Same statistics but different between business and academy

Sunday. February 26, 2023 - 2 mins

Why do we use statistics?

We all know that statistics is the scientific method for determining whether or not a hypothesis is correct. Some people, however, mistakenly believe that a p-value of less than 5% indicates empirical proof. During my teaching experiences, I have seen many students believe that the 5% rule applies to their results without exception. Some of them even used incorrect statistical methodology or had critical flaws in the method. In any case, there is widespread belief in the 5% p-value.

Statistical and practical significance

What does statistical significance mean? Simply talking is extremely difficult. The question here is, “Can statistical significance guarantee practical significance?” Apparently, the answer is NO. We understand the meaning of practical implication implicitly. Most people, however, do not distinguish between the two. We must exercise caution when interpreting statistical results, especially when dealing with human subjects. A new medication, for example, could save patients under the 5% significance level. What does it imply? Is it possible that 5% of patients will die? Is it sufficient? Sometimes 5% is inadequate. Or is a 1% level of significance satisfactory? No. We must take precautions when interpreting statistical results. The 5% is not a deity.

Business statistics and academic statistics

Many people are perplexed by the distinction because both use the same statistical method and software. The main reason is that most people have received their education from academic institutions (i.e., graduate school). Particularly if a person is educated for a master’s or Ph. D. degree, he or she may be sensitive to the 5% significance because the significance is strongly related to the degree. Following graduation, they work for companies that share the same statistical philosophy. The 5% significance level is excellent. Consider a real-world business example. There is a shopping mall that wishes to maximize revenue. As a result, the company investigated who makes the most purchases at the shopping mall. Finally, they discovered that female customers purchase significantly more than male customers. For academicians, it is very important, while it is nothing for business practitioners.

Why?

Even if we accept the outcome, what business strategy could be developed and implemented? Nothing. Do we have to convert our male customers to females? Do we have to stop serving male customers and only serve female customers? Everything is nonsense. The main point is that business statistics should not aim for a 5% level of significance. Statistical analysis should be used in business analytics to develop a strategy. The gender significane is useless in most cases. Most academic institutions, however, ignored those points.

Summary

The goal of business analytics is to use statistical analysis to develop a feasible strategy for action. NOT a 5% significance level.

Kwangmin Park

A Person who loves my family and research