Why is α (alpha) Always Used as a Reference in Research? Why Not β (beta)?
Explanation and examples analogies
Perhaps many are wondering why alpha is always used as a reference and is the focus of research, maybe even some statistics students don’t know it. On this occasion, I intend to explain this and provide an example of a simple analogy that is hopefully easy to understand.
Table of Contents
- Type I Error
- Type II Error
- Explanation & Simple Examples Analogies
Type I Error
Type I Error in statistics is symbolized by α (alpha), Type I Error is an error because it rejects the null hypothesis, even though the null hypothesis is true. That is, the alpha value is the amount of chance we are wrong in deciding to reject the research null hypothesis.
Type II Error
Type II Error or commonly symbolized by β (beta) is an error that occurs because we decide to accept the null hypothesis, even though the null hypothesis is wrong. That is, the beta value is the amount of chance we are wrong in deciding to accept the null hypothesis in research.
Explanation & Simple Examples Analogies
Maybe some people do not know why alpha should be noticed and determined as small as possible by researchers?
If I could illustrate, for example, a judge in a court who wants to decide a serious case, for example a case of theft.
Let’s just say the defendant is his null hypothesis. Then, let us apply the decision that the judge committed Type I Error that the judge decides to sentence the defendant, when in fact the defendant is innocent (the right party).
Well, with the same flow, now we try to apply the decision by committing Type II Error that the judge did not sentence the defendant, even though the defendant was actually wrong (actually stealing).
From the 2 cases above we can see that the decision of the judge with Type Error I looks inhuman when compared to the decision with Type II Error. There is a good saying “It is better to acquit a thousand guilty people than to punish one innocent person”. Therefore, statisticians decided to focus more on the Type I Error (alpha), and not the Type II Error (beta), at once showing that the alpha in statistics is more inhuman than beta.
Conclusion
Now we know the reasons why alpha is always used as a reference and the focus of research, and if you are a researcher, it’s good to know these reasons. Hopefully with the simple analogy example above you can easily understand. Continue to be someone curious about knowledge! thanks!.
Reference
[1] J. A. Nursiyono, Statistika : Mengenal Alpha dan Beta dalam Memutuskan Perkara (2015), Accessed from https://www.kompasiana.com/jokoade/54f67e08a3331191178b4be1/statistika-mengenal-alpha-dan-beta-dalam-memutuskan-perkara#
