Nowadays, the use of statistics has been a common way of giving evidence of a particular study (Wilson, 2005). For this reason, many people always find information that is supported by statistics to be utterly real. However, what many people never think about is the method that used to come up with particular statistics. Was the method used reliable, did it offer a chance for a complete representation of the study sample? Such questions are essential in establishing the credibility and reliability of any given statistics. This paper will seek to offer insights on an example of statistics presented in the article, “How to Lie with Statistics.” The author of the article asserts that, even though, trends and graphs are some of the most used data representation methods; they may not represent the actual and real situation (Huff, 2010). As such, the primary aims of this paper shall be discussing the article in areas touching on methods used to obtain the given statistics and establish chances of biases in the method and the study.
It is worth noting that statistics contains a secret language, whose purpose include among others; sensitization, inflation, cause confusion as well as oversimplification. However, there is a need to have honesty writers as well as understanding readers for any statistical results to be valid and away from semantic nonsense. From the article, one finds the statistics on the class of “24”, Yale. The statistics for this group indicate that all members of Yaleman, class of “24” make a total of $ 24,111 (Huff, 2010). Considering such a high amount, it becomes necessary to look into the statistics once more. In doing this, the emphasis is on examining the method used to come up with the data as well as establishing any form of bias in both the process and data itself. To begin with, one can see that the statistics came from a study. Possibly, the study should have chosen a number of members of the Yale man, class of “24”. For reliable data, the sample size selected ought to be sufficient to represent the whole class adequately. Just as Huff (2010) points out, it is quite hard to get all members of Yaleman, class of “24”. As such, the implication is that the study used random sampling technique. Here, the study chose a particular number of members of that class.
Evidently, the study is biased. Usually, few individuals give accurate data regarding their income (Huff, 2010). For example, a number of people would tend to lower their income due to fear of income-tax returns implications, and yet others would exaggerate it for self-satisfaction. In addition, the study is biased in that it used average in its report. The use of average in reporting this data might not be reliable since it does not explain the method of obtaining the average. Did the presenter of this information get the average through calculation of median or mean? Usually, the average regardless of whether specified, unspecified, median, or mean, presents a scenario of oversimplification (Huff, 2010). In calculating the average, one requires the total sample population. However, in this case, it can be quite hard to get all members of the “24” class. For this reason, the results given may not be representative of all members of that class.
How then can one evade the cases that come about from statistical data such as this data and data, which is inclusive? Does such a question call for every person to be a statistician so that he or she can examine any raw data presented? According to Huff (2010), one can avoid such struggle by examining the test of significance. Usually, this is a better way to find out how reliable any data set is. The test of significance presents a report of the probability that exists in a given statistical figure for it to be a representative of a whole population in any given study. As such, the test of significance shows that the representation of given data figures is not out of chance (Wilson, 2005). However, in most of the statistics presented, the presenters of data omit such a number with the thought that the reader of the data will not be able to notice the omission. In the case of Yaleman data, the method as well as the study itself is biased. The data does not include a test of significance in the report. As such, one wonders how real and reliable the data is.
From the case study of Yale class of “24”, it suffices that one can look at statistics in the perspective of both a science and mathematics. Usually, manipulations are imminent in any statistical data presentation for a given study. Even though, statistical data is important in explaining particular situations; it is important to examine the reliability of any figures presented. According to Wilson (2005), a sure way to question reality and reliability of statistical data is the assessment of the data’s test and its degree of significance. With such, it would be easy to establish the probability of a given data produced by chance. However, any statistical data can be reliable if it considers sample representation of the entire population under study (Huff, 2010). As such, for the case of Yale, the data collection method and the statistics would have been reliable if the given figure had represented all members of the said class. However, because it is hard to get all members and correct data about their income, the provided value is not something of reliance.
References
Huff, D. (2010). How to Lie With Statistics. Web.
Wilson, A. (2005). Modern statistical and mathematical methods in reliability. Singapore: World Scientific.