Question

For each question below select the best answer from those listed and give your reasoning. Your reasoning need only be a sentence or two. It is not enough to get the right answer, you must know why it is the right answer.

Question 5

Fred's friend claimed that Canadians tend to be jerks. Fred wondered if that was true, and tested it by checking to see how many Canadian jerks he could think of. Fred's cognitive strategy is

            ["the availability heuristic", "participation bias", "in-group bias", "attrition bias"]   

and is especially subject to forms of

          ["representativeness heuristic", "participation bias", "selective recall", "survivor bias"]   

such as serial position effect.

Question 6

Match each argument to its argument type.

4 of the 10 students I met in Philosophy 101 on the first day of class were philosophy majors. It’s likely, therefore, that about 40% of students in the class are philosophy majors.

      [ Choose ]            Statistical Generalization            Statistical Instantiation      

40% of students who take Philosophy 101 are philosophy majors. Janice is a student in Philosophy 101, so there is a 40% change that she is a philosophy major.

      [ Choose ]            Statistical Generalization            Statistical Instantiation      

Question 7

Cindy knows that people who are members of the Green Party candidates are far more likely to be vegan than people in general are. Based on that fact alone, when she meets people who mention that they are vegan, she thinks, "Probably this person is a member of the Green Party." This inference...

• is faulty due to a sampling effect
• is faulty due to the representativeness heuristic
• is faulty due to its use of anecdotal evidence
• Is faulty due to the law of large numbers

Question 8

A Yale recruitment officer says: “It must be true that we offer a better education than the average university—after all, our graduates tend to do better on GREs and MCATs and LSATs than graduates from the average university."

Which of the following, if true, provides the strongest challenge to this argument? (Pay close attention to the argument as stated!)

• The recruitment officer has a strong emotional bias in favor of viewing Yale's education as superior.
• Yale students have higher standardized scores than those of the average university before they even start, creating a selection effect.
• A degree from an Ivy League university tends to make people assume the degree holder is smarter or more competent than they would otherwise think.
• UCLA students and Michigan students do even better on their GREs and MCATs and LSATs than Yale students-- and also get paid better.

Question 9

After the first two weeks of the major league baseball season, newspapers begin to print the top ten batting averages. (This number is the total number of safe hits divided by the total number of opportunities to hit.)

Typically, the leading batter after two weeks has an average of about .450. Yet no batter in major league history has ever averaged .450 at the end of a season.

This is best explained by:

• the serial position effect
• the law of large numbers
• none of the other answers
• publication bias
• selective noticing

Question 10

Using anecdotal evidence to substantiate a generalization is problematic for at least two reasons.

First, the fewer examples you have available to you, the more easily you could have observed a sample like this even if the generalization is false, because of                       

["a measure of low resistance", "publication effect", "representativeness heuristic", "the law of large numbers"]

Second, even if you have enough examples, they are likely subject to                        

["attrition bias", "the law of large numbers", "a selection effect", "publication bias"]    

, which can also make it more likely that you'd have observed a sample like this even if the generalization was false.

Choose the answers that make these statements not only true but applicable to anecdotal evidence in general, rather than to a particular kind of case.

Question 11

Suppose I argued that 'regular Americans' are doing far better economically than they were in the 90s based on a chart showing a large increase in mean inflation-adjusted income per person since then. What's the main problem with this argument?

• It fails to take into account the increase in population that accounts for a large part of the rise in GDP
• It fails to take into account the effect of inflation over that time period
• It uses a measure of centrality with low resistance
• It's not based on a random sample of all American households

Question 12

Ali is wondering whether his date enjoyed her evening. The next day he texts her to see whether she enjoyed her evening and she doesn't text back for a few hours.

He knows she is really busy, so whether or not she enjoyed her evening, he didn't really expect her to text back right away. Using your common sense as well as the evidence test, does her failing to text back provide any evidence about whether she enjoyed her evening?

