Free Polling and Discrepancies Essay Sample

The ultimate purpose of an election poll is to predict the results of the actual election though not many people in the polling business deny it. As much as these pollsters ignore the truth, they know only too well that it is the truth and that a pollster's success depends on how accurate their predictions come to pass. Polling is done because in this world it is hard to carry out measurement of the real values therefore use of sampling is the best way to estimate.

Before the election in 2000, many polls put president bush in the lead by between 4 to 14 ahead of the democrat candidate Al  Gore therefore there was assumption that Bush was going to pull off a landslide victory not only in the popular vote but also Senate and GOP control. A CBS News poll in contrast gave Bush only a 2% lead thus bringing out the question, how reliable are polls when it comes to elections. Another poll by 'Portrait of America' showed Bush leading Gore by a 12 point margin just accentuated  the question of which poll in the presidential race was the most accurate.

The first thing to consider while carrying out a poll is; as was put across by Ed Goeas to see whether the poll is carried out during the weekend or weekday. He said that pollsters do not carry out the opinion polls during the weekend because of the movements that always happen during Saturday and Sundays when families go out for recreational purposes. It is hard to find someone who will give you 20 minutes of his weekend time when he is supposed to be with the family or doing something else. Goeas even went further to explain that most Democrats are found at home during weekends than weekdays, spending time with their families.

Another reason for the discrepancies in these polls was that many polls who interviewed registered voters was not as accurate as those who went to seek the opinion of the likely voters. Not everyone who is a registered voter will vote on the Election Day thus those who are more enthusiastic to vote are the best to interview for the opinion polls. In 2000, many republicans were willing to vote due to their strong anti-Clinton feelings hence the big gap by some pollsters. The enthusiasm among democrats for Al Gore was low thus less potential voters.

Many polls don't take into consideration the elements that will produce more accurate results such as geographical balance. Mathew Dowd, George Bush's director of polling at that time explained that it is very important get the exact percentage from each region to be able to reflect the truth on the ground and this must involve population too. The history of voting patterns should also be critically involved while sampling as per the prior exit polls.

The CNN/USA Today Gallup poll of likely voters made a mistake when they carried out the poll during the weekend. This automatically made it faulty thus not so reliable. Then came the CBS poll which also had an error of not only carrying out the survey on weekend but also polling the registered voters that we found earlier on to be not so accurate. The Rasmussen poll which was carried out from Saturday which is not advisable but through to Sunday and Monday which is better. The John Zogby poll which had Bush ahead with 4 points was rather confusing since it polled likely voters, which is a good thing but carried u the survey from Friday, Saturday Sunday and Monday hence reason to doubt it crept out. This poll was all mixed up leaving one confused on which poll carried out their business accurately to be able to bring out the clear picture in the race.

Sampling Techniques

The first technique vital in achieving accurate estimate is identifying likely voters for the Election Day. It is very difficult to determine the people who will actually vote in the election since many people will say they intend to vote and not actually vote. Pollsters therefore do not solely depend on the respondents' intentions to vote but a series of questions to the respondents will help the gauge the interest and likeliness of the respondent to vote and the actual election. This involves the past trend of the respondent in elections, whether they have been consistent in voting in the past. Different pollsters use different questions to respondent to determine the likely voters especially considering their background culturally, geographically, politically etc.

Determining voter preference is a technique that is very important for pollsters if they have to be accurate in their estimates. It sound so simple a task but in reality not so. Many pollsters will just ask whoever the respondent favors in the poll, which is an easy question but the truth is not many respondent have put much thoughts into the nitty gritties of the campaign. Therefore there should be questions for the respondents too to determine who they prefer but after putting many vital details on the table for them such as the running mates for the presidential candidates, the party affiliation of the candidates, the order of mentioning of the parties so that one party is named before the other and vice versa, mentioning of all the contestants including the minor ones that many pollsters ignore often.    

Gauging the accuracy of pools is vital too since very many polls in USA and all over the world has failed to predict the actual outcome of an election. The first thing to do in this is to analyze the polling track record. Some states like New Hampshire have good track record in the history of polling. It is very difficult to gauge the accuracy in that some polls are carried out early in the presidential races hence not reflective on the impact of the campaigns later on in the run-up to the election.

Setting aside enough money for multiple polls by organizations is very important because more polls are more newsworthy and therefore of value than single polls. Therefore organizations who want to carry out multiple polls will tend to compromise the procedure and sampling to be able to carry out the polls with limited budget. The polling as done in modern society is mainly by telephone because it is more convenient.  This n the first case involves randomness in choosing who to call thus not geographically representative. The organizations responsible for carrying out polls should realize that as much as multiple polls are newsworthy, having them accurate is better than some shoddy work that won't be reflective of what is on the ground. Enough money should be set aside if multiple polls have to be done to ensure that high standards are still maintained.  

Pulling is a very delicate process with a lot of elements to consider therefore should be taken seriously. Elements like voter turnout, natural interferences like weather, death, events in the run up to the election and the insignificant players in the election should all be taken into consideration with the seriousness it deserves. This is when an accurate poll will be achieved and that doesn't mean that it will be exact.

 

Have NO Inspiration
to write your essay?

Ask for Professional help