CHAPTER 17 On the Limits of Quantification

Question 1

What do data and quantitative evidence not tell us?

Answer 1

Everything we need to know to make decisions.

Question 2

What should decisions consider besides quantitative evidence?

Answer 2

The effects of our actions and our values.

Question 3

What can happen if we mistake correlation for causation?

Answer 3

Quantification can lead you astray.

Question 4

What is the danger of relying solely on evidence in decision-making?

Answer 4

There is no purely evidence-based decision.

Question 5

What does the absence of evidence not mean?

Answer 5

The question is unimportant or safely ignored.

Question 6

What is a common bias created by the demand for evidence?

Answer 6

Status quo bias.

Question 7

What is required for many new regulatory actions by the U.S. government?

Answer 7

Quantitative cost-benefit analysis.

Question 8

What is a consequence of regulations requiring quantification?

Answer 8

Narrowing the field of vision to quantifiable areas.

Question 9

What example illustrates the risks of focusing only on quantifiable data?

Answer 9

The drunk man searching for keys under the lamppost.

Question 10

What did the EPA report on arsenic regulation highlight?

Answer 10

Quantified benefits only included bladder and lung cancers.

Question 11

What conclusion can be drawn from the meta-analysis regarding flossing?

Answer 11

There is ‘very unreliable’ evidence that flossing reduces plaque.

Question 12

Why does Anthony continue to floss despite the lack of evidence?

Answer 12

Absence of evidence is not proof of absence.

Question 13

What theoretical reasons support wearing masks during the COVID-19 pandemic?

Answer 13

Masks mitigate the flow of respiratory particles.

Question 14

What do good decision makers acknowledge about quantitative evidence?

Answer 14

It only tells them so much.

Question 15

What can happen if our goals and values are shaped by the mandate to quantify?

Answer 15

We might embrace values we would otherwise reject.

Question 16

What is a risk of machine learning in decision-making?

Answer 16

Objectionable values can creep into decisions unnoticed.

Question 17

How can algorithms exhibit racial or gender bias without relevant data?

Answer 17

Through correlations with other biased variables.

Question 18

What is a common application of predictive machine learning algorithms?

Answer 18

Job placement, credit evaluation, and content recommendations.

Question 19

What issue arises in health care algorithms predicting care needs?

Answer 19

They may unintentionally discriminate based on correlated variables.

Question 20

Fill in the blank: Evidence is meant to be a tool used in service of our _______.

Answer 20

[goals and values]

Question 21

True or False: Quantitative evidence can tell us how to act.

Answer 21

False.

Question 22

What do large health care providers in the U.S. use to predict patients with complex health needs?

Answer 22

Machine learning algorithms

These algorithms aim to identify patients likely to have the greatest care needs.

Question 23

What is the correlation between health care costs and health care needs?

Answer 23

Strong positive correlation

Sicker patients tend to receive more and more expensive treatment.

Question 24

What type of data does the algorithm use to predict health care costs?

Answer 24

Data on patients’ past insurance claims, medical diagnoses, and medications

The algorithm does not receive information about race.

Question 25

What is a comorbidity score?

Answer 25

A measure of active chronic health conditions ## Footnote Used to assess true health care needs.

Question 26

How did the algorithm's predictions differ between Black and White patients?

Answer 26

Systematically under-estimated how sick Black patients were relative to White patients ## Footnote Black patients were less likely to be enrolled in special programs despite similar health.

Question 27

What is a potential reason for the algorithm's racial bias?

Answer 27

Using cost as a proxy for health care needs ## Footnote Less money is spent on Black patients on average than similarly sick White patients.

Question 28

What is welfarism?

Answer 28

The view that policies should be evaluated based on their implications for human well-being ## Footnote It often prioritizes utilitarianism.

Question 29

What is utilitarianism?

Answer 29

The view that policies should be evaluated based on their implications for the sum total of human well-being ## Footnote It often focuses on material costs and benefits.

Question 30

What is crass utilitarianism?

Answer 30

A form of utilitarianism that defines well-being primarily in terms of material costs and benefits ## Footnote It simplifies complex ethical considerations.

Question 31

What did the presenter at the academic presentation conclude about removing children from abusive homes?

Answer 31

Children from abusive homes are, on average, better off in foster care ## Footnote The benefits appear to exceed the costs of providing foster care.

Question 32

What critique was raised regarding the policy recommendation to remove children from abusive homes?

