Statistics can be used inappropriately in different ways. This may either be done intentionally or unconsciously due to lack of in-depth understanding of statistical issues. The following are some of the ways in which statistics is used inappropriately.
Lack of context or wrong context
Giving statistical values without any background it is likely that any conclusion derived is wrong. Likewise, if wrong context is applied in getting statistical values the statistical values obtained are irrelevant. Some authors use contextual vacuum to put their spin on meaning of statistical values. For example, calculating rate of road casualty using population of an area is inappropriate e.g. casualties per 1000 people. The right way to get rate of road casualties is casualties per 1000 passengers Km (Wang, 11).
Sample surveys
Statistical value calculated using biased sample or unrepresentative sample are always wrong. Biased or unrepresentative sample may result from selecting a very small sample compared to the enter population. For example, when calculating performance of shares in a stock market where over 5000 stocks are being traded, choosing a sample of 5 stocks to calculate mean price per month and drawing a trend will be inadequate to determine price trend for all stocks. Likewise, choosing a sample of stocks from companies in one industry also gives statistical trend which cannot be generalized to explain performance of all stocks. It is appropriate to determine the performance of stocks by choosing considerably large size of stocks from different industries (Wang, 16).
Using wrong statistical measure
Different statistical measures suit different situations. For example, in determining house hold income in a country some countries use average income instead of median income. However, it’s better to use median income but not average income. This is because average income is mainly rightly skewed. Likewise, in determining spread of children performance in a test standard deviation is an appropriate measure than range because it considers all score but. However, many teachers prefer range because it is simpler to calculate than standard deviation (Wang,12).
References
Wang, Chamont. Sense and nonsense of statistical inference: controversy, misuse, and subtlety. New York: Marcel Dekker, 1993. Print.
