Lies in Statistics (Essay Sample)

Lies in Statistics
Name
Institution
Lies in Statistics
Human beings are naturally not very good with numbers, most of them find it difficult to fit large numbers into their brains, and that is why scientist can lie to people and no one will realise what they have done. There is more to the, averages, graphs, trends and relationship that scientist show a common person, and the deal given in every calculation that is presented wonderfully may be a better deal than that the others. Statistic is usually applied to confuse, and oversimplify mass data on social life, the economy, opinion polls, and even census. The words used are well arranged to confuse people and make it more believable.
Most reports come from samples that are taken from people who can as well lie, a good example is, when someone is asked how much he earns it becomes a tricky question that most people avoid or lie. The moment research start with a lie, then the whole data and whatever report is generated out of it becomes a lie. Another example is in the Yale men statistics, whereby, the data that came from the people where mostly lies, since, to date the men said to have been interviewed are nowhere to be found. It is also noted when one is picking a sample they should get it from a population where it is assumed that there is a sample representation of the whole population  ADDIN EN.CITE Huff1954117(Huff, 1954)1171176Darrell HuffHow to Lie with Statistics1954New YorkNorton and Company( HYPERLINK \l "_ENREF_1" \o "Huff, 1954 #117"
Huff, 1954 p. 13). In some cases the sample may be biased in every sense, such as, it may include more male than female and vice versa.
When a statistics is done, it most involves the same kind of people. For example, when a study on Christians in a given area is to be conducted in a church, how will those who cannot walk or those that worship at home participate in the study? Looking into the statistic on opinion polls, sometimes the election may take the opinion of the voters in a biased way because the sample taken is random meaning it is from a pure chance that the sample will present the whole population  ADDIN EN.CITE Huff1954117(Huff, 1954)1171176Darrell HuffHow to Lie with Statistics1954New YorkNorton and Company( HYPERLINK \l "_ENREF_1" \o "Huff, 1954 #117"
Huff, 1954 p.18-21). The questions that are used should accommodate everyone’s culture, while the sample should give all the groups an equal chance of representation.
Are the people mandated to come up with sample for data collection trustworthy? If so what criteria will they use to approach a person? Will they be biased to choose those that are approachable and living out the unapproachable? It will be easy to choose a well-dressed person than a person who is in tatters. If the interview consist of moving from door-to-door, will the interviewee miss the people employed during the day and even if the schedule is changed to evening, will the night-clubbers still be represented? These are the dynamics that a researcher should analyse and come up with solutions for there to be equal representation of the whole population  ADDIN EN.CITE Huff1954117(Huff, 1954)1171176Darrell HuffHow to Lie with Statistics1954New YorkNorton and Company( HYPERLINK \l "_ENREF_1" \o "Huff, 1954 #117"
Huff, 1954 p. 22).
Most statistical facts contradict themselves, an example is when the study of life expectancy was done. It was concluded that life expectancy of a human being is at 55 years, whereby retirement benefits was placed at 65 years  ADDIN EN.CITE Huff1954117(Huff, 1954)1171176Darrell HuffHow to Lie with Statistics1954New YorkNorton and Company( HYPERLINK \l "_ENREF_1" \o "Huff, 1954 #117"
Huff, 1954 p.31). When we go with such a case, then no one is expected to benefit from the retirement plan because everyone will have died by the time they qualify for it. The research was conducted way back, but it should have been updated due to the changes of lifestyle among other factors, that may affect life expectancy.
Averages, are mostly used trickily to either influence opinion polls. The mean is used to get larger figures, by adding the figure in a set of data then dividing it by the number of data in that set, this can be accurate if the items on the data are uniform. On the other hand, smaller figures can be interpreted well using median since it can be used where it should reveal that half the population earns more money, and the rest half earns less  ADDIN EN.CITE Huff1954117(Huff, 1954)1171176Darrell HuffHow to Lie with Statistics1954New YorkNorton and Company( HYPERLINK \l "_ENREF_1" \o "Huff, 1954 #117"
Huff, 1954 p.35).
It is known that the possibility of tossing a coin will either be head or tail and if someone tosses it for a number of times chances will still be equal since it has only the two sides, the tail and the heads. The question that should linger into our mind is, how many times is it possible to get the result? How many times is it enough to get this result? When a study is done to get the probability of prostitute contracting HIV/AIDS in a given sample; the fact that there was no study showing the status of the people using the service makes the result biased in one way or another. The result may show that no prostitute contracted the virus, or all of them did depending on the status of the population  ADDIN EN.CITE Huff1954117(Huff, 1954)1171176Darrell HuffHow to Lie with Statistics1954New YorkNorton and Company( HYPERLINK \l "_ENREF_1" \o "Huff, 1954 #117"
Huff, 1954 p. 40-41).
Statistics also oversimplify the r...