The scope of a social problem needs to be fully understood before it can be effectively addressed. Chief editor at Africa Check, Anim van Wyk, explains the key role that data plays in development and why, in the age of fake news, it is so important to be able to interpret data and identify inaccuracies.
What does Africa Check aim to achieve?
Africa Check was founded in South Africa in 2012. It is an independent, non-partisan organisation that assesses claims made in the public arena, sorting fact from fiction and publishing the results. We now also work in Kenya, Nigeria and Senegal to reduce the spread of misinformation in these countries and across the continent. Our ultimate goal is to strengthen democracy by ensuring that policymakers and the public have access to accurate
information that they can draw from in order to make informed decisions.
Please comment on the importance of accurate data, particularly with reference to the development sector.
You can’t solve a problem unless you understand it, especially in the development sector in which organisations are tasked with solving the most pressing social problems. It’s also imperative that accurate data is used to monitor whether interventions are having the desired impact and reaching the right people. In cases where problems are not fully understood and further research is required, an investment should be made to collect better data. For example, with many organisations claiming that ‘every so many seconds a woman is raped in South Africa’, it appears that we know exactly how many women are affected. The reality is that we lack updated national estimates on how many rapes go unreported, so the scope of the problem is in fact unknown.
What is the state of data collection, accuracy and interpretation in South Africa, and how does it compare to international standards?
In South Africa, we are fortunate to have a rich throve such as Statistics South Africa (Stats SA) but, often, people interpret the available data with preconceived ideas, rather than trying to understand what it really says. For journalists, it’s often a case of time pressure and a lack of data literacy. At Africa Check, we provide factchecking
tools for the media. Through workshops and fellowships, we also train journalists to check for accuracy. We also
confer annual awards for fact-checking journalism.
Can you give an example of inaccurate data that Africa Check has debunked?
A frequently repeated statistic is that 80% of South Africans consult with a traditional healer over a medical
expert for basic healthcare. This claim was based on a throwaway line from a book published in 1983 by the World
Health Organisation (WHO), which had been incorrectly quoted over the years. Meanwhile, Stats SA’s latest General
Household Survey showed that fewer than 1% of households reported consulting a traditional healer first when a family member fell ill or had an accident. This claim can be harmful when it affects how public health programmes are designed.
How can data be effectively collected and verified?
When collecting data, surveys have to be carefully designed to ensure that the final data is generalisable and representative. To verify the data collected and to understand its limitations, it helps to rope in experts in the field that the data stems from. For example, we fact-checked a claim about which cities have the worst air quality. While the WHO has a database on air quality in various cities, some of the entries are quite dated. The last entry for Johannesburg was in 2011, which obviously doesn’t reflect the current situation. Also, experts told us that different cities measure air quality differently, so it’s not as simple as saying that, based on WHO data, a city has the
most polluted air.
What are the tell-tale signs of ‘fake news’ and poor reporting?
In both cases, a lack of clearly identified sources should set off alarm bells. Typical ‘fake news’ stories contain a lot of
exclamation marks and spelling errors. If in doubt, study the ‘about us’ page of a website – it can tell you whether the
publication has a history of publishing quality information. Some sites state upfront that they cannot vouch for the
accuracy of their stories. Africa Check put together a guide to spotting fake news and frequently publishes further tips, including how to conduct a reverse image search, which can tell you whether a picture has been used before in different circumstances.
What advice do you have for those using data in the design of their development programmes?
Developers of social programmes need to ensure that the data they use applies to their target population. For example, some surveys are only applicable to urban South Africans and will not be representative of people in rural areas. Data is also not transferable from one country to another. Data collection methods are also important. For example, internet polls will exclude people who don’t have internet access.
How can corporates support quality collection and use of data?
Corporates should set an example by using quality data sources and accurate data. They can also consider making skills – such as actuarial or accounting skills – available to development organisations and media to support with data interpretation.