The fallacy of ANAP/NOI polls
One balmy evening last week I ensconced on the sofa in my residence and watched Seun Okinbaloye on Channels Television revealed the outcome of election projection results conducted by ANAP Foundation and NOI supported by TV channel. For those who know the Lagos business circuit and the personalities behind the two outfits, they would not […]
One balmy evening last week I ensconced on the sofa in my residence and watched Seun Okinbaloye on Channels Television revealed the outcome of election projection results conducted by ANAP Foundation and NOI supported by TV channel.
For those who know the Lagos business circuit and the personalities behind the two outfits, they would not be surprised by the collaboration, which I must say is also an attempt at deepening our democratic culture through the templates of audience measurements, feedback and research.
The challenge however with embarking on these type of opinion polls is that it is primarily data-based, methodology-driven and such analysis contextualized within the framework of the instruments used to collect and interpret data.
This recent poll did not reflect the diversity in the demographics of age, gender, occupation etc of the respondents. Any student of research or precision journalism knows the role of research methodology, sample population, sample size and probability or non-probability sampling.
The ANAP/NOI polls did not define the sampling of 92 million registered voters spread across the country and how they were targeted. This population of registered voters might provide for 1000 respondents for example. Those 1000 registered voters must capture the voters using phone to respond as well as those with face-to-face application of questionnaires.
Within this population, ANAP needed to define how many respondents are from each state since it is a national survey intended to project the popularity of the presidential candidates. The sampling technique of how the choices of respondents were made must reflect not only display a national spread but also demographic and socio-economic parameters.
Considering that internet penetration is only about 49 per cent within Nigeria and essentially they are urban based, covering about 30 per cent of Nigerians who live within the cities, it totally blacks out the opinions and choices of no less than 70 per cent of the rural voters. Statistics show that Nigeria’s predominant internet users are the youth aged between 16 – 40 years.
Furthermore, in this case, some of the respondents on our social media platforms are in such far-flung countries like Vietnam, Belarus and Papua New Guinea. Permit me to bring my wide travelling experience and its geographic implications to bear on this issue. Within the Asian sub-continent, the majority of the Nigerian diaspora community are from the South East and South South. The Europe and American continents hold out more South West diaspora folks. Like wise if you want to be fair to northern candidates in any kind of survey you go to the Arabian peninsula covering Saudi Arabia, Egypt, Dubai etc.
Any sampling technique that is short of these variables will produce a misinformed and fallacious poll result. I probably agree that there exists huge risks in human face to face application of survey instruments in the crises-prone, bandit infested areas, but then it is even more dangerous creating a farcical facade of hope for otherwise not so popular candidates and their cohorts by raising their hopes and adrenaline levels to the point that when eventual election results are announced late February 2023, their grief-stricken followers now embark on an orgy of violence over results they had been fooled will go their way.
This is the credibilty and viability challenge that the ANAP NOI polls face.
Much as we need to engender the utilisation of surveys and polls for election polls forecast, the devil is in the details of such inadequacies and gross misapplication of methodology as well as the eventual outcomes.
The hiatus of undecided voters as indicated in the results of the exercise also points to the need for not only more rigorous voter education but a retooling of the process and a total discountenancing of the flawed results declaring the Labour Party candidate, Peter Obi, as the preferred choice of Nigerians. The premise is faulty, the assumptions are inchoate and therefore the supposed results are misleading and fallacious.
Danladi Bako PhD, is a member of the Presidential Campaign Council of APC