Organizational Behavior — biases, nudges & better decisions
Smallest objective by weight, but very "free-point" friendly: almost every question is spot the bias → justify the precise error → recommend a specific fix, set in a case-study situation. Learn the bias vocabulary cold, then drill the trainer at the bottom until the justification is automatic.
6
readingsMINDSPACE · Short↔Long term · Bazerman Ch12 · Capital Bias · Hardy Ch19 · Deciding How to Decide
35+
named biasesHardy & Saunders gives the master list — 3 families. Others' readings reuse the same names.
3-step
answer patternName it → show the precise error in this situation → specific remedy (from a reading).
How Behaviors Impact Your Corporate Decision-Making — MINDSPACECP311-112-26 · LO 4a, 4b
What you must get out of it: De-biasing (training people to think more rationally) only fixes the slow conscious half of the brain. Behavioural economics instead changes the context so different automatic processes kick in. MINDSPACE = the 9 most robust effects on behaviour, usable as a checklist to audit a decision (and to influence others / exploit competitor biases).
System 1 — automatic
Fast, intuitive, emotional, subconscious. Heavily shaped by context, past experience and biology. Where most biases live.System 2 — deliberate
Slow, conscious, effortful; reason dominates. Often fails to correct System 1 because decision-makers don't notice how System 1 is steering them.Conventional de-biasing = "changing minds"
Train people to recognise biases → think more logically. Aims to make existing decision strategies more accurate. Insufficient — only touches System 2.Behavioural economics = "changing contexts"
Engineer the decision environment so intuition is directed to better outcomes ("strategizing with biases"). Reframes the decision so different automatic strategies are recruited.The 9 effects — with the article's diagnostic question & remedy
MMessengerWe're heavily influenced by who communicates the information.
Q: Is the decision swayed by who provided the info (even non-stakeholders)?
Fix: Separate the message from the messenger's credibility.
Fix: Separate the message from the messenger's credibility.
IIncentivesResponses to incentives follow predictable shortcuts, e.g. strongly avoiding losses.
Q: Drawn to the easy answer — quick wins, lower-risk options?
Fix: Check how solutions are framed; seek alternative ways to tackle the problem.
Fix: Check how solutions are framed; seek alternative ways to tackle the problem.
NNormsStrongly influenced by what others do.
Q: Is the option chosen because it's non-controversial / what peers do?
Fix: Challenge whether you're aligning with peers or worried about appearances.
Fix: Challenge whether you're aligning with peers or worried about appearances.
DDefaultsWe "go with the flow" of pre-set options.
Q: Fallen back on the default, especially after time/money already spent?
Fix: Resist sunk-cost pull & easy options; think through the best course.
Fix: Resist sunk-cost pull & easy options; think through the best course.
SSalienceAttention goes to what is novel and seems relevant.
Q: Drawn to an option flagged as new/relevant (media, suppliers)?
Fix: Park the favourite; develop neglected options with different benchmarks, models, sources; compare.
Fix: Park the favourite; develop neglected options with different benchmarks, models, sources; compare.
PPrimingActs influenced by subconscious cues (words, sights, smells).
Q: Swayed by how the scene was set early (e.g. meeting those impacted)?
Fix: Sceptically review language, imagery and "staging effects".
Fix: Sceptically review language, imagery and "staging effects".
AAffectEmotional associations powerfully shape actions.
Q: Did decision-makers' emotions play a part?
Fix: Step away until feelings subside; revisit with an outside view.
Fix: Step away until feelings subside; revisit with an outside view.
CCommitmentsWe seek consistency with public promises and reciprocate acts.
Q: Was it steered toward keeping promises (to senior mgmt, customers)?
Fix: Imagine breaking the promise; build a story justifying a different approach.
Fix: Imagine breaking the promise; build a story justifying a different approach.
EEgoWe act in ways that make us feel better about ourselves.
Q: Was it guided by how decision-makers would look?
Fix: Examine the motive (self-serving?); ask others how they'd decide.
Fix: Examine the motive (self-serving?); ask others how they'd decide.
