Polling or sampling is the process by which the views of a number of people in a population are taken so as to represent the way the entire population looks at or views a certain issue. Polling is basically carried out in all aspects regarding issues that the public is entitled to have a say on. It is not possible to go round and get the views of each person in the population hence polls are used, although they are subject to errors. There will be, in this paper, a description of each of the polling techniques and the cons and pros associated with them as well as an explanation of the bandwagon effect. The conclusion will come last (Thompson, 1995, p.10).
So as to get these views from the population, various methods are employed. There is random sampling, the quota system as well as cluster method. In the random sampling, also called probability sample, all people within the population that is targeted get a fair chance to be selected to take part in the polling process. In this method, there are extensions to which when a certain percentage of participants is reached, it is no longer important to collect more data as they can hardly change the overall results. This method is called random because no specific order is followed in choosing who gets a chance to take part in the poll. The means used in data collection are by way of telephone (calling the participants via phone and questioning them), online survey via email or an interview that involves face-to-face contact with the interviewee. By telephone surveys, the random digit dialing is used in which case, the interviewer will choose telephone numbers from a list of the ones published recently, then he or she will go on to call these numbers randomly such that even those individuals who got telephones most recently, also participate. This way, random sampling comes out as fair, lacking bias (Thompson, 1995, p.12).
Cluster sampling on the other hand, refers to interviewing people who emanate from one neighborhood. In cluster sample, the interviewer goes from one home to another whereby after several cluster samples are carried out in selected neighborhoods, the results are combined and the poll is summed up. This method has an advantage over random sampling because samples of in-home views from varied locations but at random, the cost of having to travel many kilometers to get these in-home results is a discouraging factor. However, the fact that various geographical regions in a place are chosen randomly, it remains true that margin errors will occur (Deming, 1970, p.23).
In the quota sample, there is a variation in terms of the individuals who take part in the poll. Various individuals are chosen with an effort being made to ensure that the demographic variables including sex and age are as varied and as distributed and as much as possible. In this method, sampling is so important such that if the researcher is for example sitting at a specific place, they will ensure they pick on people basing on demographic variables such that if the researcher had interviewed people of a certain age limit before such as youth and boys mainly, he or she will now target those of an older age mainly the old so as to distribute and balance the survey well. The problem with this method however, is that there are chances of the interviewer being biased as he or she chooses on who to select based on the demographic variables (Deming, 1970, p.25).
Bandwagon effect or cromo effect refers to the situation in which people get to believe, and also do what other people do without having to decide on their own. This effect is associated with behavioral psychology whereby beliefs can quickly spread in groups of people such that within a short time, many people tend to believe and do the same thing. In another way, it can be seen that these people who follow what the crowd is doing are afraid of thinking on their own. Relating this to polls, there can be a situation whereby if the interviewer is openly interviewing people in the open, some will simply say or answer exactly what others have said, without it being their sincere views. In this case, the sample will not be a genuine representation of what the large population thinks or holds (Deming, 1970, p.27).
Conclusively, we note that due to the feature called random digit dialing in telephone surveys, random sampling comes out as the fairest and most reliable polling method. This method ensures that the demographic variable gets a fair chance hence the population receives a fair representation.
References
Thompson, S. (1995). Sampling techniques. New York: John Wiley & Sons Inc.
Deming, W. (1970). Sampling Theories. New Jersey: McGraw-Hill.