Site icon CDD Fact Check

FACT SHIELD: All you need to know about pre-election opinion polls

By Aisha Ibrahim Abba & Raji Olatunji


As the 2023 general elections draw nearer, opinion polls are becoming more popular and will get even a lot more popular in the days ahead, as more organisations conduct and publish results from these polls. But what exactly are opinion polls? How relevant are they to the forthcoming elections? Do they determine the outcome of the elections?

This fact shield addresses these pressing questions in an easy-to-understand language. 

What is an opinion poll?

An opinion poll is an activity in which many people are asked the same set of questions to find out what most people think about an issue or anything at all. In an opinion poll, interviewers would ask questions of people who are chosen at random from the population being measured. The people give their responses and interpretations are made based on the results. A typical example of an opinion poll is one conducted to determine who is likely to win a forthcoming presidential election in a particular country. 

Thus far, Nigeria has seen a number of opinion polls ahead of the forthcoming elections, some of which include the opinion polls commissioned by ANAP Foundation as conducted by NOI polls, and The Bloomberg Poll. Based on data from its survey, these polls predicted the possible outcome of the 2023 presidential elections.

Conducting an opinion poll requires mainly three major components: the questionnaire, which is the set of questions to ask; the set of people to interview, called the sampling; and finally the mode of interview.

What are the methods of conducting an opinion poll?

Opinion polls used to be largely traditional–– which typically entails phone interviews or person-to-person contact. However, in recent times, polls are now conducted via online poll surveys or even on social media platforms such as Twitter, Facebook, and Instagram.

However, during elections, Polling methods can take different forms:

Benchmark or baseline polls: This is generally the first poll taken in a campaign, it measures the initial thoughts and wants of the likely voters long before the campaigns shift hearts and minds.

Tracking Poll: This is a continuous poll in which responses are obtained in a number of consecutive periods, for instance daily, weekly, or monthly. The results are calculated using a moving average of the responses that were gathered over a fixed number of the most recent periods, for example, the past seven days.

The ANAP polls conducted to understand the position of the people on Nigeria’s presidential election in September and December 2022 are an example of a three-month tracking poll.

How are the participants of an opinion poll chosen?

There are 93.4 million registered voters in Nigeria. Can a sample size of 3,000 people adequately represent these 93.4 million people? The answer is nuanced and would depend largely on the sampling method used.

Random sampling and representative sampling are the most common method of sampling used in opinion polls. In the case of random sampling, a pollster can choose 3,000 people at random to answer survey questions and draw conclusions from that sample.

For representative sampling, the pollster carefully selects 3,000 people that represent the entire 93.4 million people. Depending on the nature of the data being sought, the methods have their different benefits. For elections, a representative sample is desirable.

This ensures the pollster covers all ages, education statuses, religions, ethnic groups, knowledge of politics, and all other factors that influence voting patterns among the 93.4 million voters.

Should you trust the outcome of opinion polls?

This depends largely on how much information the opinion pollster provides about the methodology of the poll. When deciding to trust the results of a poll, certain factors should be considered. We explain some of them below.

Who was interviewed? The polling of registered voters to determine the outcome of an election is more likely to be accurate than polling underaged and unregistered individuals. Also, the more representative the number of participants, the more it is likely to be accurate.

Who conducted the poll? Of course, pollsters that exhibit professionalism and experience cannot be compared with novice pollsters or even social media influencers. The track record of the pollster matters in making the decision to trust the poll.

When was the poll conducted? As a general rule, the more current the poll is, the more likely it is to provide an accurate or near-accurate result. Also, the state of affairs in the nation at the time the polls were conducted is likely to influence the outcome of the poll. For instance, if the polls to determine the nation’s preference in a presidential election were conducted a day after one of the candidates made a very disappointing gaffe, this may affect their performance in the polls. Opinion polls are known to overstate such events.

How was the interview conducted? The type of questions being asked, the atmosphere — free from bias or coercion. These issues could affect the outcome of the polls.

What was the percentage of error? Pollsters that indicate the potential for error in their poll are deemed to be more accurate than those who do not consider that there could be some error in their polling. Always look out for this when judging a poll.

Can opinion polls be partisan?

Yes. Polls can be partisan, and during political seasons, they are often very partisan. To decide how credible a poll is, it is important to determine who commissioned the poll, what their political leanings are, and how much influence they have over the process.

Beware of partisan opinion polls.

What opinion polls should you (not) take seriously?

There are opinion polls that you should not take seriously at all: An example is an opinion poll where the methodology and sample size are unknown. Social media opinion polls can also fall into this category.

If the method and the sample size are public and verifiable, the poll is likely to indeed reflect the pulse of the people.

What’s the history of opinion polls in Nigeria?

Opinion polls have existed in Nigeria for decades but really became popular in the country around 2010, when organisations, including the NOI Polls, International Republican Institute, THISDAY and Covenant University, all published results of surveys ahead of the 2011 election.

NOI Polls, for example, has accurately called the winner of every presidential election in Nigeria since 2011.

The newer players in the polling space in Nigeria are Bloomberg, Nigerians Decide, Market Trends International, just to name a few.

Do opinion polls determine election outcomes?

In every electoral cycle, opinion polls play three major roles which include; forecasting election outcomes, understating voter behaviour, and planning campaign strategy. While they can influence election outcomes, they do not directly determine election outcomes.

The public uses polls to understand who among the party candidates is popular on the ballot, the candidates on the other hand use them to understand who the public prefers ahead of the elections, and what they need to do or say to win hearts and minds.

One way polls influence the outcome of an election is that voters are more likely to vote for parties whose candidates they assume are more popular on the ballot.

The polls conducted by ANAP Foundation ahead of the 2015 and 2019 presidential elections projected incumbent Nigeria’s President Muhammadu Buhari as the winner and he won on the platform of the ruling All Progressives Congress (APC).  

Meanwhile, the predictions from polls conducted by several polling organisations ahead of the 2015 UK general elections and the 2016 US presidential elections were off the mark. The polls did not predict the actual outcome accurately.

Conclusion

Opinion polls are important, and are capable of influencing elections and electoral voting patterns, but they do not determine the outcome of an election. While track record in polling matters, even the best pollsters have been wrong and can make wrong calls.

Exit mobile version