• Yes, because it's at least a bit more likely that she'd text back right away if she did enjoy her evening than if she didn't, so it's at least some evidence
• No, because she wasn't likely to text back right away whether or not she enjoyed her evening
• There is no statistical pool of data to draw from, so assessing probabilities in this context does not make sense, so neither does talking about evidence.
• No, because whether or not she texts is actually independent of the hypothesis

Question 13

Suppose we take a sample, and find that a third of individuals in that sample have feature F. We take this to be evidence for the hypothesis that roughly a third of individuals in the larger population have feature F. Call this hypothesis "H".

To know how strongly this observation supports H, all we need to know is...

• whether there were members of every statistically relevant subgroup in the sample
• whether the sample was randomly selected so we can ensure that there is no selection bias
• how large the sample was, so we can ensure that it is not subject to the law of large numbers
• how much more likely it is that our sample would look like this if H were true than if H were false

Question 14

Suppose we get this piece of evidence:

77% of adults interviewed in three Philadelphia shopping malls (650 people) say that they will vote Democratic in the next presidential election.

Which of the following hypotheses does this piece of evidence most strongly support? (Remember that strength of evidence is defined by the strength factor: P(E | H) / P(E |~H).)

• Most adult residents of Philadelphia will vote Democratic in the next presidential election.
• Most people will vote Democratic in the next presidential election.
• Most people who shop at malls in Philadelphia will vote Democratic in the next presidential election.
• Most people in Philadelphia will vote Democratic in the next presidential election.

Question 15

Aliyah checks United Nations census data for all countries and finds that many of the countries with the very highest percentage of women in the population are very small island nations. Aliyah guesses that there must be something about the lives of people in these nations that explains this fact. (For example, maybe men in these places fish for a living, and fishing is a dangerous occupation, so more men die.) The clearest concern about her assumption that the explanation must concern some difference in how people live is...

• publication bias
• sampling bias
• representativeness heuristic
• the law of large numbers
• survival bias

Question 16

Anita conducts a survey to determine if Americans are willing to support the arts by giving money directly to local theater groups. One night she and her assistants interview 500 people who are attending a performance of a musical. To help ensure random selection, they select every other patron for interviewing. They ask: “Are you willing to support the arts by giving money to local theater groups?” 94% answer “yes”. Anita reports that a large majority of Americans are willing to support the arts by giving money to local theater groups.

The clearest and most significant problem with this inference has to do with...

• selective recall
• representativeness heuristic
• insufficient sample size and the law of large numbers
• sampling bias

Question 17(5 pts)

According to one large and careful study, Democrats and Republicans have significant misperceptions of the composition of the other party. For example, Democrats guess on average that 44% of Republicans make $250k or more, and 43% are 65 or older. (The actual numbers are 2% and 21%, respectively.). And Republicans guess on average that 36% of Democrats are atheists or agnostics, and 38% are LGBT. (The actual numbers are 9% and 6% respectively.)

The study also found that people who consumed the most news media had far more skewed perceptions than those who consumed the least.

Explain how the news media's incentive to frame stories in terms of inter-group conflicts may create these skewed perceptions, even if all the news reports themselves are entirely accurate. Use the concepts of selection effect, availability heuristic, and representativeness heuristic in your explanation. 150-250 words should suffice.

Question 18 (5 pts)

Consider this argument:

"Most people are fed up with celebrities talking politics. Every time I see an ordinary person interviewed on TV about this-- and I've seen at least a dozen-- they always say they're angry about it and wish celebrities would keep their political opinions to themselves."

Clearly employing the framework of the strength test, explain why this argument is weak. Make reference to sample size, response bias, and at least one other selection effect at work. (It needn't be a selection effect with a special name in the text.) 150-250 words should suffice.

Answer 1

5. Answer: Availability heuristics

It is a shortcut method used in problem solving and decision making. Here, the individual reaches a conclusion by referring to information that readily or easily comes to mind regarding a topic. Here, Fred thinks of people he knows and this depends on what will be available in his memory and the conclusion reached cannot be guaranteed to be true.

This is especially subject to forms of selective recall. This is a kind of bias that memory one's memory is subjected to. Here, the person will have a tendency to remember negative details better than positive details, regarding an event.