Sample Errors on Political Surveys - Literature review Example

Literature Review on Sample Errors on Political Surveys Literature Review on Sample Errors on Political Surveys Introduction Opinion polls or political surveys play a vital role in informing the public about the performance trends of the various political aspirants. It helps to provide the ratings that are associated with the varied aspects that the public is likely to consider while voting for their preferred candidates. In this sense, it is important that the methods and techniques used during the performance of these research activities are reliable and suitable. This helps to obtain results that can be sufficiently replicated during the actual voting process. This aspect will be discussed in this paper to identify the most suitable and appropriate methods and techniques that help to get the desired results and offer the public the much required information. Definition and explanation of the terms used Political surveys are also commonly known as the opinion polls. They are at times simply referred to as polls. They are conducted in the form of polls or public opinions that are obtained from a particular sample. These political surveys or opinion polls are common in a majority of the countries all over the world. They are mainly witnessed during the national elections or in rare cases when there are votes to be taken in the different corporate organizations. The main purpose of the political surveys is that they are designed to act as a representation of the opinions of an entire population through the conduct of a series of questions. They are in this case conducted to reflect the actual results of the true operations. Anderson and Pickup (2010, p. 243) illustrate that these are followed by the extrapolating generalities of ratios that are done within the present confidence intervals. Thus, the research methods that are used to conduct the opinion polls must be as suitable and reliable as possible in such a way that they minimize the rate of bias and can reduce the disparity between the results collected during the opinion polls and those that are obtained on the actual performance of the voting process. An explanation of the methods or techniques under review During the conduction of opinion polls, there are a number of methods and techniques that are used. These are meant to ensure that the results obtained in the surveys are as accurate as possible. In addition, they provide the ability to get precise results. Over the years, the various opinion polls have been maintained through telecommunications or through contacts between one person and another. Thus, there are varied methods and techniques. However, they are acceptable from region to region. Some of them include the ballots, verbal and processing forms which are commonly used. There are also the other types, which include surveys, complicated matrices, and the systematic methods that are more complicated than the past orthodox procedures. The types of methods that have been used over time vary as a result of the need to modify and improve on the different methods to reduce bias. For instance, the telephone method of collecting data in a pollster sample has been found to create an error since most of the people who have possessed telephones over time have been richer than those who do not. As such, the use of telephones does not create a true reflection of the opinion of the entire population (Kenneth 2002, p. 201). Understanding on the Sampling Bias in Political Surveys Despite the use of appropriate techniques and methods there are several errors that are linked to the conduction of political surveys. Most of them are related to the sampling procedures. These have been used in opinion polling for a long time within which the polling itself has developed into popular applications through the popular thought. According to Kenneth (2002, p. 200), although the response rate for some of the surveys have been declined, there are others that have been confidently accepted while others have had differentiating results. The sampling for the political surveys is done by a number of organizations that include YouGov, Zogby and the Angus Reid Public opinion. They are popularly known for the use of the internet survey methods. In this, as is indicated by Catrin & Stuck (2001, p. vii), the sample is collected from a large panel of willing volunteers. The results are then weighted so that they can reflect the demographics of the population being studied. In contrast to these, Lynne (2007, p. 58) contends that there are the popular web polls that draw on any individual who wishes to take part in the study. As such, it does not include the scientific sample of the population under study. It is for this reason that they cannot be said to be professional. In the recent past, there has been the use of statistical learning methods, which have been proposed with the intention of exploiting the contents of social media such as the posts on the micro blogging platform on twitter. These are used for predicting the voting intention polls (Brendan et al. 2010, 130). Sources of Error The polls that are based on a sample of the population are likely to have some bias. These are the sampling errors that reflect the effects of having chances or uncertainty in the sampling process. Usually, the uncertainty is provided as an error in the sampling process and is given as the margin of error. It is this margin of error that is defined as the radius of the confidence interval for a certain statistic from the survey. One of the proper examples is the percentage of people who prefer to use product A over product B. When one global margin of error is reported for a survey, it is used to show that the maximum margin of error for all the reported percentages using the full sample from the survey. In the event that the statistic used is a percentage, the maximum margin of error can be obtained as the radius that belongs to the confidence interval for a certain reported percentage of 50%. Vasilensos, Daniel and Trevor (2013, p. 993) show that for a poll with a random sample of 1000 individuals, there is a margin