Salience example — Oxbridge
Elite UK law firms over-hire Oxford/Cambridge grads: the degree is a salient stereotype cue so those CVs get considered further. Suboptimal — misses "false negatives" from other schools. Clifford Chance adopted a "CV-blind" policy to break the subconscious reaction.Priming examples
- Words: describing a time you had power → feel more powerful/confident
- Sights: bigger food containers → eat more
- Smells: cleaner scent → keep table cleaner
- Business: proposal cover colour — blue raises acceptance, red lowers it
Memory hook"My Inner Nerd Doesn't See People's Automatic Choices Easily." Group them: who/what (Messenger, Incentives, Norms) · set-up (Defaults, Salience, Priming) · self (Affect, Commitments, Ego).
How it gets testedA scenario (e.g. Blue Jay Tire choosing a supplier the CEO's friend recommended; Blue Jay keeping the default vendor) → "Identify MINDSPACE effects at play and recommend how to counter them." Score by quoting the diagnostic question and giving the matching remedy applied to the facts. Also: "Why is de-biasing training alone insufficient?" → System 1 vs 2.
Traps
- MINDSPACE is a behaviour-change / checklist tool — don't describe it as a list of biases only; say it's used to design the context.
- Defaults ≠ Norms: defaults = pre-set option; norms = what others do.
- Generic fix ("be aware of bias") scores little — use the specific remedy.
Self-test: Why does the article say de-biasing is insufficient, and what does it propose instead?
De-biasing trains the conscious, slow System 2, but System 2 often fails to correct automatic System 1 because people are unaware of its influence. Instead, change the context (choice environment) so that different automatic processes are recruited toward better outcomes — e.g. MINDSPACE checklist, CV-blind hiring.Self-test: A Blue Jay Air exec pushes a new AI scheduling vendor "everyone is talking about at conferences." Which effects + fix?
Salience (novel, flagged by media/suppliers) and Norms (what peers are doing). Fix: set aside the favoured option, develop neglected alternatives with different benchmarks/models/sources and compare on objective criteria; challenge whether the choice is about appearing current.Tying Short-Term Decisions to Long-Term Strategy — the behavioral angleCP311-108-26 · also in Obj 1 → see Obj 1 page
What you must get out of it: McKinsey survey: outperformers make bold, long-term resource reallocations. The behavioural blockers are consensus-seeking → groupthink and loss aversion; the cures are more voices, active debate and rewarding "noble failures".
GGovernanceInfluential, involved leadership; CEO backed by a strong FP&A / strategy team; clear strategic direction to BUs.
PProcessesThe nimbler the better: 3–7 yr plan in ≤3 months, annual budget in ≤2 months; reallocate resources within the year.
AAnalyticsRigorous, standardized financial analysis; rank programs on financial outcomes — consistently.
DDecision makingDe-bias to pursue bold investments — fight groupthink & loss aversion.
Bias → effect
- Consensus across many execs stifles debate → groupthink
- Loss aversion (losses felt more than gains) → avoid bold bets with higher value than low-risk ones
- People hesitate to speak up, esp. against a senior leader
Fixes in the reading
- Breaking groupthink starts at the top
- Include more voices — research says decisions improve
- Reward noble failures: courageous, responsible, well-executed initiatives that miss but teach
- Take "the right level of risk"; invest for the long term (≈2× more likely to outperform on growth & ROIC)
Memory hookGPAD — "Good Plans Are Decisive." Behaviour lives in the D.
How it gets testedCase company's planning takes 6 months and every exec must agree → "Assess the planning process and recommend improvements." Hit: process too slow (benchmarks 3/2 months, in-year reallocation), consensus → groupthink/loss aversion, recommend more voices, debate, reward noble failures, standardized analytics.
TrapsDon't recommend "get everyone to agree" — the reading says consensus-seeking is the problem. Bold ≠ reckless: still requires rigorous analytics.
Self-test: What is a "noble failure" and which bias does rewarding it counter?
