CHAPTER 3
Answer each question
in a sentence or two.
1. Why is it not always a good idea to get the best evidence possible?
2. The book imagines that you sample voters in a mall in Montreal.
Give two good reasons to think that this sample is unrepresentative.
a)
b)
3. What's the difference between a primary property and
a secondary property? [Do not confuse these terms from
Martin with the standard philosophical distinction between primary
and secondary qualities.]
4. A relevant criticism of the representativeness of a sample
is finding secondary properties which, it's reasonable to think,
might possibly have these two characteristics:
a)
b)
5. Explain, using the notion of secondary property, what
sample match is.
CHAPTER 4
Suppose you have
the job of finding out how people in your town feel about a proposed
garbage dump to be built. You've decided to conduct a poll by
telephone, from a random sample of people in your town.
I . What does it mean to say that a sample of people from your town is random?
2. Suppose that
all the numbers starting with '3' you could dial without an area
code are in your town, and suppose that all the numbers in your
town start with '3'. So you program your computer to produce a
random seven-digit number beginning with '3', and you call this
number and talk to whoever answers. Give three different reasons
why it could be argued that the sample you obtain this way is
not random.
a)
b)
c)
CHAPTER 5
1. TRUE/FALSE Circle
T for True, or F for False. On multiple-item questions
there may be one, or more than one, or no True answers.
T F You sample from a population of 100 Individuals by first numbering
them I - 100. Next you ask ten people each to pick a number from
I to 100 and write it down on a piece of paper.You then include
in your sample the individuals corresponding to each number picked
. This is SAMPLING WITH REPLACEMENT.
T F You take a sample from the coin box by making a. small hole
in it, and shaking the box until 10 lumps fall out. This is SAMPLING
WITH REPLACEMENT.
T F When you sample with replacement, you may get the same individual
more than once in your sample.
T F When you sample with replacement, the size of the population
is irrelevant to the strength of the evidence you get.
T F When the population contains millions of individuals, the
strength of evidence you get with a sample of 50
taken without replacement Is much greater than the strength
of evidence you get with a sample of 50
taken with replacement.
T F When the population contains millions of Individuals, the
strength of evidence you get with a sample of 50
taken without replacement is much less than the strength of evidence
you get with a sample of 50 taken without replacement.
2. Suppose you have a bag of 1000 jelly beans. You pick
10 of them at random. Five out of ten are red.
T F The most probable number of jelly beans in the bag that are red is 500.
T F It's quite improbable that the number of red jelly beans In the bag is 500.
T F You should draw the conclusion that exactly 500 jellybeans in the bag are red.
T F A precise conclusion about the number of red jellybeans in the bag is better than an approximate one.
T F Usually the more precise
a claim is, the more likely it is to be right.
3. A confidence level of .95 means:
T F The real frequency will be within the margin of error around
the observed frequency 19 times out of 20.
T F It's 95% probable that the observed frequency is the exact
frequency of the property in the
population.
T F The margin of error has to be ±.05.
T F The prediction will be wrong about 5% of the time.
4. A confidence level of .95
T F is the one all reputable scientists use
T F would not be appropriate
for some scientific procedures.
5. Draw a rectangular diagram to represent:
A population of 10,000 jellybeans
A sample size of 1000, of which 450 are red.
A margin of error of ± .03.