Richard L Epstein
Page 54
15. Generalization: (unstated) Lots of people fail Dr. E's course.
Sample: The three people the speaker knows.
Sample is representative? No reason to believe so.
Sample is big enough? More like anecdotal evidence.
Sample is studied well? Yes, they failed.
Good generalization? No.
Unstated premise and conclusion: You shouldn't take a course you might fail. You shouldn't
take Dr. E's course.
19. Generalization: (unstated) A high percentage of women think men with beards are sexy.
Sample: The women who responded to the survey.
Sample is representative? No reason to believe so. Lee doesn't even have reason to think the sample is representative of the women who read that magazine. After all, they may have got
only 10,000 out of 200,000 sent out, and mostly women who like men with beards responded.
Sample is big enough? Yes, if only we had reason to believe it is representative.
Sample is studied well? Probably.
Good generalization? No.
21. Tom is not making a generalization; he's using one. Almost all pro basketball players are over
six feet tall, and people that tall won't fit into Suzy's car. Therefore, (unstated) You shouldn't
use Suzy's car to pick up the basketball player. Needs an unstated premise: You shouldn't pick
up someone in a car he can't fit into. Pretty good argument.
23. Generalization: Aquarians are scientific but eccentric.
Sample: Copernicus, Galileo, and Thomas Edison.
422 Answers to Selected Exercises
Sample is representative? No.
Sample is big enough? No.
Sample is studied well? Probably O.K.
Good generalization? No.
24. Generalization: The pacifier will stop the baby from crying.
Sample: All the times the speaker has given the pacifier to the baby.
Sample is representative? Who knows?
Sample is big enough? We don't know how often they've done it.
Sample is studied well? Possibly, or possibly bad memory.
Additional premises needed? None.
Good generalization? Weak, but there's little risk in that course of action.
29. Generalization: Questioning partners about their sex lives is not an effective strategy for reducing the risk of acquiring AIDS for young people.
Sample: 665 students attending colleges in California, aged 18-25.
Sample is representative? No.
Sample is big enough? Don't know.
Sample is studied well? Only if you trust the responses people give to questions about their sexual history and what they'd do to get sex.
Good generalization? Not as bad as it may seem. Remember, the authors are trying to disprove a generalization, rather than prove their own!
32. The analogy is comparing Maria's rice cooker to the one that Zoe wants to buy. The
generalization needed is: (Almost) all Blauspot rice cookers will have a serious defect. Though
the generalization is only anecdotal evidence, Zoe might decide that the risk is enough to go with
that weak argument.
33. The analogy is between chimpanzees and humans. It requires a a generalization that (almost) all
chimpanzees will become obese if fed one pound of chocolate per day in addition to their regular
diet. The analogy depends on the similarity of chimpanzee physiology to human physiology, and
assumes that the equivalent of one pound of chocolate for a chimpanzee to 1 % of body weight for
a human. And how much exercise did they get? A pretty poor argument: The conclusion is more
plausible than the premises.
Chapter 15
Sections A. 1-A.3
1. Causal claim: The police car's siren caused me to pull over.
Particular or general? Particular.
Cause (stated as a claim): The police car had its siren going near me.
Effect (stated as a claim): I pulled over.
2. Causal claim: Dick getting a speeding ticket caused his insurance rates to go up.
Particular or general? Particular.
Cause (stated as a claim): Dick got a speeding ticket.
Effect (stated as a claim): Dick's insurance rates went up.
3. Causal claim: People getting speeding tickets causes their insurance rates to go up.
Particular or general? General—generalizing over all examples like Exercise 2.
4. Causal claim: Your being late caused us to miss the beginning of the movie.
Particular ox general? Particular. Cause (stated as a claim): You were late.
Effect (stated as a claim): We missed the beginning of the movie.
6. Not a causal claim. (Sometimes "make" means "causes," and sometimes not.)
Answers to Selected Exercises 423
7. Not a causal claim. Inductive evidence is offered for a generalization that might be used in
establishing a general causal claim.
8. Causal claim: Someone ringing the doorbell caused Spot to bark.
Particular or general? Particular.
Cause (stated as a claim): Someone rang the doorbell.
Effect (stated as a claim): Spot barked.
10. Causal claim: Drinking coffee causes me not to get a headache in the afternoon.
Particular or general? General. Perhaps too vague: How much coffee?
Section A
3. The normal conditions.
4. We can't see how to fill in the normal conditions. It's just like when we say a premise isn't
relevant to the conclusion of an argument.
7. Reread Chapter 5.
Sections A and B
1. Causal claim: Someone pulling in front of Maria caused her to slam on her brakes.
Cause: Someone pulled in front of Maria.
Effect: Maria slammed on her brakes.
Cause and effect true? Apparently so.
Cause precedes effect? Yes.
