Opinion polls, or to be more precise election polls are important aspects of democracy, especially during elections.

They do not only seek to correctly predict the outcomes of elections, they also help politicians to understand how the public, or electorates think.

Political polls, like all research works, are expected to be scientific, professional, accurate and most importantly, credible. The outcome of a poll may not be as exact as predicting sunrise the next day, but a credible poll must be near perfect, with a reasonable margin of error, at least within +5 or -5%.

But when predictions of a supposed professional polls are completely wide off the mark by over 20% margin, plus or minus, it leaves much to be desired, especially when the researchers preach independence.

MUSAH DANKWA, GLOBAL INFO ANALYTICS AND THEIR FAILED PREDICTIONS

One of such polls, whose real outcomes were ridiculously over the bar, was a so-called poll conducted by Global Info Analytics led by its leader, Mussa Dankwa, ahead of the NPP's recently-held Super Delegates Conference in August.

Prior to the NPP’s Super Delegates Conference, Mussa Dankwa and his Global Info Analytics predicted that the top 5 candidates would be Dr. Mahamudu Bawumia (36%); Alan Kyeremateng (33.3%); Kennedy Agyepong (24.8%); Kwabena Agyapong (1.5%), and Dr. Afriyie Akoto (1.3%).

The above prediction, emanating from a supposed poll conducted by Global Info Analytjcs, turned out to be a figment of the imagination of Mussa and his team, as the evidence clearly indicates. Not only did the so-called polls fail woefully in terms of the top 5, the order of arrangement, in terms of positioning, and predicted percentages, were embarrassingly off the mark.

Global Info Analytics predicted 36% for Dr. Bawumia, but at the end of the election, the Vice President scored nearly 69%, a difference of about 33%! Which serious, professional, and credible poll gets a margin of error of more than 30% plus?

Again, Global Info Analytics predicted 33.3% for Alan Kyerematen in the Super Delagates, but at the end of the election, Alan got about 10%, meaning Global Info Analytics' prediction failed by nearly -23%! Again, which credible poll gets this high percentage of margin of error?

Global Info Analytics' so-called polls for the Super Delegates was also badly exposed by their failed Kennedy Agyapong prediction. They predicted 24.8% for him at 3rd place, but at the end of the polls, Kennedy Agyapong rather had 14.30% % and he surprised Alan Kyerematen at 2nd position. How come a so-called credible poll could not predict Alan won't be 2nd?

Again, the so-caled polls predicted that Kwabena Agyapong would be among the top 5, and Addai Nimo would place last among the 10 people who contested the Super Delegates. None happened! Kwabena Agyapong did not make the top 5, and Addai Nimo did not finish last. On the contrary, Addai Nimo was joint 5th with Boakye Agyarko, and he is now in the final race next month.

From the above, where lies the credibility of Global Info Analytics as far as political polls are concerned?

A research organisation that conducts a poll, so-called, and its predictions fail by a whopping +30% and -20% margins should be concerned about its credibility and the credibility of its polls.

CREDIBILITY IN RUINS, APOLOGY REQUIRED

Musah Dankwa and his Global Info Analytics should be apologising and explaining to Ghanaians why their NPP Super Delegates prediction failed miserably, rather than unashamedly coming out with another prediction for the main congress on November 4.

Interestingly, similar to the concocted Super Delegates prediction, Mussa Dankwah and his research group's forecast for the main delegates conference in November, surprisingly put Dr. Bawumia 34.8% ahead of Alan Kyeremateng 27.9%, and Kennedy with 12.5% - before Alan resigned

Mussa Dankwa further predicted that in case of a run-off, Alan would edge out Bawumia 38% to 37%. He even went on to say that 72% of those who voted for Kennedy in the first round would vote for Alan. Again, Mussa Dankwa’s poll predicted that Alan would win a head to head election (54% vrs. 46%) against Dr. Bawumia within the general population.

Isn't the above data ridiculous, considering how the Global Info Analytics' own predictions for the Super Delegates, from same polls they claimed to have conducted?

The fact that Alan performed so woefully, contrary to their prediction, and he is no more in the race after bowing out, greatly expose Mussa Dankwa and his group.

It appears Mussa Dankwa and his Info Analytics have not learnt any lesson from the embarassing failure of the last 'polls'. They are still making wild predictions ahead of next month's congress.

For instance, in their most recent polls ahead of the NPP's November 4 election, Mussa Dankwa and his team puts Kennedy Agyapong at 39% against Dr. Mahanudu Bawumia at 38% in some so-called “swing regions.”

If I may ask Mussa and his team, which region in Ghana can be classified "swing", as far as NPP internal politics is concerned? Which data did he use to determine a swing region in an NPP internal election? Didn't Dr. Bawumia win all 16 regions in the Super Delegates? And for the purpose of trend analysis, didn't Nana Akufo-Addo win all 10 regions in the 2014 NPP super delegates and main congress?

Election poll is a serious business, which should not be reduced to the levels of lotto forecasting.

The people of Ghana, and the NPP delegates Global Info Analytics seek to serve with their polls, deserve respect and an apology. They must apologise for a grand deception cloaked as political poll, for the NPP Super Delegates.

Comparing the Global Info Analytics predictions of the NPP Super Delegates, with the real outcomes, clearly shows that the so-called poll was one cooked in the kitchen of Mussa Dankwa and his group.

Sadly, some renowned media organisations tied their blue-chip reputations to Mussa Dankwa’s “analytics” and claimed joint-ownership of those polls. And these media organisations are still encouraging another grand deception by projecting same organisation and same 'pollster'!

The media have a responsibility to challenge Mussa Dankwa and his Global Info Analytics about its methodology, instead of giving him free and unchallenged platforms to give Ghanaians baseless, muddled and confused presentatios.

Ghana's democracy is growing, and part of the growth should be promoting credible democratic institutions, including strong, firm and professional research institutions.

Our democracy does not need self-serving polls that are cooked in a kitchen and served to clients ready to eat.

A credible research does not lie. It may not be exact, but the margins of error would always be reasonably understandble. But a pre-determined polling results, cooked in a researcher's kitchen for clients, instead of the people, will always be exposed!

As it is universally acknowledged a credible election poll should predict what voters think. But it is obvious, from the evidence of the failed NPP Super Delegates polls that, Mussa Dankwa and his Global Info Analytics are rather in the business of telling the public what they think.

Let's say not to fake election polls, which deceive the public and adds no value to our Democratic dispensation.

---

By Dr. Ekow Acquah, Political Communications Consultant