A courageous, responsible, well-executed initiative that doesn't achieve its goals but provides valuable lessons. Rewarding it counters loss aversion (and encourages bold investment).Improving Decision Making — 7 strategiesCP311-113-26 · Bazerman & Moore, Judgment in Managerial Decision Making, Ch 12
What you must get out of it: Decisions are routinely impaired by overconfidence and over-reliance on experience/intuition. Know the 7 interventions well enough to pick the right one for a situation and explain why it works.
Moneyball (Oakland A's)
One of the lowest payrolls, yet near top in wins — by replacing intuition with objective analysis. Typical mistakes found: overgeneralizing from personal experience / what's readily available, and overweighting recent performance vs long-term data. If closely watched baseball teams err like this, any organization can. Switching takes time and courage.| # | Strategy | Key content | Use when… |
|---|---|---|---|
| 1 | Use decision-analysis tools | Expected value; multi-attribute utility for complex choices; linear models (weight & add predictor variables) for repeated decisions (hiring, loans). Beat intuition because people know which factors matter but don't weight them properly or consistently; complex models add little. Why resisted: people want their gut. Interviews overrated due to affect heuristic (quick feelings on superficial traits), representativeness (speaks well ⇒ performs well), correspondence bias (take interview answers at face value, ignore years of work in a GPA). If you must interview, use structured questions and feed them into the model. | Routine, repeatable judgments |
| 2 | Acquire expertise, not just experience | Feedback improves judgment only if accurate & immediate (learning to corner a car). With randomness & delays, outcomes mislead. Analyse the decision process, not just the outcome. | Noisy, delayed-feedback domains |
| 3 | De-bias your judgment | Unfreeze (show old strategy fails — e.g. quiz, "invisible gorilla"; obstacles: status-quo satisfaction, risk aversion, preferring known outcomes) → Change (3 steps: clarify the specific deficiency, explain its roots, reassure it's no threat to self-esteem; general tool: "consider the opposite") → Refreeze (routine check-ups of recent decisions so you don't revert). | Known recurring bias in a person/team |
| 4 | Reason analogically | Take common lessons from 2+ situations → more generalizable insight (well supported by research). | Learning transfer, training |
| 5 | Take an outsider's view | Insider view: "my situation is unique" → overconfident. Outsider: base rates across similar cases (home renos always run over). | Forecasting cost/time/success |
| 6 | Understand biases in others | 3 steps: (1) perceive/analyse the context; (2) identify potential biases of the decision & decision-makers; (3) make the logical adjustment. (XP example: over-weighting last exam's topics.) | Using others' forecasts/advice |
| 7 | Nudge wiser & more ethical decisions | Choice architecture/defaults. Organ donation: opt-in ≈ 4% vs opt-out ≈ 96%. Same for retirement savings auto-enrolment. | Designing choices for others |
Memory hook"The Expert Doctor Always Offers Useful Nudges": Tools · Expertise · De-bias · Analogies · Outsider view · Understand others · Nudge. De-bias = UCR (Unfreeze, Change, Refreeze); Change = CER (Clarify, Explain, Reassure).
How it gets tested"Darwin's underwriters override the scoring model based on their interview impression — critique and recommend" → linear model beats intuition (consistency), interview biases (affect, representativeness, correspondence), structured interview as model input. "Snappy's CEO forecasts launch in 3 months" → outsider view. "Increase enrolment in a benefit" → opt-out nudge.
Traps
- Experience ≠ expertise — say why feedback is misleading in that setting.
- Linear models are for repeated decisions; one-off strategic bets need other tools.
- Nudges must be in the decision-maker's best interest (ethical) — mention it.
Self-test: List the three steps of "Change" in the de-biasing framework.
(1) Clarify the existence of specific judgmental deficiencies; (2) explain the roots of those deficiencies; (3) reassure that they aren't a threat to self-esteem. General strategy: "consider the opposite."Self-test: Why do linear models outperform expert intuition?
