Chapter 12 Quiz: Self-Assessment

Instructions: Answer each question without looking back at the chapter. After completing all questions, check your answers against the key at the bottom. If you score below 70%, revisit the relevant sections before moving on to Chapter 13.

Multiple Choice

Q1. The term "satisficing" was coined by:

a) Daniel Kahneman b) Gerd Gigerenzer c) Herbert Simon d) Gary Klein

Q2. Satisficing means:

a) Settling for an inferior option due to laziness or ignorance b) Searching for an option that meets a threshold of acceptability and stopping c) Comparing all available options and selecting the best one d) Randomly choosing among available options

Q3. Bounded rationality refers to:

a) The irrationality of human decision-making b) The limits on rational decision-making imposed by finite information, cognitive capacity, and time c) The boundaries between rational and irrational thinking d) The rational choice to ignore available information

Q4. The vertebrate retina is "wired backwards" (photoreceptors pointing away from incoming light). This is evidence that evolution:

a) Optimizes for long-term performance b) Makes random design choices unrelated to fitness c) Satisfices -- retaining designs that are good enough to function, even if not optimal d) Deliberately introduces flaws to promote future adaptation

Q5. Deep Blue defeated Garry Kasparov by:

a) Evaluating every possible chess position and selecting the optimal move b) Using heuristic evaluation and search pruning to find good-enough moves c) Memorizing all known chess games and replaying the best moves d) Using quantum computing to solve chess completely

Q6. The Shannon number (approximately 10^120) refers to:

a) The number of atoms in the observable universe b) The estimated number of possible chess games c) The number of neurons in the human brain d) The number of species that have ever existed on Earth

Q7. Von Moltke's Auftragstaktik (mission-type tactics) is an example of satisficing because:

a) It specifies the optimal action for every possible contingency b) It specifies objectives and leaves methods to subordinate commanders who adapt to local conditions c) It eliminates all uncertainty from military planning d) It relies entirely on luck rather than planning

Q8. Engineering tolerances are:

a) Statements of optimization -- specifying the single best value for a dimension b) Statements of satisficing -- specifying a range of acceptable values c) Evidence of poor engineering practice d) Only relevant in mass manufacturing, not precision engineering

Q9. The paradox of choice refers to the finding that:

a) People always prefer more options to fewer b) People with fewer options are unable to make good decisions c) Expanding options beyond a certain point decreases satisfaction and increases decision difficulty d) The best choice is always the most expensive one

Q10. According to Barry Schwartz's research, satisficers compared to maximizers tend to:

a) Make worse choices and be less happy b) Make similar-quality choices and be happier c) Make better choices and be equally happy d) Make no choices at all

Q11. The recognition heuristic works by:

a) Computing a weighted average of all available information b) Choosing the recognized option over the unrecognized one c) Recognizing the optimal choice through intuition d) Comparing all recognized options on multiple criteria

Q12. Gigerenzer's less-is-more effect shows that:

a) Less effort always produces worse results b) Simple heuristics using less information can sometimes outperform complex models using more information c) Less information is always better than more d) Complex models always outperform simple heuristics

Q13. The 1/N rule (allocate equally across N options) is notable because:

a) It is the mathematically optimal allocation strategy b) It uses zero information about individual options yet often matches or outperforms sophisticated optimization methods c) It only works when all options are identical d) It requires extensive computation to implement

Q14. Gary Klein's research on recognition-primed decision making found that experienced decision-makers:

a) Systematically compare multiple options before choosing b) Use optimization algorithms learned through training c) Generate a single option based on pattern recognition, test it mentally, and usually adopt it d) Make random choices and rationalize them afterward

Q15. Ecological rationality refers to:

a) Rationality about environmental issues b) The fit between a decision strategy and the structure of the decision environment c) The rational exploitation of ecological resources d) The ecological impact of rational decision-making

Q16. According to the chapter, the 80/20 rule (Pareto principle) supports satisficing because:

a) 80% of people are satisficers and 20% are maximizers b) The first 20% of effort captures roughly 80% of the value, making additional effort disproportionately costly c) 80% of decisions are trivial and only 20% matter d) Satisficing works 80% of the time and fails 20% of the time

Q17. Goodhart's Law states that:

a) Good measures always lead to good outcomes b) When a measure becomes a target, it ceases to be a good measure c) All measurements are equally valid d) The best metric is always the one that is easiest to measure

Q18. The chapter argues that overfitting (fitting data too precisely) is related to satisficing because:

a) Satisficing always leads to overfitting b) Satisficing is a natural defense against overfitting, because accepting a "good enough" solution avoids fitting noise c) Overfitting is the same thing as satisficing d) Neither overfitting nor satisficing is relevant to real-world decisions

Q19. The chapter's threshold concept -- "Bounded Rationality Is Not Irrationality" -- means:

a) Humans are perfectly rational b) Using heuristics and satisficing strategies under real-world constraints is a rational adaptation, not a failure of reasoning c) Irrationality does not exist d) Bounded rationality is inferior to unbounded rationality but acceptable as a compromise

Q20. Which of the following is the best summary of the chapter's central argument?

a) Optimization is always superior to satisficing, but satisficing is a practical necessity b) Satisficing is always superior to optimization in every situation c) In complex, uncertain, resource-constrained environments, satisficing is not a compromise but often the genuinely optimal strategy, and it appears independently across every domain examined d) Satisficing and optimization are identical strategies with different names

Short Answer

Q21. In two to three sentences, explain why the Schlieffen Plan's failure illustrates the brittleness of optimization. Use the concepts of "optimization" and "satisficing" in your answer.

