Question

True or False: Suppose we use a t-test for one population mean to test H0: μ = 3418 g and we reject H0 at α =.05, two-tailed. When we state that the hypothesis test result is "statistically significant at the 5% level", we mean that (mark each statement as true or false):

A) There is a 5% (long-term) chance that the decision to reject Ho is incorrect

B) There is strong evidence against the null hypothesis

C) The study did not have sufficient statistical power

D) If the null hypothesis is false, we would expect to obtain a test statistic as extreme or more extreme by chance alone in approximately 5 samples per each 100 random samples

E) The observed difference between x̅ and μ0 is definitely important clinically

Answer 1

A. True. It is called Type I error in hypothesis testing. In this example, there is a 5% or less than 5% chance of rejecting the fact that the mean value is equal to 3418g.

B. True. 95% of the samples will not have the mean value equal to 3418g. It will be significantly higher or lower than the mean value of null hypothesis claim.

C. False. The probability of accepting null hypothesis when alternative hypothesis is true. Since the test is statistically significant at alpha = 0.05 level, probability of observing the H0 is less than 5%. This test has a sufficient statistical power.

D. False. Since the null is false, we will have 95 samples per 100 random samples with more extreme values.

E. True. If Sample mean differs from population parameter, then it is very important unbiased point estimation