Big Data can be a valuable tool in helping those of us in the social sector reach our analysis goals. Used well, some types of Big Data can answer critical questions for governments and nonprofits, and help us gain insights that wouldn’t otherwise be available. In short, there is a powerful argument to be made in favour of nonprofit Big Data projects.
But as I discussed in my last post on nonprofits using Big Data, the social sector can’t just apply the power of Big Data to every problem we need solved.
So when do nonprofit Big Data projects make sense?
Big Data is best leveraged by social sector organizations to answer two types of questions:
- Monitoring (or process) questions, or
- Predictive questions
What Kind of Question are You Asking?
Much of social sector analysis focuses on three types of questions organizations are trying to answer. Big Data is an invaluable tool for two; not so useful for the third.
- Process and Descriptive
These questions include the monitoring aspect of M&E and generally, try to answer the question “what is happening here?” Monitoring the migration of birds, conducting a sentiment analysis on Twitter, or tracking the spread of a disease — these are all process questions.
This type of question is usually asked for project management or intervention purposes and can be helpful when determining where and when to deploy resources. Big Data is very helpful in answering this type of question.
Predictive questions are exactly what they sound like — questions we use to figure out what is most likely to happen next. What is the most probable result of a severe drought in a specific country? Where will the next outbreak occur during an epidemic? Who is most likely to commit a crime or pay their rent on time? These questions are vital to making efficient and effective decisions.
Predictive modelling is where most corporate Big Data work is done. For example, when Netflix recommends a movie or show to you, it is predicting what you will like based on Big Data analysis. In many circumstances, nonprofit Big Data projects can help answer these questions, although occasionally, Big Data can do more harm than good. (The distinction, as Cathy O’Neil reminds us, between harm and good lies largely in the quality of the data itself and the equity of the algorithm being used for predictions.
Causal questions are also more commonly called impact analysis. They’re asking, “Why did this change happen?” Impact analysis is not a place for nonprofit Big Data projects, for a number of reasons.
Big Data is usually collected using methods that are convenient, rather than designed. This can result in selection bias, which prevents the data from being able to answer causal questions. If data is biased, having more of it does not decrease the bias; it only makes that bias harder to understand — which is much worse.
Nonprofit Big Data Projects That Work
Provided we confine our use of Big Data to descriptive and predictive questions, it can be very helpful in addressing three key problems often present in social sector data:
- Lack of data volume
- Lack of data source variety
- Lack of current data
Big Data provides an often innovative source of data to help triangulate more traditional data. And, by definition, there’s lots of it — meaning it provides higher volumes of information. Additionally, Big Data is often collected in real time (or close to it), so managers, researchers, and frontline workers can have a more immediate, more nuanced understanding of what’s happening — and respond appropriately.
Here are a few great examples of nonprofit Big Data projects that are making a difference.
Weather Forecasting and Refugee Migration
The United Nations Office of the High Commissioner for Refugees is using weather data to predict where and how harsh winter weather conditions might affect refugee settlements, allowing them to forecast which populations are most likely to move, how many might need assistance, and where they might go.
“In late January… below-normal temperatures plunging to lows of -5 to -15 C, accompanied by heavy snowfall, are forecast for the southern and eastern Balkan Peninsula, Turkey, the Eastern Mediterranean and the Middle East. Across this vast area, millions of refugees — most from Syria and Iraq — live in camps and settlements. Thousands of other people are on the move, many on foot or in flimsy, open boats.”
Measuring Poverty With Machine Roof Counting
Under the United Nations Global Pulse initiative, Pulse Lab Kampala is using satellite imagery in an effort to monitor poverty in Uganda. Because traditional thatched roofs can harbor pests and disease (not to mention the amount of maintenance they require!), families will upgrade to metal or tiled roofing as soon as they can afford to. By teaching computers to differentiate between different roof types, the team can accurately monitor poverty levels across the country without the need for door-to-door surveys.
“As the Government of Uganda, we are now more convinced than ever before that development challenges need to be addressed with data innovation.”
Dr. Ruhakana Rugunda, Prime Minister of Uganda
NASA Data to Support Agricultural Development
Here at Datassist, we’ve partnered with Farm Shop, a Kenya-based agricultural input franchisee, to use data from NASA satellites to conduct analyses that support the roll-out of location-specific agricultural products, training, and support.
“The program aims to support tens of thousands of small farms — many run by women — providing sustainable jobs and income for farmers, labourers, and their families as well as a reliable source of healthy, fresh food for the people of rural Kenya.”
Need Help With Your Data?
There are plenty of examples of Big Data being used in awesome ways to support social initiatives. Unfortunately, there are just as many examples of projects where Big Data is being misused — either naively, accidentally, or intentionally — to misrepresent reality; these projects harm, rather than help, in the social sector. We’ll examine some examples of nonprofit Big Data projects that shouldn’t be in my next post.
If you’re part of a nonprofit or social sector organization that would like to leverage the power of Big Data — or any data, really — Datassist is here to support you. Our team of experts can assist you in collecting, analyzing, and presenting your data in a way that educates and engages. Get in touch with us today.