It's nearly impossible for the cause to be true and effect false? Yes, given some plausible
normal conditions.
Cause makes a difference? It seems so, but we need to know more about what was happening at
the time. Was Maria paying attention?
Common cause? Possibly, if the other driver was trying to avoid hitting someone.
Evaluation: Plausible if nothing else unusual was happening at the time.
2. Causal claim: Wearing new shoes caused Suzy's feet to hurt when she was cheerleading.
Cause: Suzy wore new shoes cheerleading.
Effect: Her feet hurt.
Cause and effect true? Apparently so. Suzy ought to know.
Cause precedes effect? Yes.
It's nearly impossible for the cause to be true and effect false? We need to know the normal conditions. Was everything like it usually is when Suzy is cheerleading? Apparently so, from
what she says.
Cause makes a difference? Suzy says it did, by comparing it to all the other times when she
didn't have sore feet.
Common cause? None apparent.
Evaluation: Pretty plausible.
3. Causal claim: Dick pigging out on nachos and salsa caused his stomachache.
Cause: Dick pigged out on nachos and salsa last night.
Effect: Dick has a stomachache.
Cause and effect true? Apparently so, but Zoe could be exaggerating.
Cause precedes effect? Yes.
It's nearly impossible for the cause to be true and effect false? We need to know more.
Cause makes a difference? Can't say without knowing more.
Common cause? No.
Evaluation: Suspend judgment.
424 Answers to Selected Exercises
Causal claim: Marriage causes divorce. General causal claim.
Evaluation: This is tracing too fa
r back: Getting married is part of the normal conditions for getting a divorce.
Causal claim: (unstated) My not going to the game causes the team to lose. It's a general
causal claim.
Evaluation: Anecdotal evidence. Post hoc reasoning. No reason to believe it.
This isn't cause and effect, it's a definition.
Causal claim: The dark sky caused Zoe to be depressed.
Cause: The sky was dark.
Effect: Zoe got depressed.
Cause and effect true? Apparently so.
Cause precedes efect? Yes.
It's nearly impossible for the cause to be true and effect false? Can't say. We'd need to know a lot more about Zoe's psyche, or else rely on a generalization that Zoe gets depressed every
time it's dark in similar circumstances.
Cause makes a difference? Perhaps, but we need to know what happened to Zoe before that
might have made her depressed.
Common cause? None.
Evaluation: Suspend judgment until we know more.
Causal claim: The sun being strong yesterday caused me to get a sunburn.
Cause: The sun was strong yesterday.
Effect: I got a sunburn.
Cause and effect true? Apparently so.
Cause precedes efect? Yes.
It's nearly impossible for the cause to be true and effect false? Depends on what we call the normal conditions.
Cause makes a difference? Yes.
Common cause? None.
Evaluation: This is a good candidate for a cause, not the cause. Zoe being outside with her skin exposed for a long period of time is not a normal condition.
Causal claim: Lou's getting a college education is a cause of his getting a high-paying job the year after he graduated.
Cause: Lou graduated college.
Effect: Lou got a job the next year.
Cause and effect true? Apparently so.
Cause precedes efect? Yes.
It's nearly impossible for the cause to be true and effect false? Unlikely.
Cause makes a difference? Don't know. What are the normal conditions? Does Lou's dad own
the factory where he got the job?
Common cause? Perhaps Lou's parents are wealthy.
Evaluation: Plausible as a cause, if the normal conditions are right.
Causal claim: Dick telling Sally that Zoe killed Puff caused Zoe to be miserable now.
Cause: Dick told Sally that Zoe killed Puff.
Effect: Zoe is unhappy now.
Cause and effect true? Apparently so. Dick doesn't deny it!
Cause precedes efect? Yes.
It's nearly impossible for the cause to be true and effect false? Can't tell.
Cause makes a difference? Can't tell.
Answers to Selected Exercises 425
Common cause? None apparent.
Evaluation: Tracing the cause too far back. A psychiatrist might say Zoe's right. But spelling out what she believes are the normal conditions might show she's wrong. It's like the Treaty
of Versailles example, and on top of that it's subjective. Just have broad shoulders, Dick.
19. Causal claim: (General) Smoking marijuana causes heroin use.
Evaluation: First, at best it's can cause not causes, since we all know examples of people who smoke marijuana and don't use heroin. But it's also post hoc reasoning. They probably all
drank milk, too.
23. Clear possibility of common cause: Their parents are richer and/or spend time with them, which
is why they get breakfast and do better.
Section C
2. There isn't a causal claim here. Rather, Flo is overlooking one. Perhaps coincidence is just our
ignorance of real cause and effect. To be sure, our knowledge is limited. But not commonly as
limited as with children.
3. Causal claim: The pedals are making a clicking sound on Dick's bike.
Cause: The pedals are defective. ?? Effect: There is a clicking sound.
