Hypothesis Testing and Inferential Statistics Question 1
HypothesisTesting and Inferential Statistics
Higheralpha threshold implies there is weaker evidence, while a smalleralpha threshold implies stronger evidence. When there is weakerevidence, there is a possibility of carrying more tests in order toincrease chances of certainty. Examples of situations where higheralpha threshold may be okay entails testing for survival of a patientand testing for the evidence of a certain killer disease such ascancer or HIV AIDs through medical screening. On the other hand,examples of situations where lower alpha threshold would be necessaryentail testing the chances of winning a game and testing the evidenceof strong performance in a business.
TypeI error describes a situation where a null hypothesis becomesrejected and the null hypothesis is true. Examples include a medicaltest showing that a patient has a disease, when the patient does nothave the disease and going off of a fire alarm to indicate a fire,when there is no fire. On the other hand, type II error describes asituation where null hypothesis is not rejected, when the alternatehypothesis is true. Examples of such a situation include a clinicaltrial failing to indicate that a certain medical treatment works,where it really works and a situation where a blood test fails todetect a disease it is designed to detect.
Inthis case, strong evidence is required in order to determine whetherthe method can be used in the future. This implies that smaller alphathreshold would be required so as to offer stronger evidence. Forpractical significance, I would consider alpha of 0.01. This isbecause the smaller the alpha value, the less the likelihood ofrejecting a true null hypothesis.