sampling error of about 3%. This is for the estimated percentage of the entire population. Potential solutions to the high percentages of bias In an effort to prevent sampling errors in the conduct of opinion polls, the researchers can place several potential solutions. New technological techniques and methods can therefore be used. It has been discovered that when a margin error of 3% is produced, then the same procedure has been used numerous times. It, therefore, means that the average population will be within the confidence level of 95% for about 95%. This statistic has an error margin of 3%. As such, some of the most appropriate methods include the reduction of the margin of error through the use of a larger sample. Nonetheless, if a pollster also wishes to reduce the margin error to about 1%, then it is necessary to get a sample of about 100,000 people. In essence, it is vital for pollsters to balance the cost of performing a research study using a large population versus the need to reduce the sampling error and having to use a sample size of about 500-1000 participants. However, this is usually a huge compromise in political polls. The other source of error arises from the faulty methods that are used in the demographics. Pollsters use them to weight their samples through the use of particular variables such as the party identification to be used in the election. Although cell phones have been linked to coverage bias, they still remain the most appropriate method or technique to use in sampling of political polling. This is mainly because they are widely spread and are easy to use in the sense that asking questions and getting feedback is done conveniently (Anderson & Pickup 2010, p. 270). Discussion on the strengths and weaknesses of the cell phone sampling method suggested The most appropriate methods and techniques that can be used to ensure that political surveys are conducted in a satisfactory way include the use of large samples of the population. Although this presents the possibility of getting sampling errors, it provides more accurate results that can be well associated with the actual results of the political survey. It has also been found that the sampling errors in political surveys can be reduced by relying on the poll averages. This is done through the creation of the assumption that the procedure is quite similar to the many polls. It also uses a large sample size of each of the polls to create a poll average. This has been used a number of times including the polling average of the 2008 presidential elections in the United States. Some of the strengths associated with this method or technique are that the pollsters are able to get complete responses that arise as a result of the ability to include thousands and many of the other participants. The cell phone method of sampling is appropriate since in many places, there are people who own the mobile phones. As such, sampling can be done on an entirely random basis. Nonetheless, the method of sampling through the use of mobile phones also contains some weaknesses. These can be highlighted as the rules and regulations that prohibit the calling of any numbers. This is based on the fact that some of the cell phone owners may be charged for receiving the calls. The suitability of your suggested method or technique in different research contents Despite the few weaknesses associated with the use of mobile phones to collect samples during political surveys, it still remains one of the most suitable owing to reliability and rate of accuracy. This method or technique is highly suitable in varied research contents in the sense that many people can be reached to provide their opinions on different subjects The generalisability of results obtained using your method / technique By making use of the cell phone method or technique, there is a great possibility of obtaining results that are both accurate and precise. The results can be properly generalized as they are obtained from a wide sample of the population from completely diverse areas. Therefore, it generally acts as a full representation of the entire sample. Knowledge on the Presence of Alternative Research Methods Based on the literature obtained on the topic of sampling errors in political surveys, there is a need to utilize other methods that are likely to minimize the level of bias and error in the results obtained. Some of these alternative research methods include the highly technological methods. These include the use of the internet as well as other web based services. The idea comes about as a result of some of the expenses and inconveniences associated with the use of the cell phone sampling methods. These present the opportunity of reaching so many people at a particular time in an instant manner. Nonetheless, with some of these advanced methods, there is need to perform proper research. This will help to reach a broader sample as opposed to individuals who possess the same mindset about a particular candidate. With this, any form of sampling bias will be significantly reduced. In turn, the results that the public receives from the survey results will be as close as possible as the ones obtained from the actual polling activities. Bibliography Anderson, Cameron & Pickup, Mark, 2010. "10". Voting Behaviour in Canada. Vancouver: UBC Press. pp. 243–278. Brendan OConnor, Ramnath Balasubramanyan, Bryan R Routledge, and Noah A Smith, 2010. From Tweets to Polls: Linking Text Sentiment to Public Opinion Time Series. In Proceedings of the International AAAI Conference on Weblogs and Social Media. AAAI Press, pp. 122–129. Cantril, Hadley & Strunk, Mildred, 2001. "Public Opinion, 1935-1946". Princeton University Press. p. vii. Kenneth F. Warren, 2002. "in Defense of Public Opinion Polling." Westview Press. p. 200-1. Lynch, Scott M. 2007. Introduction to Bayesian Statistics and Estimation for Social Scientists. New York: Wiley and Sons Vasileios Lampos, Daniel Preotiuc-Pietro and Trevor Cohn. 2013. A user-centric model of voting intention from Social Media. Proceedings of the 51st Annual Meeting of the Association for Computational Linguistics. ACL, pp. 993-1003. Web Read More