Evaluation: Good method, but a false dilemma starts it. The clicking could also come from the gears. Have Zoe put her ear close to the pedals when Dick is turning them.
Section D
9. Suzy thinks that being in the army causes men to abuse their wives. But there's a possible
common cause that hasn't been ruled out: Men who are prone to abuse their wives like violence
and hence are more likely to join the army.
11. Causal claim: A high-fiber diet can cause less colon cancer.
Type of cause-in-population experiment: Uncontrolled: cause-to-effect.
Evaluate the evidence: Clear correlation. Not clear if sample is representative, though large.
No mechanism given for explaining why there is the correlation. Equally likely is the reverse
cause and effect: People eat more fruit and fiber because their digestion is good. Until that
is ruled out, there's no reason to believe the claim.
Further tests? Controlled studies seem in order to rule out the reverse cause and effect. Try to find an explanation for the correlation.
12. Causal claim: Suggests without saying it ("Some studies on day care have found it's not bad at all"): Day care causes behavioral, emotional, and physical health problems for children.
Type of cause-in-population experiment: The first two bulleted items are not studies at all, just post hoc observations. The third bulleted item appears to be an uncontrolled cause-to-effect
study, but it's hard to say, since not enough information is given. The fourth bulleted item
appears to be an uncontrolled cause-to-effect study.
Evaluate the evidence: All this is just port hoc ergo propter hoc. There's no reason to think there isn't a common cause of parents putting their children in day care and children's
problems, namely, parents are too busy to give time to their kids. Or parents who leave their
children in day care—on the whole—have pressures that make them not parent well. Or bad
parents prefer to put their children in day care. Or . . . There's no reason to believe the causal
claim based on what's said here.
Further tests? Uncontrolled studies that factor out common threads. Controlled studies.
15. Causal claim: Giving sick people better health care causes them to be hospitalized more.
Type of cause-in-population experiment: Controlled cause-to-effect. The control group is
composed of those patients at those hospitals before the experiment, apparently. Or perhaps
those that weren't given better care. It's not made explicit.
426 Answers to Selected Exercises
Evaluate the evidence: (1) The experiment can't be generalized to anyone other than poor,
seriously ill veterans. Very dubious to generalize to all people. Especially dubious to
generalize to people who are in good health to begin with.
(2) Overlooks other possible causes, such as normal deterioration in the patients' health
during that period. (They were seriously ill to begin with.) Or the doctors, knowing the
patients didn't have to pay, were more willing to hospitalize them.
Overall, there's not much you can conclude from this experiment.
Further tests? Do the same experiment with other populations throughout the country, ensuring that the samples are taken randomly. Choose control groups better. Assign doctors randomly
and don't tell them the health care is free.
16. Causal claim: Having bad hair causes people to lack self-confidence.
Type of cause-in-population experiment: Sort of a cause-to-effect controlled experiment, except the subjects weren't interviewed on days they actually had bad hair, but about times when
they had bad hair.
Evaluate the evidence: Hopelessly badly done. No reason to think the sample is representative even of that small age group. No reason to
think that the subjects remembered correctly.
Possibly reversing cause and effect. Research is sponsored by a company that benefits from
the results that were obtained, so the authority of the researcher is called into question
(possible conscious or unconscious bias of the researcher).
Further tests? I can't even begin to imagine any. It's a waste of time and money (except for Proctor & Gamble).
Review Exercises for Chapters 12-15
1. A collection of claims that are intended to show that one of them, the conclusion, is true.
2. The argument must be valid or strong, the premises must be plausible, and the premises must be
more plausible than the conclusion.
3. For a valid argument it is impossible for the premises to be true and the conclusion false. For a
strong argument it is unlikely the premise could be true and the conclusion false at the same time.
4. No. It could have a dubious premise or beg the question.
5. A comparison becomes reasoning by analogy when a claim is being argued for. On one side of
the comparison we draw a conclusion, so on the other side we should conclude the same.
6. 1. Is this an argument? What is the conclusion?
2. What is the comparison?
3. What are the premises? (one or both sides of the comparison)
4. What are the similarities?
5. Can we state the similarities as premises and find a general principle that covers the two
sides?
6. Does the general principle really apply to both sides? What about the differences?
7. Is the argument strong or valid? Is it good?
7. a. Add all the numbers in the collection. Divide by the number of items in the collection.
b. Same as the average.
c. The midway number: As many numbers in the collection are greater than it as are less than it.
d. The number that appears most often in the collection (there may be more than one mode in a
collection).
8. A comparison where the base is unknown.
9. a. A generalization is an argument concluding a claim about a group from a claim about some
part of the group,
b. The sample. c. The population.
Answers to Selected Exercises 427
10. One in which no one subgroup of the population is represented more than its proportion of the