People generally know which variables matter but fail to weight them properly and consistently across repeated decisions; a linear model applies the weights rigorously and consistently. Research: more complex models add little marginal gain.Capital Bias — biases in capital planningCP311-105-25 · LO 4b
What you must get out of it: Three biases distort capital project decisions (products, equipment, acquisitions). Two cures recur: many inputs/experts and objective, consistent metrics to compare projects instead of narratives.
| Bias | What it is | Symptom | Fix |
|---|---|---|---|
| Optimism | Over-optimism despite evidence → overconfident estimates | Estimates systematically beat by reality; stories filling data gaps | Survey past performance (estimates vs actuals); decide on data, not narratives |
| Expert | Accepting one expert (or the last "hot" expert) without other sources on complex projects | The top engineer's projects always approved | Multiple experts / second opinions; diverse proposal sources; objective criteria (experts' future predictions ≈ random) |
| Narrow framing | Splitting broad problems into isolated decisions / judging on a single attribute; gain-vs-loss framing flips choices | Each sub-department argues with its own metric | Metrics aligned with company goals; evaluate all capital decisions together (portfolio) |
US Navy (narrow framing)
Huge capital budget decided piecemeal by sub-departments with varying metrics → created universal objective metrics (cost & energy reduction) → better frame to compare very different projects.Telecom (expert bias)
Top engineers' projects always won → criteria on risk-adjusted benefits vs cost; insights converted to objective data → faster, better capital decisions.Two process questions
- How are proposals submitted? — seek them from a diverse set of employees/departments (beats narrow framing & expert bias).
- How are proposals assessed? — replace expert & optimistic narratives with consistent, objective metrics.
Is it really a bias? (instructor examples)
- Refusing to swap $10 for $5 — not status-quo bias; it's a rational comparison.
- Repeating a study method that keeps failing — status-quo bias is plausible.
- CEO: "Don't worry about the well-financed entrant, we always win" — optimism bias (ignoring real uncertainty).
- Using a plumber is fine; relying on last year's lucky economic forecaster for a complex forecast — expert bias.
Memory hookO-E-N = "Over Eager Narratives." Cures: Many voices + one yardstick.
TrapsOptimism is only a bias if it ignores evidence. Expert advice isn't automatically a bias — it's a problem in complex/uncertain domains with a single source.
Self-test: Blue Jay Tire approves each plant upgrade separately using the sponsoring manager's payback estimate. Which biases + fixes?
Narrow framing (each decision isolated, single attribute — payback) and likely optimism (sponsor's own estimate) — fix: evaluate projects together with consistent goal-aligned metrics (risk-adjusted benefit vs cost); back-test past sponsors' estimates vs actuals; base decisions on data, not narratives.Psychology of risk — the master bias listCP311-106-25 · Hardy & Saunders, Quantitative ERM, Ch 19 §19.1–19.4
What you must get out of it: Definitions of every bias in the 3 families, group dynamics (cohesion vs conformity, groupthink), and the 18 ways to overcome bias. Exam style: "spot the bias, justify, offer a solution."
Rational model
Economics assumes expected utility of wealth (captures risk aversion). Real people are more complex, contradictory, irrational.Bounded rationality
Constraints (time, info, cognition) on people's ability to make strictly optimal decisions.Heuristics
Mental shortcuts — usually reasonable ("actuaries are smart"), sometimes lead to biases ("actuaries can't party").Cognitive vs social bias
Cognitive = systematic errors processing info (self-deception + info processing); you think you were rational but weren't. Social = distortion caused by social forces.Memory hook — the 3 families"Me, my Mind, my Mob": Self-deception (how I see me) · Information processing (how my mind handles data) · Social (what the mob/group does to me).
Self-deception starts with "over-": over-precision, over-confidence (over-estimation, over-placement) + imposter, self-attribution, optimism, hindsight, outcome.
Info processing "CFA-CARRG": Confirmation, Framing, Anchoring, Conservatism, Availability, Recency, Representativeness, Gambler's fallacy.
Self-deception starts with "over-": over-precision, over-confidence (over-estimation, over-placement) + imposter, self-attribution, optimism, hindsight, outcome.
