Homework 4

Due 2/22/22 at Midnight

Feel free to write up and submit in whichever way is easiest for you, though there is a special place in my heart for those who submit PDFs

  1. Is the null hypothesis a hypothesis about (i) the sample or (ii) the population

  2. Suppose a scientist carries out 100 hypothesis tests. Unbeknownst to her, in all 100 cases, the null hypothesis is true. That is, \(\mu = \mu_0\). If she uses a cutoff \(p\)-value of \(p < 0.15\), how many mistakes would you expect that she makes? Why type of error is this? Is there anything that she can do to avoid making any mistakes? Why or why not?

  3. A researcher performs a study and collects a test statistic finding a \(p\)-value of 0.51. What does this statment mean with regards to the null hypothesis and the data collected?

  4. An investigator wishes to examine whether there is a relationship between a particular candidate gene and susceptibility to a specified clinical syndrome. In actuality, there is an association, but the sample size is modest and the investigator fails to identify this relationship, i.e., he does not reject the null hypothesis that there is no association. What type of error is this? What factors may have contributed to his incorrectly failing to reject the null hypothesis?