All parts please (5pts each) Consider a test of the null hypothesis Ho: -o again
ID: 3324152 • Letter: A
Question
All parts please (5pts each) Consider a test of the null hypothesis Ho: -o against the alternative hypothesis H: o performed at the significance level. (a) Define "p-value" in two sentences or less (b) What is the probability of committing a Type I Error? (c) Data obtained to test this hypothesis s used to form a 95% confidence interval. what does it mean to be .V5% confident"in two sentences or less? (d) While looking at the data, a researcher notices a scientific anomaly. The re- searcher then forms an additional null and alternative hypothesis. Using the exact same data, the researcher rejects the null in favor of the alternative. True on false (no justific ation needed) There are no signiflcant problems with
Explanation / Answer
1)
P- value is probability of getting the observed value atleast as extreme as when null hypothesis is true.
2)
Probability of Type I error is probability of rejecting Ho when it is true. In this case it is nothing but significance level i.e. alpha
3) If you repeat the process number of times (finding confidence intervals) 95% of times these confidence interval contains population mean.
4) No, there is no significant problems in methodology.