Answer 32

The researcher had not estimated all relevant costs and benefits ## Footnote Benefits to abusive parents were not considered.

Question 33

What problematic suggestion did Larry Summers make regarding pollution?

Answer 33

Encouraging the migration of dirty industries to less developed countries ## Footnote This was based on the economic logic of cost-effectiveness.

Question 34

What is the issue with valuing outcomes based on willingness to pay?

Answer 34

It can imply that the well-being of the rich is more important than that of the poor ## Footnote This raises ethical concerns about equity.

Question 35

What is the role of cost-benefit analysis in policy decision making?

Answer 35

It quantifies costs and benefits to provide bounds on decision making ## Footnote However, it does not eliminate subjective moral opinions.

Question 36

What is a potential moral concern with transferring toxic waste from rich to poor countries?

Answer 36

It may violate principles of fairness, justice, and responsibility ## Footnote Rich countries should take responsibility for their own actions.

Question 37

What is the relationship between quantification and values in decision making?

Answer 37

Quantification can shape the goals and values we pursue ## Footnote We often become utilitarians because we quantify.

Question 38

What is an ethical position consistent with quantifying consequences?

Answer 38

It can be flexible and include considerations like rights and responsibilities ## Footnote Quantitative analysis can incorporate equity considerations.

Question 39

What moral arguments should be considered when discussing toxic waste disposal?

Answer 39

The manner in which decisions are made regarding toxic waste disposal should be considered, especially in terms of consent and compensation for poor countries.

Question 40

What was Summers's stance on shipping toxic waste to poor countries?

Answer 40

Summers initially supported the idea but later criticized it, stating that the sentiment was obviously wrong and that the expression of the idea was not constructive.

Question 41

What are the potential consequences of quantification in moral arguments?

Answer 41

Quantification can lead to crass utilitarianism, pushing towards ruthless conclusions, but it can also provide valuable insights into trade-offs.

Question 42

What is the danger of focusing on quantifiable data in ethical discussions?

Answer 42

Focusing on quantifiable data can obscure rights, dignity, and fairness, which are harder to quantify and may lead to objectionable utilitarianism.

Question 43

What should be the goal of using quantitative analysis?

Answer 43

The goal should be to estimate important quantities without distorting the underlying values and goals against which choices are evaluated.

Question 44

True or False: It is acceptable to use data to mislead others if one is skilled in quantitative thinking.

Answer 44

False

Question 45

What responsibility comes with being a savvy quantitative thinker?

Answer 45

To be transparent about the strengths and weaknesses of evidence and to help others think clearly with data.

Question 46

What fundamental concepts should a quantitative thinker understand?

Answer 46

* Selecting on the dependent variable * Statistical vs. substantive significance * Reversion to the mean * Publication bias * Sources of cosmic habituation * Correlation vs. causation * Research design

Question 47

What is a key critique of the economic benefits presented for low-income neighborhood amenities?

Answer 47

Questions regarding the accuracy of the cost and benefit estimates should be asked before proceeding with the plan.

Question 48

Fill in the blank: The exercise of quantifying costs and benefits can lead to a serious argument for policies involving _______.

Answer 48

consent and compensation

Question 49

What is the importance of thinking clearly about data?

Answer 49

It is essential for understanding the world and making it a better place.

Question 50

What does the concept of 'publication bias' refer to?

Answer 50

The tendency for journals to publish positive results more than negative or inconclusive ones.

Question 51

What is the significance of the meta-analysis on flossing?

Answer 51

It found statistically significant evidence that flossing reduces gingivitis.

Question 52

What does the term 'crass utilitarianism' imply in the context of moral reasoning?

Answer 52

A form of reasoning that focuses solely on material costs and benefits, neglecting ethical considerations.

Question 53

What should individuals consider when making decisions without quantitative evidence?

Answer 53

The factors leading to their decisions and how quantitative studies could provide more compelling evidence.

Question 54

What is the role of transparency in quantitative analysis?

Answer 54

To clarify the strengths and weaknesses of the evidence used in analysis.

Question 55

What does 'reversion to the mean' refer to in statistics?

Answer 55

The phenomenon where extreme values in a dataset tend to be followed by values closer to the average.

Question 56

What is a critical takeaway regarding the relationship between correlation and causation?

Answer 56

Correlation does not imply causation.

Question 57

True or False: All quantitative evidence is free from bias.

Answer 57

False

Question 58

What does the term 'cosmic habituation' refer to?

Answer 58

The tendency of individuals to become desensitized to certain stimuli over time.