Statistics are used to emphasize or provide an interpretation of data in order to draw people’s attention to believe on a piece of information. They could be presented in several places such as reports, journal or public advertisement. The statements which contain statistics seem to be powerful since they reflect strong support the arguments. However, not all statistics that are presented are reliable. In other words, some statistics are used to mislead and make people to conclude things wrongly. This essay aims to elucidate three main problems that statistics are utilized to mislead people, including generalization from small sample size, conclusion from a single factor, and variation of definition of terms.
To begin, statistics are possibly not reliable when generalizations are made from small sample size. This means that small sample size that does not represent the populations well, the reliability of generalization is very weak. For instance, if Cocacola Company wants to study the number of people who drink coke every day in Phnom Penh city in 2010, the number of people should be selected as sample. If, only 100 people are chosen to interviewed, and the result proves that 80 of 100 people respond positively, the conclusion is drawn that 80% of people in Phnom Penh city drink Cocacola every day. This is a kind of misleading statistics since the result was devised from overgeneralization from small sample size. Meanwhile, many studies support the same thing that the reliability of generalization mainly depends on how well the sample size represents the populations. In other words, if the sample does not mirror the population, the statistics will not be accurate, (Rowntree, 1982).
Second, conclusions from one single factor may result in misleading statistics. This means that statements could possibly be unreliable when all relevant variables are missed including in the study because the change in one thing is basically manipulated by several factors. Rowntree D (1982) bears that the generalization made through one variable can be inaccurate because it is considered as an overgeneralization. The study of potatoes size is the clear example about this. The conclusions, which is made that bigger size of potatoes rely on fertilizer, could be untrue because other factors such as weather, soil, or diseases may also influence the potatoes’ size. Therefore to avoid misleading statistic, overgeneralization form a single factor should not be made. In other words, we need to include all relevant factors to make a certain conclusion and ensure the reliability of statistics.
Last, another problem which also brings about misleading statistics is when definition of terms is different. But, giving definition of term likely varies from one to another because different people may have different ideas. However, ensuring accurate data in scientific research is beyond the different points of view of people on the term. In other words, statistics can be false when definition of term is determined in a different way. Similarly, Alden (2002) indicates that the difference in definition of terms possibly lead statistic to be confusing. To prove this, the example of child abuse rate in two cities of USA, Alaska and Pennsylvania, is emphasized. In Alaska, child is abused when his or her health or welfare is injured or endangered. Base on this definition, among 1000 population, there are 37.1 cases of children abuse. On the other hand, in Pennsylvania, there are only 1.9 cases in the same amount of population because the term child abused is defined when recent act or failure to act. Therefore, because of the ambiguous statistics, people may think that Alaskans are terrible parents. (Alden, 2005)
In conclusion, statistics which are presented in studies, reports or any advertising are very useful to make people more interested. However, those statistics can be used to mislead when errors occur in research methodology such as inappropriate sample size, statement from one aspect, and variation the meaning of terms because these kinds of major problem could make statistics unreliable. Therefore, readers should be more careful and be skeptical when statistics are presenting in order to avoid misunderstanding and making wrong conclusion which may possibly affect any crucial decisions in life.
Bibliography
Alden, L. (2005). econo class.com. Retrieved March 17, 2011, from Misleading statistics: www.econoclass.com/misleadingstats.html
Rowntree, D. (1982). Statistics without tears: a primer for non-mathematicians. Harmondsworth: Penguin.
No comments:
Post a Comment