Q22. Describe a specific example (not from the chapter) where a fast-and-frugal heuristic outperforms or could outperform a more complex decision method. Explain why the heuristic succeeds.

Q23. The chapter argues that there is no universally optimal decision strategy -- only an "adaptive toolbox" of strategies matched to environments. In your own words, explain what this means and why it is important.

Answer Key

Q1: c) Herbert Simon Section 12.2 -- Simon coined the term by combining "satisfy" and "suffice."

Q2: b) Searching for an option that meets a threshold of acceptability and stopping Section 12.2 -- Satisficing defines a threshold, searches until an option meets it, and stops.

Q3: b) The limits on rational decision-making imposed by finite information, cognitive capacity, and time Section 12.2 -- Bounded rationality is rationality within real-world constraints, not irrationality.

Q4: c) Satisfices -- retaining designs that are good enough to function, even if not optimal Section 12.3 -- The backwards retina works well enough; evolution had no mechanism to redesign it.

Q5: b) Using heuristic evaluation and search pruning to find good-enough moves Section 12.4 -- Deep Blue searched a subset of possibilities and evaluated positions heuristically, not exhaustively.

Q6: b) The estimated number of possible chess games Section 12.4 -- The Shannon number demonstrates that exhaustive search of chess is computationally intractable.

Q7: b) It specifies objectives and leaves methods to subordinate commanders who adapt to local conditions Section 12.5 -- Auftragstaktik defines "good enough" (the objective) and trusts commanders to find a workable method.

Q8: b) Statements of satisficing -- specifying a range of acceptable values Section 12.6 -- Tolerances define a range of "good enough" rather than demanding the single optimal value.

Q9: c) Expanding options beyond a certain point decreases satisfaction and increases decision difficulty Section 12.7 -- Iyengar and Lepper's jam study and Schwartz's research show diminishing and eventually negative returns from more options.

Q10: b) Make similar-quality choices and be happier Section 12.7 -- Satisficers experience less regret, less anxiety, and greater satisfaction despite similar objective outcomes.

Q11: b) Choosing the recognized option over the unrecognized one Section 12.8 -- The recognition heuristic uses a single cue (recognition) and ignores all other information.

Q12: b) Simple heuristics using less information can sometimes outperform complex models using more information Section 12.8 -- Less information can mean less noise; simple models avoid overfitting.

Q13: b) It uses zero information about individual options yet often matches or outperforms sophisticated optimization methods Section 12.8 -- Equal allocation avoids the estimation errors inherent in optimized allocation.

Q14: c) Generate a single option based on pattern recognition, test it mentally, and usually adopt it Section 12.9 -- Klein found that experts satisfice through recognition rather than comparison.

Q15: b) The fit between a decision strategy and the structure of the decision environment Section 12.12 -- A heuristic is ecologically rational when its assumptions match the environment's structure.

Q16: b) The first 20% of effort captures roughly 80% of the value, making additional effort disproportionately costly Section 12.11 -- Diminishing returns mean the satisficer captures most of the value at a fraction of the cost.

Q17: b) When a measure becomes a target, it ceases to be a good measure Section 12.10 -- Optimizing a proxy for the real goal distorts the proxy; satisficing avoids this by not optimizing any single metric.

Q18: b) Satisficing is a natural defense against overfitting, because accepting a "good enough" solution avoids fitting noise Section 12.10 -- Tolerances and satisficing thresholds build in margins that absorb variation, preventing overfitting.

Q19: b) Using heuristics and satisficing strategies under real-world constraints is a rational adaptation, not a failure of reasoning Section 12.12 -- Simon argued that bounded rationality is the correct model of rationality, not a deficient approximation of optimization.

Q20: c) In complex, uncertain, resource-constrained environments, satisficing is not a compromise but often the genuinely optimal strategy, and it appears independently across every domain examined Sections 12.12-12.13 -- The central argument: satisficing is a universal pattern that emerges from the structure of real-world decision problems.

Q21. Sample answer: The Schlieffen Plan was optimized for a specific set of assumptions about enemy behavior, troop movements, and logistical conditions. When reality deviated from those assumptions -- Belgian resistance, faster British deployment, quicker Russian mobilization -- the optimized plan shattered because it had no margin for adaptation. A satisficing approach, like Von Moltke's Auftragstaktik, would have specified objectives and allowed commanders to adapt to actual conditions, trading theoretical precision for practical robustness.

Q22. Sample answer: When a doctor in an emergency room sees a patient clutching their chest, sweating, and reporting radiating arm pain, she diagnoses a heart attack and begins treatment immediately rather than running a comprehensive battery of tests. The fast-and-frugal heuristic (recognize classic symptoms, act immediately) outperforms the more thorough approach because the cost of delay (potential death) far exceeds the cost of occasional false positives, and the classic symptom pattern has high diagnostic validity in the emergency room population.

Q23. Sample answer: No single decision strategy is best in all situations. A fast-and-frugal heuristic that works brilliantly in a noisy, time-pressured environment may perform poorly in a data-rich, low-stakes environment where careful analysis is possible. The skill of good decision-making is not mastering a single strategy but maintaining a repertoire of strategies -- an adaptive toolbox -- and knowing which tool to select for each situation. This matters because it shifts the focus from "what is the right way to decide?" to "what kind of decision environment am I in?" -- a question that is both more realistic and more productive.