bridge the gap in their understanding for better programme outcomes. 60% of occupations have at least 30% of activities that automatable (4). We want our employees to work high value work, low value work will be automated. To reach our goal of 15 million children accessing quality education3, predictive modelling will be used to study available data and will be used to improve the existing programmes and to value add to them. In this blog I purposefully avoided mentioning COVID-19 but it has long term repercussions on day to day life., We have lost almost 5.6 million lives till now5. More painful thing was we were helpless to help those people who were dying without ventilators and required instruments. Data will help us to understand the gaps and appropriate ways to bridge the gap. Data is a key to creating an impact and I’m happy to be part of Impact.
References:
- Simpson, J. A., Weiner, E. S. C., & Oxford University Press. (1989). The Oxford English Dictionary. Oxford: Clarendon Press.
- [1.1][1.2][1.3] The United Nations. (2015). Transforming our world: the 2030 Agenda for Sustainable Development
- Educate Girls 2022, < https://www.educategirls.ngo/Who-We-Are.aspx#about-us>
- Driving impact at scale from automation and AI. (2019). McKinsey
- WHO 2022, < https://covid19.who.int/>