Responses

1. Correlation is not Causation

One of the major misconceptions about correlation is that a relationship between two variables means causation; that is, one variable causes changes in the other variable. There is a particular tendency to make this causal error, when the two variables seem to be related to each other.

Describe an instance where you have seen correlation misinterpreted as causation.

Correlation is not causation is a mantra of mine. At one time, I was not aware of this reality. The biggest mistake I made in regard to this was the flu shot. I believed erroneously that I would develop the flu after getting the flu shot. I have since learned that not only was I not getting the flu, I was developing an illness that may have been more closely related to my allergies. The CDC says that we cannot be infected with the flu virus by the flu vaccine. (CDC) I believed that I was because I would develop a low-grade fever and ultimately a respiratory infection. My doctor would tell me that it was impossible but could not explain why I would develop this illness immediately after receiving the vaccine. The time of year that I would get the flu vaccine was shortly after the children returned to school. In September, fall allergies are at their peak. (Oehler, 2019) Whether it is mold from leaves falling, illness from the children returning to school and being exposed to new viruses, or fall pollen allergies, it would cause mild asthma symptoms and cause serious sinus and respiratory symptoms. Often I would develop sinus and respiratory infection within a week of receiving the flu vaccine. It seemed obvious that the correlation between the vaccine and the symptoms was proof of causation. Now that I have a formal diagnosis of asthma and am undergoing treatment, it is less likely that I will develop the acute symptoms that I once did this time of year, vaccine or no. I now know that the correlation between the flu vaccine and the development of flu-like symptoms and infection is not proof of causation.

2. Linear Regression

The linear regression is one of the regression analyses that used to assess or estimate the relationships between one dependent variable and one or more independent variables. The independent or explanatory variable is used to predict the other variable. And the dependent variable responds to the independent variable that also depends on the value of the independent variable we pick. The equation of linear regression is in the form of Y = 1 + bx, in which the value of y (dependent variable) depends on the value of x (independent variable).

In my chosen major of Accounting, we would use the linear regression to analyze finance or forecast financial statements for a company. An accountant would often do multiple regression analyses, such as linear regression to determine how changes in certain factors of the business will impact the upcoming revenue and expenses of the company. An accountant would use linear regression to determine the relationship between the expected amount of sales and the cost of goods sold. For example, there may be a high relationship between the cost of the production machines or the wages of employees. An accountant can also use linear regression in forecasting the company’s performance based on different revenue, production or other factors. For example, there may be a high correlation based on the number of productions, the number of revenue that the company generates, or the number of the company’s liabilities.