Chapter 13:
1) When one says that they have “proved” something, they are calling it a fact. When conducting a study, no matter what the subject or field is, you need to have sufficient backing before you call something a fact. Much of the time, this means that experiments have been run multiple times, in multiple environments, by multiple individuals. Otherwise, it is safe to say that the data “provides strong evidence”.
2) A p value is a number that helps to accept or refute the null hypothesis. In this case, despite the trivial travel time, there was still enough evidence to say that one was faster than the other. As stated in the paper, the p value is not the end all be all of the study. It is a strong statistical tool.
3) Statistics can be used for many things and in many fields. Marketing, forensics, police work, countless research projects and many other fields all use statistics. Forensic practitioners need to be able to say what the odds are that a match has been made (csafe goals). Marketing staff need to know who their target audience is.
Chapter 14:
1) When we say the probability of the data under the null hypothesis, we are saying that we are looking to see if the data will refute the it. While our end goal is to see if the data will support the hypothesis, we need to be aware that the tests will not prove it, but will show that they do not support the null assessment.
2b) Null: African American men do not have a harder time then white men hailing a cab in NYC.
2c) Null: No, Hospitals in New York State do not have a higher than average rate of cesarean sections.
2e) No, patients taking the new chemotherapy drug will not have response rates at least as good as the standard.
Chapter 15:
1) If the p value is at .049999 or .050001 you should look at the number of decimal places in the value. It has been said that at multiple points in this reading that the cutoff for being statistically significant is .05. While this is true, I would say that readings that are this close to .05 should be re-run for safety sake and in the end, applied to the .05 rule. Large numbers of decimal places should be noted and applied.
2) The test was run on an incredibly small sample size. In most cases, a sample of 30 or more is required for a solid hypothesis test to be run.
3) When you are testifying that findings are significant to put someone away for a crime, a p-value would be, at least, a solid starting point to say yes or no. As has been said in past chapters, a p-value dose not confirm but will support the hypothesis.