730 CHAPTER 15 Holistic Statistics can give us deep insight that might not be otherwise obtained. Instead of relying on a single statistical method, use a variety of different methods. Use parametric methods if they satisfy distribution requirements, and also use nonparametric methods that don’t require particular distributions. Use randomization and bootstrapping methods. ■ Multiple Technologies It is wise to not rely solely on a single technology. Some technologies are a bit quirky and can provide results that can be easily misunderstood or misinterpreted. ■ Replicate Repeat the entire study after collecting a new set of sample data. Compare results from the original study and the new study. 1. Differences from Matched Pairs Use the second-day 8 AM temperatures paired with the second-day 12 AM temperatures from Data Set 5 “Body Temperatures.” Test the null hypothesis that the matched pairs have differences with a mean equal to 0°F. Use relevant parametric and nonparametric methods. 2. Differences from Matched Pairs Use the first-day 12 AM temperatures paired with the second-day 12 AM temperatures from Data Set 5 “Body Temperatures.” Test the null hypothesis that the matched pairs have differences with a mean equal to 0°F. Use relevant parametric and nonparametric methods. 3. Correlation Use the second-day 8 AM temperatures paired with the second-day 12 AM temperatures from Data Set 5 “Body Temperatures.” Test for a correlation using the linear correlation coefficient r and the rank correlation coefficient rs. 4. Correlation Use the first-day 12 AM temperatures paired with the second-day 12 AM temperatures from Data Set 5 “Body Temperatures.” Test for a correlation using the linear correlation coefficient r and the rank correlation coefficient rs. 5.Fevers Use the sample statistics from the study conducted by Mackowiak et al. (x = 98.20°F and s = 0.62°F). Find the 99th percentile to be used as the cutoff for determining that a patient has a fever. Compare the result to the 99th percentile that would be obtained by using the mean of 98.6°F instead of 98.2°F. Compare the results. The body temperatures included in this chapter used only the values of the 12 AM temperatures from the second day (Day 2) as listed in Data Set 5 “Body Temperatures.” For Exercises 6-8, use the indicated samples of temperatures from Data Set 5 “Body Temperatures” and repeat the test of the claim that the mean body temperature is equal to 98.6°F using at least several different methods. Also, compare the results to those found previously in this chapter. 6. Replication Use the 8 AM temperatures from the second day to test the claim that the mean body temperature is 98.6°F. 7. Replication Use the 12 AM temperatures from the first day (Day 1). 8. Replication Use the 8 AM temperatures from the first day (Day 1). 9. Simulation This chapter included a simulation method for testing the claim that the mean body temperature is equal to 98.6°F. Use a technology to generate samples of size 106 (as in the sample), a normal distribution (as in the sample), an assumed mean of 98.6°F, and a standard deviation of 0.6228965°F (as in the sample). Find the means of the simulated samples, then arrange them in ascending order, and determine how many of the simulated means are at least as extreme as the sample mean of 98.2°F that was obtained in the 1992 study. Write a conclusion and provide an explanation justifying the conclusion. Holistic Exercises
RkJQdWJsaXNoZXIy NjM5ODQ=