ANC secretary-general Ace Magashule. Picture: Nigel Sibanda
South African Citizen Surveys defends the integrity of its latest report comparing the favourability ratings of prominent political figures, saying the insights from the report were not the opinions of SA Citizen Surveys, but a reflection of the views of ordinary citizens.
The survey, which caused an outcry from the ANC, found that secretary-general Ace Magashule was the least-liked politician because news about him had generally been negative.
ANC spokesperson Pule Mabe said the survey was a ploy to cast Magashule in a negative light.
However, SA Citizen Surveys strategic research director Reza Omar said sampling sites were selected through scientific probability sampling in all provinces and across the metro, urban and rural divide.
“We cannot speak to the opinion of the political parties. We can only speak to the information in the data and we stand by the integrity of that data.
“The sample is based on the latest Statistics SA estimates of the population aged 18 and older and sample weights are applied to ensure the sample represents the most recent national population figures with respect to province, race, gender, age and area type.”
Omar added: “Quarterly data is released on a total of 3 900 respondents, which produces results with a margin of error of ±1.5 percentage points at the 95% confidence level.
“Monthly data released with a sample of 1 300 respondents results in a margin of error of ±2.5 percentage points at the 95% confidence level.”
The organisation’s website said Magashule’s low rating was no surprise as he had been in the news for the wrong reasons.
“In one recent example, he was blamed for the rand losing value after saying the ANC wanted to change Reserve Bank policy.”
He claimed he was merely expressing ANC policy.
“His dealings while he was Free State premier and the ANC leader were also exposed in Pieter-Louis Myburgh’s book Gangster State, which almost certainly added to his low level of support.”
SA Citizen Surveys said the drip-feed of bad news involving Magashule from the Commission of Inquiry into State Capture might be playing a big role too. “The inquiry has heard testimony of how proper procedures were flouted to ensure the Estina dairy farm project company ended up managing R200 million.”
Magashule’s several appearances on television and at press conferences had also not improved his image, it said.
Study possibly ticks all the boxes
Fact-checking website AfricaCheck says there are ways to judge the accuracy of surveys and data: “An opinion poll that samples the views of a few dozen – or even a few hundred – people is unlikely to be representative of a population of millions. Most polling organisations suggest a well-chosen sample of 1,000 people is the minimum required for accurate results.”
According to SA Citizen Surveys, their study polled more than 3,900 respondents.
AfricaCheck said: “Even large-scale surveys that do not look in all the right places can give an inaccurate picture. This is known as the ‘black swan problem’ – a reference to the erroneous assumption made in Europe for centuries that all swans are white. It was only when 17th century Dutch explorer Willem de Vlamingh returned to Europe after discovering black swans in Australia, that people began to know better.”
AfricaCheck also notes that sample size is not all that matters.
“In an opinion poll, pollsters have to include people of all relevant social groups – both genders, all ages, from different races, social-economic groups and different regions – and in the right proportions.”
SA Citizen Surveys said the sample was based on the latest Statistics SA estimates of the population aged 18 and older and it was weighted to ensure it represented the most recent national population figures with respect to province, race, gender, age and area type.
AfricaCheck, an African fact-checking website, was quoted in this story about the SA Citizen Survey which, among other things, found that Ace Magashule was the country’s least liked politician.
The Africa Check comments were in relation to assessment of surveys in general – and not to the specific survey carried out by the SA Citizen Survey. Africa Check has not evaluated that survey.
The information used in the article came from AfricaCheck’s online training material.
The Citizen apologises for any confusion that may have been created.
For more news your way, download The Citizen’s app for iOS and Android.
Download our app and read this and other great stories on the move. Available for Android and iOS.