Class, February 3

  1. The dataset spanish-test.data gives pre-test and post-test scores on the MLA listening test in Spanish for 20 high school spanish teachers who attended an intensive summer course in Spanish. Does attending the institute improve listening skills?
    1. Check the data graphically and numerically for outliers or anything else untoward. Report what you see.
    2. Give a 90 percent confidence interval for the mean increase in test score.
    3. State appropriate null and alternative hypotheses for a hypothesis test.
    4. Carry out your hypothesis test and report the p-value.
  2. The data set teks-confidence.data gives reported confidence of in-service seventh grade teachers in teaching a particular topic. Confidence is rated on a scale of 0 to 5. Also included is whether the teacher is "emergency certified", essentially teaching without being fully trained. Here 0 means not emergency certified (i.e. fully trained) and 1 means emergency certified.
    1. Is there a difference in confidence between the two groups? If so, what is the difference. State appropriate null and alternative hypotheses if you carry out a hypothesis test.
  3. The data set guinea-pigs.data gives the survival times of 72 guinea pigs after injection with tubercle bacilli in a medical experiment.
    1. Give a 95 percent confidence interval for the mean survival time.
    2. Check the data graphically and numerically for outliers, or anything else untoward. Report what you see. (Yes, you should ordinarily do this first, but that disrupts the flow of this problem.)
    3. Take the logarithm of the data (use the log command in R) and repeat the checks above.
    4. Give a 95 percent confidence interval for the mean log survival time and transform this (use the exp command in R) back to an interval of untransformed units (i.e. undo the logarithm).