Experts clash on poll survey
Zomba-based Institute of Public Opinion and Research (Ipor) has said it stands by the results of a survey it conducted between April 25 and June 3 which gave the pair of Lazarus Chakwera and Saulos Chilima of the Tonse Alliance an edge over the pair of Peter Mutharika and Atupele Muluzi in next Tuesday’s fresh presidential election.
Director of research at Ipor, Boniface Dulani, was Tuesday reacting to sentiments by Professor Jimmy Namangale, a statistician at Chancellor College, who trashed the report, saying the respondents might have been influenced by the political climate created by opposition political parties Malawi Congress Party and UTM that allegedly demonised the Democratic Progressive (DPP) and the court outcome of the presidential election case.
According to Namangale, reading from the circumstances, there is a high likelihood that DPP supporters shunned the survey.
But Dulani said in as much they respect fair criticism as Ipor, they also note the errors that are in Namangale’s analysis which he described as gross and dangerous.
He also expressed concern that the latter’s write-up is based on inflated figures from the voters’ register where, for instance, for the central region he is working on 2,247,157 registered voters, while the Malawi Electoral Commission (Mec) indicates that the region has 2,919,398.
“Am at pains in trying to understand the reasoning he is trying to advance, but we respect him as a fellow academic. I want to believe that the mathematical errors are indeed genuine; unfortunately, all his other arguments are based on the false numbers.
“He inflated figures. It is kind of disingenuous on his side to say that respondents were influenced by the political environment, but at the same time wants to tell us that the numbers that candidates got in 2019 have not changed,” Dulani explained.
He added that the Ipor report has scientifically represented the changes in voter behaviour with evidence and not mere assumptions.
“The political environment does change the people yes, and this is how they have changed. We asked the people and this is what they have said,” Dulani said.
Political scientist Joseph Chunga has also reacted to Namangale’s analysis in a separate write-up, saying the latter’s theoretical assumptions, analysis and conclusions are misleading, among other reasons, by overlooking the net effect of electoral alliances.
Chunga added that the premises used and assumptions are shaky and suspicious.
Chunga corroborated Dulani on the apparently falsified figures, saying it is important to get the figures right and wondered why Namangale in his analysis excludes over 670,000 potential voters.
On the point of DPP supporters’ social desirability to take part in the survey Chunga has said: “Evidence from Malawi surveys shows no significant relationship between one’s views of voting intentions and who they think is conducting the survey. Malawians are generally free to air their views. By the way, why would supporters of a ruling party and its partner be afraid of sharing their views?”
But Namangale in a follow-up interview said he is contributing to an academic discourse, adding his assumptions might be true or false.
According to the statistician, Dulani and his team erred by using the same sample size across the country’s geographical reasons despite the regions having different populations and different numbers of registered voters.
Namangale also indicated that that the annulled presidential election was not rigged but was only marred by irregularities, insinuating that the 2020 result may not be different from that of 2029 where President Peter Mutharika was declared winner.
But on this, Chunga said: “If he indeed believes that the 2019 presidential elections figures stand valid and at the same time believes the electoral alliance Mathematics go by simple addition of alliance partners’ support, the conclusion can’t be that the DPP-UDF alliance have an upper hand.”
A vibrant writer who gives a great insight on hot topics and issues
