Nowadays, we are floundering with tremendous data and information in the information age. Especially in the case of statistics and graphs, they are rooted everywhere in real life. For example, people are charting the results of all the things in their expert areas including extremely private problems like figures of the average seasonal bath of British men and women which is surveyed by the Ministry of Labour. Should we accept all information as it is even though it had exceeded the acceptable range? Surely, we should not. The foundation of the statistic is math, but its actual content is not science but art because we can do a lot of fabrication and distortion in the given information. Therefore, we need to think with selective thoughts to accept the correct information that we want.

▲ We always have to consider the tremendous data that it is right or wrong

According to one article, Yale University graduates, who graduated in 1924, have an annual average income of 25,111 dollars. This may look plausible to catch a glimpse. If we consider that the per-capita annual average income of 1950s Americans is nearly 1,900 dollars, it can be no doubt because they are all from prestigious universities, and the number just exactly to the one's place makes us feel relieved. In statistics, it uses this kind of mental defect to give a feeling that is plausible by enumerating the exact number which is useless. For example, there is a decimal notation. if it is 25,111 dollars and 87 cents, we will trust that it is an ideal figure! Also, when people are given a question about their income, people who are driven by vanity answer will be much higher than the actual, and some may contract the number because of the report of income tax.

▲ The example of the wrong statistic

Then, how should we identify correct and healthy data? First, we should know who reported the information. We have to know the source. What we have to look carefully is finding the purpose of the data. For example, we need to consider the side of the labor union or entrepreneur of the data. Second, we should be warned of the survey method that is used. There is a lot of statistics in the finding the correlation coefficient about the result of what he or she wants by generalizing with extremely few samples. Lastly, we should examine the information that is obviously correct and survey the part which is doubtful.

July 28th, 2019

by Kim Dae-Hwan