Info processing "CFA-CARRG": Confirmation, Framing, Anchoring, Conservatism, Availability, Recency, Representativeness, Gambler's fallacy.
| Bias | Definition | Quick example | Antidote |
|---|---|---|---|
| ① Self-deception biases | |||
| Over-precision | Ranges for unknowns far too narrow | "Japan has 10–500 islands" (6,582); tight stress-scenario ranges | Widen intervals; outside data; plausible worst case |
| Overconfidence ("illusion of knowledge") | Believe ourselves smarter than we are; two forms ↓ | CEO skips due diligence | Metacognition; devil's advocate; Board challenge |
| Over-estimation | Own ability better than it objectively is | "We can launch in 3 months" | Back-test past estimates; outsider view |
| Over-placement | Believe you're better than peers, underestimating them | 95% think they're above-average drivers | Benchmark vs peers objectively |
| Imposter syndrome | Opposite: underestimate self, overestimate peers | Expert stays silent in meeting | Structured individual input; encourage debate |
| Self-attribution | Illusion of control: successes = my skill; failures = bad luck / others' bad faith | In a crisis, blame-shifting/denial | Process reviews; transparency |
| Optimism | Underestimate likelihood of adverse events; low-prob/high-severity under-managed | Firm treats 1% VaR tail as impossible → games capital rules | Plausible worst-case scenarios; past-estimate survey |
| Hindsight | "I knew it all along" — claiming you predicted an unpredictable event | Post-crisis everyone "saw it coming" | Document predictions in advance |
| Outcome | Judge a strategy by its outcome, not objectively | Unhedged put options expire worthless → "great strategy" | Evaluate the decision process & risk taken |
| ② Information-processing biases | |||
| Confirmation | Downplay contradictory info; overweight supportive info | Reading only favourable parts of a risk report | Devil's advocate; consider the opposite |
| Framing | Choice depends on presentation; positive frames win agreement | ERM pitched as "better decisions/value" vs "avoid harms" | Reframe — see if decision changes |
| Anchoring | Over-influenced by an initial piece of info even if worthless | First offer in a negotiation; first impression of a person | Avoid early solutions; individual prior evaluation |
| Conservatism | (Sub-type of anchoring; related to confirmation) reluctance to fully update beliefs on new info | Value investor ignores negative news | Formal re-assessment triggers |
| Availability | Probability judged by how easily examples come to mind; rare-but-newsy events seem common | Plane-crash headlines → overprice crash risk; unseen rare risks ignored | Long-term objective data |
| Recency | Decisions based only on recent experience | Last 2 good years drive the forecast | Full-cycle data |
| Representativeness | Judge A by similarity to known B / stereotype; "law of small numbers" | Small sample assumed to behave like large one | Base rates; credible sample sizes |
| Gambler's fallacy | Expect reversal after a run of independent events | 5 tails → "heads is due" (still 50%) | Statistical independence check |
| ③ Social biases (incl. loss-aversion family) | |||
| Groupthink | Excess value on conformity to the collective view | "Don't rock the boat"; only 1–2 options considered | Devil's advocate; outsiders; rotate members |
| Polarization | Group decision more extreme than any individual's start | Risk-taking group ends up taking even more risk | Debate; individual written views first |
| Endowment effect | Extra value placed on what you own (WTA > WTP) | Refusing a fair price for a legacy unit | Outside/independent valuation |
| Authority effect | Obey authority even against own beliefs/knowledge | Analyst signs off the CEO's numbers | Ethics & responsibility training; don't reward deference |
| Conformity / herding | Instinct to follow the herd; overrides own judgment | Students start cheating when peers do | Individual accountability; diversity |
| Information cascade | Sequential decisions pile onto the bandwagon ignoring contrary signals | Each insurer copies the last one's product | Independent analysis before seeing others' choices |
| Loss aversion | Losses felt more strongly than equal gains; hard to drop losing course | Keep funding a failing project to avoid booking the loss | Reframe; reward noble failures |
| Sunk-cost fallacy | Take more risk once some investment made | "We've spent $20M, can't stop now" | Decide on future cash flows only |
| Regret aversion | Fear of future regret inhibits deciding | Board won't approve any change | Plausible worst case of not acting |
| Disposition effect | Hold losers, sell winners (hope to recoup paper losses) | Portfolio manager keeps sinking bonds | Rules-based rebalancing |
| Ambiguity aversion | Prefer known uncertainty to unknown uncertainty (about knowledge, not risk) | Avoid new market with unknown odds | Scenario analysis to frame unknowns |
| Status quo bias | Prefer current state even when change is objectively optimal (safe harbour from regret) | Keeping an obsolete risk process | Reframe; plausible worst case of status quo |
| Inertia bias | Unwilling to decide → status quo passively adopted | Decision deferred meeting after meeting | Rules/deadlines for deciding |
| Hyperbolic discounting | Innate preference for immediate over delayed reward | $100 today over $110 tomorrow | Long-term metrics & incentives |
| Probability distortion | Small probabilities exaggerated or treated as zero ("unlikely" = impossible) | After a crash, odds of another felt far higher | Quantitative, objective analysis |
Also in the reading: information bias — requesting more and more information as a way to avoid deciding (caveat to technique 15). Loss aversion links to framing (negative frame) and endowment (losing what you own hurts more).
