Philosophy of Opinion Polls

Public opinion is formed of individual opinions which differ from individual to individual. We can measure public opinion by counting the individuals holding different opinions. Estimation has to be made of the number of opinions which are held on the issue under study, the number of people supporting each view, and the extent of influence or the intensity of the opinions of these people. By intensity is meant the extent to which the opinion-holders might be ready to back up their opinions by resorting to effective action or pressure on the government, such as by resorting to some sort of protest, when their opinions are neglected, which may range from strikes, protest meeting, processions, to revolt and revolution, to press their views if disregarded by the government or the men in power.

We may take a normal example of a community in which people hold different opinions; some are conservatives, others liberals and still others socialists. We can measure the strength of the three types of opinions by counting them by means of an opinion poll. It will show us that the number of me.The usual method is to collect the views of a selected number of sons, which is called a sample, while the people or population from ich the sample is chosen is called ‘universe”. For reasons of accuracy of opinion polls, the sample is so chosen that it is representative of or rtionate to the views of the various groups and parties among the nion-holders or citizens, called respondents or interviewers.

The views obtained from the replies or responses of the respondents by the rviewers according to the questions, the list of which, called stionnaire, is already prepared by the organisers of the public polls, lied pollsters. The views thus collected are known as opinion data, or ply data. Inferences are drawn from the data by the pollsters regarding distribution of opinions among the people, the “universe”, and forecasts made about the probable results of the coming elections and, therefore, t the success or defeat of the various leaders and parties in the tions. This whole process of collecting views or opinion data and of wing inference and making forecasts about future trends in the coming ions is called opinion polls or public opinion surveys.

There are old and new methods of undertaking opinion polls, ich are divided into three kinds: accidental sampling, quota sampling and dom sampling.

We shall first give a few examples of accidental method of opinion lls. This method was first used by newspapers, magazines or radio dcasters to obtain information on such questions as how many people a given newspaper or magazine or listen to the radio. It was also loyed by market researchers to find out what are the likes and dislikes the people or buyers about various consumer goods. For this purpose, llots are printed in newspapers or magazines, which could be clipped and rned to the poll researchers. This method is still used by newspapers in istan to assess the views of the people on some issues, either public or cerning the newspapers in question. Ballots can also be left in shops and es, from where they are collected by the pollsters. Yet another teresting method was to collect the opinions of the crowds on the roads, s-stands or in the railway trains at certain hours of the day. The people in places were interviewed, that is, question asked, and their opinions ly recorded by the interviewers. 

More frequently, the ballots are sent to certain selected persons ose addresses have been found from telephone directories or from

It may be possible in some future lime when electronic devices and computer technology are so developed that the whole people may press electronic buttons and the result may be shown or recorded on a display board.

conservative opinion-holders is, say, about 20%, that of the socialist opinion-holders 50%, and of the liberal opinion-holders is 30%-. Now, by measuring the proportion of various opinion-holders we will know beforehand the voting behaviour of these people at the time of election, if it is contested by three parties. Thus opinion surveys enable us to forecast election results in advance. However, in one-party elections, as under a dictator or autocrat, including even the so-called partyless elections, there is no need for opinion polls, for election results are a foregone conclusion: the dictator will always get, say 99% votes.