Biostatistics Discussion Response (mod 3)
Reply to student’s discussion on biostatistics
After having the opportunity to review some of the background readings and additional sources, it was evident that as with any other type of testing, hypothesis testing can yield Type I and Type II errors other wise known as alpha and beta levels. The alpha value or “error rate that one is willing to accept” represents the existence of a Type I error state and is usually set at .05 or .01, while the beta value represents a Type II error state. In addition, a Type I error means supporting the alternative hypothesis when the null hypothesis is in fact true while a type two error is NOT supporting the alternative hypothesis when it is indeed true (Essoe, 2015). Power, a closely related concept, is “the probability that you will reject the null hypothesis when you should and thus avoid a Type II error (Zint, 2000). An example of Type I and II errors as they pertain to their significance in health care and current events is saying that a vaccine for COVID-19 is effective when it is not (Type I) or saying that the vaccine is not effective although it is (Type II).
Changing the alpha or beta also alters the specificity of the test, either increasing or decreasing the chances of type I or type II error to occur, almost like a seesaw effect. An alpha value below .05 is said to be statistically significant (Zint, 2000). According to Ranganathan, Pramash & Buyse (2015), statistical significance focuses more on sample size of a group while clinical significance is more often determined by the provider and patient and based on the patient’s history amongst other factors. Just because a value is clinically significant doesn’t automatically make it statistically significant and just because it is statistically significant doesn’t automatically make it clinically significant.
Essoe, J.K. (2015). An Illustrative Guide to Statistical Power, Alpha, Beta, and Critical Values. Retrieved from https://www.psychologyinaction.org/psychology-in-action-1/2015/03/11/an-illustrative-guide-to-statistical-power-alpha-beta-and-critical-values
Ranganathan, Pramash & Buyse (2015). Common pitfalls in statistical analysis: Clinical versus statistical significance. Retrieved from http://www.picronline.org/article.asp?issn=2229-3485;year=2015;volume=6;issue=3;spage=169;epage=170;aulast=Ranganathan
Zint, M. (2000). Power Analysis, Statistical Significance, & Effect Size. Retrieved from http://meera.snre.umich.edu/power-analysis-statistical-significance-effect-size
250-275 words, APA format, scholarly source required