Cohesion × Conformity — which group decides best?
Low conformity
High conformity
High cohesion
✓ BEST — vigorous, respectful debateConsiders a range of options, flexible updating, acknowledges errors → best evidence-based decisions. Drifts right over time as friendships form.
Groupthink (comfortable clique)Single clique; nobody wants to disagree because conformity feels good.
Low cohesion
FactionsSplits into sub-groups making different, weak/ineffectual decisions; members resentful, withdraw.
Groupthink (authoritarian)Dominated by an authority/clique; no disagreement permitted; dissenters excluded; inadequate analysis.
Cohesion (working together to a common goal) = good. Conformity (pressure to support without dissent) = bad.
8 characteristics of groupthink (Janis)
- Illusion of invulnerability
- Collective rationalization (discount challenges, even evidence of failure)
- Belief in moral superiority
- "Us vs them" / stereotyping outsiders
- Groupthink is comfortable → pressure on doubters
- Self-censorship
- Illusory unanimity
- Mind-guards shield leaders from dissent
Warning signs
- Dissent discouraged / only outside plenary
- Unanimity celebrated
- Social misfits ostracized / excluded
- Homogeneous group (age, gender, background)
- Outsiders dismissed, mocked, demonized
- New members picked to conform, not for skill
- Members leave feeling pressed into agreement
- Dominated by 1–2 powerful, dissent-intolerant members
Memory hook — 8 groupthink signs"I R M U C S I M" → "Invincible Rationalizing Moralists Unite, Comfortably Silent, Illusorily Monolithic" (Invulnerability, Rationalization, Moral superiority, Us-vs-them, Comfortable, Self-censor, Illusory unanimity, Mind-guards).
19.4 — 18 ways to overcome bias
- Metacognition — think about thinking (overconfident CEOs may find it intolerable)
- Devil's advocate — appointed dissenter; signals dissent is permitted
- Board challenges the CEO — "what could go wrong?"; directors should bear consequences (D&O moral hazard)
- Rules for decisions — written input before/after; document alternatives & reasons rejected
- Avoid early solutions — they anchor
- Change group membership — comfort breeds conformity
- Involve outsiders — SMEs, independent critical review
- Encourage debate — "don't shoot the messenger"
- Discuss contentious issues individually — but not as mind-guarding
- Avoid narcissists — overconfident, demand deference, manipulative
- Promote diversity — education, experience, gender, nationality, neurodiversity (cuts availability bias & groupthink)
- Emphasize ethics & responsibility — train against herding / deference
- Evaluate ideas individually in advance — avoid anchoring
- Realistic time — don't squeeze dissenting views
- Sufficient objective data — but beware information bias
- Reframe decisions — expose loss aversion / framing
- Transparency test — front page of tomorrow's newspaper?
- Plausible worst case — of the decision and alternatives
Memory hook — group the 18Think (1 metacognition, 16 reframe, 17 newspaper, 18 worst case) · Challenge (2 devil, 3 Board, 7 outsiders, 8 debate) · Structure (4 rules, 5 no early solutions, 13 individual prior views, 14 time, 15 data) · People (6 rotate, 9 one-on-ones, 10 no narcissists, 11 diversity, 12 ethics).
