Menu

RE: Mussa Dankwa, Global Info Analytics and 'cooked' election polls

Voters?resize=484%2C364&ssl=1 File photo

Fri, 20 Oct 2023 Source: Mussa Dankwah

I have read an opinion piece by one Dr Ekow Aquah, a Political Communication Consultant, published by 3News in which he sought to discredit Global InfoAnalytics polls. I would like to use a few words to respond to the claims in the article.

First, Dr Ekow Aquah suggested that our prediction of the NPP superdelegate's election was completely off-mark because our poll in July gave Dr. Bawumia 36%, Alan, 33.3%, Kennedy, 24.8%, and Afriyie, 1.5%. The claim is COMPLETELY FALSE.

This was a poll of Ghanaian voters, not super delegates. I expected Dr Ekow Aquah to appreciate from research methodology that in commenting and criticizing research work, you must first understand who was sampled. Was the poll among Ghanaians, delegates, or superdelegates?

Dr Ekow failed this basic test. The views of women will not necessarily be the same as those of men in surveys. The views of voters walking freely on the streets of Ghana will not be the same as those of senior party executives within touching distance of the corridor of power.

Researchers who are sent to critique our work themselves fail the basic test of research. It appears our polls are proving contrary to what Dr Ekow Aquah has been portraying to his clients, hence the urgency to discredit Global InfoAnalytics to protect their work.

I have attached some of the polls we have conducted this year for Dr Ekow Aquah, our followers, and his clients to see how close we have been within our margin of error.

Again, Dr Ekow Aquah claims that when the poll result is too wide from the margin of error, then the poll must be wrong. To some extent, I agree with him.

However, in any poll, the proportion of voters who said they are undecided is 10 times the margin of error, we have never gone ahead to predict without using computer simulation. Pollsters don't determine the proportion of voters who would be undecided, so we deal with it in the report conclusion.

In the just-ended Nigerian election, the final poll published by NOI Polls Limited, Peter Obi had 21%, Bola Tinubu, 13%, Atiku Abubakar, 10%, Rabiu Kwankwaso, 3%, undecided 23%, Refused to answer, 30%. Anyone who thought this poll was going to pan out must have been dreaming.

Next time they want to criticize, they should first understand what they are being asked to criticize. Critics have rather come to expose their ignorance on the very subject they claim to be experts in.

We welcome constructive criticism.

Columnist: Mussa Dankwah