Sampling
Question 5.1
(a) Considering that the higher the population the more the probable chances, then a population size of 100 will be appropriate for creating equal chances of tails and heads.
(b) A population size of 1000 times of baseball hits would be appropriate.
(c) Presidential elections are mostly involved with large or voters, in millions, thus in this case, the appropriate population would be ten million.
Question 5.2
In most airplanes crashes, it is only a small number of people who survives, thus the recommendation provided is not persuasive as most of the accidents are fatal and there are little chances of surviving.
Question 5.7
The first study used a large population that the second study hence resulting to more probable outcomes compared to the second study which used half the population used in the first study.
Question 5.14
In this sample, selection bias is applied in that only women who immigrated to the US and US women are studied; the study thus is not using a randomly selected sample which is likely to reflect a true state of economic mobility in women in the US.
Question 5.15
Survivor bias is used in the sample in that it is only those students who will proceed to senior-level College that will likely get average grades. Students who will not get to college thus will have no chances of getting the average grades, thus is has a survivor bias.
Question 5.16
The problem with the statement is that a biased sample is used to represent the whole population which is statistically wrong.
Question 5.17
Data is important is prediction of possible future outcomes. From the given data, the month of June has high number of divorce compared to the month of January. It is therefore, highly likely that couples who marry who marry in June will divorce than those who marry in January.
Question 5.18
What that means is that of the sick population, male teenagers and female elderly are more likely to visit the doctor.
Question 5.19
The conclusion is flawless as it is based on research facts with unbiased sample.
Question 5.35
(a)Selection bias is in this study as it only sample couples with children and those who have divorced. It does not study couples without children, hence incorporating selection bias.
(b) Many divorces occur in early stages after marriages that are two to ten years of marriage and many couples are still with a small number of children.
(c) Divorced couples with many children and married couples with few or no children can perfectly be used as a control experiment.
Chapter 6
Question 6.1
(a) Sampling error
(b) Systematic error
(c) Sampling error
(d) Sampling error
Question 6.2
(a) Sampling error
(b) Systematic error
(c) Systematic error
(d) Systematic error
Question 6.3
It will be an unbiased estimate as a random date 6 is chosen and the study is conducted weekly.
Question 6.5
(a) Would produce unbiased estimates while (b) and (c) would produce biased estimates.
Question 6.7
The USA is known for high income disparity hence the sample mean is likely to be skewed to the left.