Instituition
1. A survey form is placed in a magazine that is only read by a particular category of people,say group D. Analysis survey mainly indicates the views of group D and not those of others. This is the case of nonrandom sampling method.
2. Injecting laboratory rats used for experimental purposes with a small dosage of drug over long periods did not cause any metabolic changes in the rat. But if the rat is injected with a large dose of drug, metabolic changes can be observed which was not present earlier. This is the case where an experimental condition changes the nature of system being tested.
3. The results of a study concluded that 60 Hz power line radiation caused cell changes. Other laboratories were unable to duplicate the results. Investigations proved that the "researcher" had discarded all the data that didn't fit the theory, and had based the results on one tenth of the data. This is a case where the samples that do not fit the theory have been discarded.
4. The incidence of the disease rickets was found to be strongly correlated as being present in certain families.Early scientists had come to the conclusion that the disease was hereditary. Later it was proved that rickets was the result of malnutrition. In fact poverty was inherited. Here correlation has been confused for causality.
5. Withdrawal of salt from diet has a remedial effect in some cases of high blood pressure. Some people including government regulators concluded that salt causes high blood pressure. Removal of salt abnormally thins the blood, allowing it to easily flow through constricted spaces This is a case of confounding a corrective effect with separating the cause.(Abusing statistics for politics and profit)
6. Politicians and news sources reported that U.S. automobile companies pay their workers $76 an hour. All people imagined that amounted to $3,000 a week salary. Auto workers were outraged at the report since their weekly earnings was around $1,100. This was far less than the widely reported statistic. The per hour salary of $76 was a true fact. Financial reports of the big three auto companies in US confirmed this.The two facts represent the same thing. How could there be a discrepancy? The answer was simple. Statistics rely on the data used to derive them. The $76 amount represented the total labor cost for the automobile company on a hourly basis. The amount of $76 was obtained by adding up the money spent in the category of labor:wages, benefits, supporting pay to laid-off employees, retirement benefits to retired employees and disability pay to injured workers. This amount was then divided by the number of working hours. $76 accurately reflects the company’s hourly labor costs. No worker earned $76 an hour, even after adding the benefits. The truth is that traditional costs (retirement benefits and health insurance costs paid to retired workers and obstinately protected by the trade unions) produced an artificially inflated hourly labor amount.(How Politicians Misuse Statistics )
7. A sample is used to represent a large group. It must be guaranteed that the people in the sample are fairly representative of the larger group. The survey is about the amount of allowance received by people in the age group 10 to 15. It is to be decided whether a mall is the ideal choice to conduct the survey. The probability of finding a fair sample of people in the above age range is very less. Obtaining a sample at the mall may not represent adequately those people who receive a small allowance or no allowance.(Misleading Graphs and Statistics)
8. Pseudo Samples: A survey of 4 doctors was done. Three recommended a brand, drug a1. If the survey consisted of only 4 doctors,then it is implied that chance contributed to the results. If the survey had 100 doctors the result would have been different. The desired result would not have been obtained.
9. Vague averages: There are four commonly used measures: mean, median, mode and midrange,that are called averages. For the same data set, these averages differ considerably. People aware of this fact, select an average that supports their position.
10. Change of subject : Expressing a 3% increase in expenditure as $6,000,000 makes it look like a very enormous increase. A question is to be asked: which quantity better represents the given data? (What are the misuses of statistics that you came across?)11. Highway deaths occur four times more at 7 p.m compared to 7 a.m. A newspaper suggested
that it was more dangerous to drive in the evening than in the morning. This of course is not implied. A study of road accidents leading to casualty revealed that 98 percent of mortalities occured while driving on the left side and 2 percent occured when people travelled in the middle of the road. It was wrongly understood that it was secure to travel in the middle of the road than than on the side.
12. An example from the banking sector. Reserve Bank of India claims that inflation has come down in the recent past, whereas the the prices have not decreased. The truth is that inflation is indicated as a percentage change. Even when the percentage is slumping, it means that the prices are rising at a slow rate. The prices would reduce only when the inflation becomes less than zero.(Uses and misuses of statistics)
13. Statistics quoted based on non-representative samples.
14. A change in variable is expressed in terms of terms of actual values or percentages. This has the effect of inflating or deflating their importance subjectively. (Will you be happy if your assets expanded by $10,000,000. What is your reaction if that represented only 0.3% gain?) .
15. Usage of isolated statistics with no comparison between two quantities
16. Causal connections are implied between variables without performing a well-designed experiment. To quote the doctor " taking the drug twice a day may result in reduction of weight up to 30 kilograms in the initial 2 weeks!"
17. Changing the general appearance of graph to deceive the eye.
18. The questions in the survey are designed to bias the results.(Some abuses)
References
1. Abusing statistics for politics and profit. (n.d.). Retrieved June 14, 2016, from http://midimagic.sgc-hosting.com/statabus.htm
2.How Politicians Misuse Statistics. (n.d.). Retrieved June 14, 2016, from http://www.onlinesociologydegree.net/2011/02/10/how-politicians-misuse-statistics/
3. Marcel, F. B., Dr. (2005, July Tuesday). Misleading Graphs and Statistics. Retrieved June 14, 2016, from http://faculty.atu.edu/mfinan/2043/section31.pdf
4. What are the misuses of statistics that you came across? (n.d.). Retrieved June 14, 2016, from https://www.quora.com/What-are-the-misuses-of-statistics-that-you-came-across
5. Chakrabarty, K. C., Dr. (2012, March 20). Uses and misuses of statistics. Retrieved June 14, 2016, from http://www.bis.org/review/r120327c.pdf
6. Some Abuses. (n.d.). Retrieved June 14, 2016, from http: // www .oxfordmathcenter.com/ drupal7/node/245
