In the world of data, the pace of innovations and technological advancement over the past decade has been unprecedented. Cloud storage, open data, smartphones, drone imaging, data visualization, social media: these innovations and many others are presenting statisticians with promising opportunities for increasing the efficiency and efficacy of data collection, dissemination, and use. At the same time, the vast number of such innovations, a lack of capacity to use them effectively, and being unsure of what is a fad and what is sustainable, presents serious challenges.
As a component of the project Informing a Data Revolution, PARIS21 hosted a half-day seminar on 4 April 2014 “From Data to Policy Action in Low-Income Countries: How Can Innovation Help?” Held as part of the PARIS21 Annual Meetings, the seminar brought together representatives from national statistical offices (NSO), regional and international institutions, funders, and the private sector to discuss innovations in data and statistics in general, and specifically on: collection, new sources of data, dissemination, visualisation, non-official data, and public-private partnerships.
To begin the seminar, Pieter Everaers, Director of Cooperation in the European Statistical System and International Cooperation, EUROSTAT, framed the discussion by noting that the world of data and statistics is changing. Statisticians must think about what role they will play in the face of big data, open data, and a host of other innovations.
Following this introduction, Matjaž Jug, former Chief Information Officer, Statistics New Zealand, presented the draft of a PARIS21 discussion paper “Information Revolution”. This paper, to be released in the coming weeks, walks through a series of innovation case studies, such as mobile phone data collection in Tanzania, the National Data Archive (NADA) tool for data dissemination, and cloud-based data storage and analysis for healthcare in Nigeria. Following a brief discussion on these examples, participants broke out into smaller groups to learn more about a series of specific innovations or fields of innovation, and discuss what these could mean for the data revolution and national statistical systems in general.
“Collection: Computer-Assisted Personal Interviewing” by Dr. António Duarte, President, National Statistics Institute of Cabo Verde)
Following the successful event in March 2014 in Praia, Cabo Verde, “Regional Workshop on the use of Mobile Technology for Data Collection and Statistical Production in Africa” this session explored the promise of using computer-assisted personal interviewing for data collection. While efficiency and validity gains are quickly apparent, challenges remain in terms of harmonization, funding, and capacity.
“New sources: Using call logs for evidence-based decision making” by Nicolas de Cordes, VP Marketing Vision, Orange
This session explored the ways in which call logs from mobile phone companies can be used for extrapolating certain data patterns useful for policy making. These could include a poverty index based on telephone use, redefining urban or rural boundaries, or following the transmission of disease. Questions remain on how to ensure a sustainable legal framework for countries to make use of this data without infringing upon privacy, and to what degree these data will be monetized by the private sector in the future.
“Dissemination: Tools for data Anonymisation” by Shuang Chen, Junior Programme Officer, PARIS21
In the push for open data, many have raised concerns over how to protect privacy. This session explored tools used for data anonymisation, including removing or randomizing certain data points, adding perturbations and introducing noise, or otherwise treating the data such that it remains valid but distinct data points cannot be tied back to individuals. Ultimately there must be a trade-off between the rate of data disclosure, and the ability to keep people’s private data safe.
“Visualisation: Data visualisations for making data understandable to a wider audience” by Trevor Fletcher, Senior Project Co-ordinator, PARIS21
One of the most successful methods of getting data into the public hands is data visualisation. This session highlighted the importance of catching the attention of policy makers by reducing the gap between the language of statisticians and data users. Participants noted that statisticians sometimes forget that they are service providers; while they focus mostly on collecting data, there should be an equal focus on presenting that data clearly and useably to whoever needs it, so that it can lead to better policy.
“Filling gaps with non-official data: Using geospatial tools and targeted household surveys for urban mapping” by Ronald Roberts, Managing Director, Household Insights International
In this group, the discussion was on the use of geospatial data for the purpose of urban mapping, for example overlaying aerial photographs with a three year time gap to see the growth in the number of residences in a given neighbourhood. This kind of data collection, combined with household surveys, can be a faster and more efficient way of gathering data that may be otherwise difficult to obtain. As described by one participant, these data exist whether they are used or not.
“Public-private partnerships: Exploring the possibilities of PPPs for statistical production” by Johannes Jütting, Secretariat Manager, PARIS21
This session explored the concept of public-private partnerships (PPP) for statistical production. The main question arising from this session was how to define a PPP in this context – do NGOs count, or subcontracting for surveys? Further discussion revolved around the legal and ethical implications of data in the hands of the private sector, a lack of incentives, and a means of ensuring that any PPP is sustainable in the long term.
The seminar closed with the thought that a series of words continued to appear throughout the morning: confidentiality, sustainability, methodology, and governance. Addressing these issues, regardless of the innovation, will be central to ensuring that statisticians in developing countries can make best use of new technologies, processes, and institutional arrangements – all necessary for a successful and meaningful data revolution.