How it gets tested"(1 pt) Joe cites last week's plane crash to exclude crash cover from life policies. Describe the bias, justify." → Recency or availability — both accepted if justified (over-weighting a vivid recent event versus the very low long-term frequency and immaterial mortality impact). Then "recommend a way to overcome" → objective long-term data / outside actuarial review.
Traps
- Justify = show the precise error in the decision, not restate the definition.
- Not every optimism is bias (Sarah prepared well → reasonable).
- Cohesion is good; don't recommend "reduce cohesion" — reduce conformity.
- Risk aversion ≠ ambiguity aversion; loss aversion ≠ risk aversion.
Self-test: Distinguish over-estimation and over-placement.
Over-estimation: believe your ability is better than it objectively is. Over-placement: believe, without foundation, you're better than those around you (underestimating peers). Both are forms of overconfidence.Self-test: What is polarization and why does it happen?
A group's final decision is more extreme than any member's initial position, because members seek reasons to agree with the initial group position and ignore reasons to disagree — evidence for the initial stance gets exaggerated (risk-taking group takes even more risk).Self-test: Which quadrant tends to slide into groupthink, and why?
High cohesion / low conformity drifts to high cohesion / high conformity if kept together long — friendships form and members become less willing to disagree. Fix: change group membership.Deciding How to Decide — the behavioral angleCP311-111-26 · tools & 5 situations on Obj 3 page
What you must get out of it: Choosing the right decision tool depends on (1) do you have a causal model? (2) can you predict outcomes / probabilities? (3) is information centralized? — but human biases corrupt the diagnosis itself.
1. Executives don't know what they don't know
Most underestimate the uncertainty they face → pick tools for a more certain world.2. Cognitive bias creeps in
Biased assessment of uncertainty points them to the wrong tool (overconfidence, over-precision).3. Organizational processes get in the way
Need general decision protocols — political/behavioural pitfalls when money or power at stake.4. Reliance on a single tool
Over-use of DCF / simple scenarios even in ambiguous contexts; often need combinations.5. Ignoring the option to delay
"Deciding when to decide" matters; under high uncertainty use learning-based, iterative experimentation.Quick tool mapCausal model + predictable → capital budgeting (DCF/NPV). Causal + range & probabilities → quantitative multi-scenario (Monte Carlo, decision analysis, real options/pilots). Causal but can't predict → qualitative scenarios + case-based. No causal model → case-based decision analysis (authors' favourite: similarity-weighted average of analogous cases). Dispersed info → aggregation: Delphi, prediction markets (need a specified outcome range), incentivized estimates, similarity-based forecasting.
Memory hookThe 5 complications: "Unknown Biases Obstruct Single-tool Deciders" — Uncertainty underestimated, Bias, Org processes, Single tool, Delay ignored.
How it gets testedCombined with Obj 4: "The CEO insists on a 10-year DCF for entering a brand-new market and wants a decision this week — critique." → underestimating uncertainty / overconfidence, single-tool reliance, no causal model → qualitative scenarios + case-based analogies, consider delaying / piloting (real option).
TrapsPrediction markets only work when the range of outcomes can be specified. Case-based analysis weights more similar analogies more heavily.
Self-test: Name the two alternatives to information markets and how they work.
Incentivized estimates — people with diverse info estimate an outcome; the closest gets a payoff. Similarity-based forecasting — people rate how similar a decision/asset is to past ones; ratings aggregated statistically to forecast revenue, time or cost.Spot-the-bias trainerPick the bias, then read the model justification & fix. Some items accept two answers; some have no real bias.
How to use: Say your justification out loud before clicking — the marks are in the "precise error" sentence, not the label. Buddy mode: one reads the scenario, the other answers in the 4-part template.
Score: 0 / 0
Answer template for bias questions
- Name the bias (one, maybe two if both defensible).
- Precise error in THIS situation — "X is over-weighting/ignoring [specific fact] relative to [what the evidence actually says]."
- Consequence — what goes wrong for the company (mispricing, under-reserving, missed opportunity, value destroyed).
- Specific remedy from a reading, tied to the facts (devil's advocate at the pricing committee; back-test past estimates; opt-out default; outside review…).