The Role Of Public Opinion Polls In Our Emerging Democracy

: The Case Of Internet Polling.

Public opinion polling has become an essential tool in public policy decision-making, election campaigns, and media reporting. Among others, politicians and policy makers want to know citizens? satisfaction with service delivery, their perceived national priorities, their political preferences, and their attitudes on the state of the economy.

Because polling exerts tremendous impact on politics and public policy debate, it is important that those who undertake polls exercise due diligence and conform to minimum professional ethics and practices. Not only do polls affect public policy debate, but they also serve as one of the most significant communication links between governments and the governed. There is also a huge debate on whether polling undermines the democratic process by influencing electoral behavior and election results.


It is because of these direct and indirect effects that pollsters are to refrain from selecting research tools and methods of analysis that have the capacity to yield misleading conclusions. They are to make sure not to allow interpretations that are inconsistent with the available data. In addition, interpretations should not be given greater confidence than the data actually warrant. It is against this background that I commend Ghanaweb?s disclaimer to its polls. In its recent disclaimer, Ghanaweb advises its readers that its polls are not scientific and reflects the opinions of only those internet users who have chosen to participate. It writes:

The poll is not scientific and reflects the opinions of only those internet users who have chosen to participate. The results cannot be assumed to represent the opinions of internet users in general, nor the public as a whole. The vote sponsor is not responsible for poll content, functionality or the opinions expressed in it.

This is commendable in view of the growing controversy in the political circles about the reliability and accuracy of internet polls. This controversy exists even in cases where pollsters take great care to ensure that those who take part in their surveys are demographically and politically representative of the larger population. Critics of internet polling argue that, the sample cannot be a good representative of the entire population. This is because only a portion of the population has access to the internet. If the aim of pollsters is to generalize from the opinions of sample of population to the entire population, then it is important the sample be representative of the population as a whole. The debate over the usefulness of internet polls in Ghana may intensify in the future because of the popularity of Ghanaweb polls. Many media outlets in Ghana have started quoting these unscientific polls as if they are real polls. The Ghanaweb polling system allows anyone that has access to internet to cast votes as many as they want if they use different computers. As its disclaimer points out, polls based on voluntary responses are unscientific in their approaches and any inference drawn from the results may be fallacious. In fact, voluntary responses from Internets are not considered as polls. Polls by definition involve sampling as a way of estimating figures for a larger population. Defined this way, voluntary responses from internet visitors are neither polls nor surveys.


The extent to which the results of a sample can be applied to the general population is based on two principles. First, the sample of the population represents all the population if the sample is sufficiently large. Second, the sample is representative of the population if it is generated through a method that ensures the sample is random. Random sample means that everyone in the target population has an equal chance of being selected. This is the basis of probability sampling and fundamental principle for survey researches or polls. Getting representative sample of Ghanaian voters from the internet is a difficult enterprise, because only a few Ghanaians own computers equipped with internet.


Like many African countries, access to internet facilities in Ghana is limited to the elite groups who live in the capital cities, and Ghanaians living outside the country. Similar to newspapers, active use of the internet requires literacy in English to enjoy the benefits. This means that many people will not be able to access the internets even if they are available. Besides, local dial-up facilities are inadequate or unavailable. Other factors that militate against internet use in Ghana are high access costs, poor quality phone lines, and unreliable power supplies. Given these conditions, internet polls may not be meaningful at this stage in our democracy.


Visitors to the Ghanaweb could not speak for all Ghanaians. A representative sample of Ghanaians voters must therefore include the rich, the poor, the educated, and the uneducated, the rural and the urban dwellers. It should include the youth and the old, as well as all segments of Ghanaians from the various ethnicities. To ensure acceptable representation in the sample from each ethnic region, the sample must be stratified by region to reflect our ethnic differences and our voting patterns. This will help make easy statistical generalizations about opinions prevailing in each region and among people in the same category.

George Gallup prefers to explain random sampling with his soup analogy. Gallup contends that one spoonful of soup can accurately represent the taste of the whole pot so long as everything is WELL-STIRRED. The STIRRING is the key. That is what introduces the random element that is so important to scientific sampling.


It is because of lack of randomness in internet polling that one cannot take internet voting or straw polls as scientific polls. Michael Silberman MSNBC.com executive editor recently lamented, ?internet polling is notoriously unscientific.? Richard Morin, Washington Post Polling Director bemoaned, ?I hate online polls.? Internet polls unlike random sampling generally rely on volunteer samples; not everyone, especially older people, minorities and the poor is computer literate even in America.


Voluntary response data are worthless as indicators of opinion among Ghanaian voters. People who respond are people who feel strong enough about the issue to take the trouble to vote. Voluntary response sample consists of people who choose themselves by responding to an appeal to express an opinion about an issue. They often over represent people with strong opinion about the issues being discussed.


Murray Edelman maintains, ?For a public opinion survey to be representative of the American public, all Americans must have a chance to be selected to participate in the survey.? He asserts that web-based surveys fail to represent the views of all American and thus give a misleading picture of the public. If web-based survey is difficult for American public for lack of access to computers, then it is unimaginable for someone even to dream of using the method to conduct a poll for Ghanaian public.


Polls are as good as the theoretical and scientific bases on which they are conducted. Even besides the problem of sampling there are still execution issues that we need to settle. There are problems with question wording, question order, and confidence intervals. Also most polling data are reported in statistical terms, and can be contentious without proper interpretations. For example, what exactly is statistically significant to the average person? The impact that polling information has on the public is still a matter of debate.

There are some political analysts who argue the publication of polls gives an unfair advantage to the candidates who are leading in the polls. This phenomenon described as ?bandwagon effect? assumes that knowledge of the popular ?tide? will likely change voters intentions in favor of the winning candidates. Others counter-argue that not all voters are inclined or have the proclivity to follow the winning candidates. While some voters may want to be on the winning side, at least, there are some who want to support the losing candidates-the ?underdog effect.?


Furthermore, there are others who argue that polling can have a demotivating effect on the losing candidate or party. On one hand, poor polling results can actually demoralize party workers and supporters and make them less effective. On the other hand strong polling results can encourage party workers to redouble their efforts. For example, the recent Ghanaweb poll that indicates 62.2 % for President Kufuor reelection can demoralize the main opposition party workers, while at the same bolstering NPP workers. It is also maintained that polling can also create for candidates a false sense of security.


Though polls tread a fine line between informing and manipulating the public, they are valuable since they help political leaders to respond to the expressed will of the people. There are political analysts who ask whether political leaders should lead or follow? Should a president sell his vision to the public, or attempt to discern where the public wants to go, and facilitate the achievement of the public?s goals and preferences? There are no easy answers to these questions. In democracy, good leaders are always in search of finding the proper connection between leadership and responsiveness?


In the absence of scientific polls, various members of parliament need to meet members of their constituencies regularly to discuss affairs of public importance. Such discussions could shape public opinion, giving direct shape to their needs, their opinions, and their interests while influencing political practice.