Intro to Organizational Behavior (OB): Perception and Decision Making - Decision Making Models, Rational Model Bounded Rationality Intuitive Decision Making

Decision Making Models (Rational Model, Bounded Rationality, Intuitive Decision Making)

What This Is

Decision-making models explain how individuals and organizations choose between alternatives. These models matter because managers must balance speed, accuracy, and stakeholder buy-in—poor decisions can waste resources, demotivate teams, or derail strategy. For example, Netflix’s shift from DVD rentals to streaming was a high-stakes decision: a purely rational analysis (data on declining DVD sales) combined with intuitive leaps (betting on broadband growth) and bounded rationality (limited foresight about piracy or competition from Disney+).

Key Theories & Models

• Rational Decision-Making Model (Herbert Simon, 1947): A prescriptive model assuming decision-makers are fully informed, objective, and logical. Steps: (1) Define the problem, (2) Identify criteria, (3) Weight criteria, (4) Generate alternatives, (5) Evaluate alternatives, (6) Select the best option. Practical implication: Useful for structured problems (e.g., Amazon’s warehouse automation—data-driven cost-benefit analysis of robotics vs. human labor). Limitation: Rarely achievable in real-world complexity.

• Bounded Rationality (Simon, 1957): Decision-makers satisfice (accept "good enough") due to cognitive limits, time constraints, and incomplete information. They use heuristics (mental shortcuts) to simplify choices. Practical implication: Explains why managers might pick the first viable option (e.g., Zappos’ early hiring decisions—Tony Hsieh admitted to hiring fast based on cultural fit, not exhaustive analysis). Tool: Use "pre-mortems" (imagining failure before deciding) to counter overconfidence.

• Intuitive Decision Making (Klein, 1998): Unconscious, pattern-based decisions relying on expertise and tacit knowledge. Often used in high-pressure or ambiguous situations (e.g., firefighters, ER doctors). Practical implication: Effective when data is scarce or time is critical (e.g., Southwest Airlines’ rapid gate turnarounds—pilots and crews rely on intuition honed by experience). Risk: Bias if intuition isn’t grounded in real expertise (e.g., a manager promoting a "gut feeling" hire who later fails).

• Prospect Theory (Kahneman & Tversky, 1979): People evaluate decisions based on gains/losses relative to a reference point, not absolute outcomes. They’re risk-averse for gains but risk-seeking for losses (e.g., preferring a sure $50 over a 50% chance of $100, but taking a 50% chance to lose $100 over a sure $50 loss). Practical implication: Framing matters—Google’s "20% time" policy (sold as a gain: "innovate freely") was more motivating than "lose 20% of your work time."

• Garbage Can Model (Cohen, March, & Olsen, 1972): Decisions emerge from chaotic interactions of problems, solutions, participants, and opportunities—often in organizations with unclear goals (e.g., universities, startups). Practical implication: Explains why Twitter’s (X) chaotic product changes (e.g., removing headlines from links) seem reactive—decisions are driven by shifting priorities, not strategic planning.

• Vroom-Yetton-Jago Decision Model (1973): A contingency model prescribing how much participation to use based on decision quality, team expertise, and time. Options: Autocratic (AI/AII), Consultative (CI/CII), or Group (GII). Practical implication: Patagonia’s environmental policies—Yvon Chouinard often used consultative (CI) or group (GII) decisions for sustainability initiatives to ensure buy-in.

Step-by-Step Application

How to choose the right decision-making model for a problem:

1. Assess the problem’s structure:
2. Structured (clear goals, data available)-Rational Model (e.g., Toyota’s just-in-time inventory).

3. Unstructured (ambiguous, novel)-Intuitive or Bounded Rationality (e.g., Apple’s iPhone launch—limited market data, high uncertainty).

4. Evaluate time pressure:

5. High-Intuitive (e.g., ER doctors diagnosing chest pain) or Bounded Rationality (e.g., Netflix’s 2011 price change—quick decision, backlash).

6. Low-Rational Model (e.g., Walmart’s supply chain optimization).

7. Check for cognitive biases:

8. Use pre-mortems (bounded rationality) or devil’s advocacy (rational model) to counter overconfidence or groupthink.

9. Example: Google’s "Project Aristotle" (studying team effectiveness) used data to debias intuitive assumptions about "ideal" team composition.

10. Determine stakeholder involvement:

11. High need for buy-in-Vroom-Yetton-Jago (consultative/group) (e.g., Microsoft’s shift to open-source under Satya Nadella).

12. Low need-Autocratic (e.g., Elon Musk’s "hardcore" work culture mandate at Twitter/X).

13. Pilot or prototype:

14. For high-risk decisions, test alternatives (e.g., Amazon’s A/B testing for website changes).

15. Reflect and learn:

16. Post-decision, analyze outcomes (e.g., Netflix’s post-mortem on the Qwikster failure—separating DVD and streaming services).

Common Misconceptions

• Misconception: The rational model is always best for organizations. Correction: It’s ideal for structured problems but impractical for complex, ambiguous, or time-sensitive decisions. Example: Blockbuster’s failure—a rational analysis might have shown declining DVD rentals, but bounded rationality (ignoring streaming) and intuition (dismissing Netflix as a "niche") led to bankruptcy.

• Misconception: Intuition is just guessing. Correction: Expert intuition is pattern recognition (e.g., Chesley "Sully" Sullenberger’s Hudson River landing—decades of experience enabled split-second decision-making). Novices relying on intuition, however, often fall prey to bias (e.g., a manager promoting a "charismatic" employee without performance data).

• Misconception: Bounded rationality means laziness. Correction: It’s a cognitive limitation, not a choice. Example: Zappos’ holacracy experiment—Tony Hsieh’s bounded rationality (limited understanding of self-management’s complexity) led to a costly, failed reorganization.

• Misconception: More data always improves decisions. Correction: Data overload can paralyze (analysis paralysis) or confirm bias. Example: Kodak’s digital camera failure—despite inventing the technology, they over-relied on film sales data and ignored market shifts.

• Misconception: Group decisions are always better than individual ones. Correction: Groups can suffer from groupthink (e.g., NASA’s Challenger disaster—engineers’ concerns were suppressed) or groupshift (e.g., Reddit’s WallStreetBets—collective risk-taking led to extreme trades). Use nominal group technique (silent idea generation + voting) to mitigate.

Exam / Case Interview Tips

1. Spot the model in disguise:
2. Rational Model questions often involve cost-benefit analysis (e.g., "Should we outsource IT?").
3. Bounded Rationality questions hint at time pressure or incomplete data (e.g., "The CEO needs a decision in 24 hours—what do you do?").

4. Intuitive questions involve expertise or crisis (e.g., "A cyberattack is underway—how does the CISO respond?").

5. Compare and contrast:

6. Rational vs. Bounded Rationality: "Why might a manager satisfice instead of optimizing?"-Time, cognitive limits, or uncertainty (e.g., Tesla’s Gigafactory locations—Elon Musk picked Nevada quickly despite other options).

7. Intuition vs. Rationality: "When should a leader trust their gut?"-High expertise + time pressure (e.g., Steve Jobs’ iPhone design—no market research, just intuition).

8. Avoid the "either/or" trap:

9. Real-world decisions often blend models. Example: Netflix’s content strategy—rational (data on viewer habits) + bounded (limited foresight on competitors) + intuitive (Reed Hastings’ bet on original content).

10. Link to other OB concepts:

11. Groupthink: "How might bounded rationality contribute to groupthink?"-Satisficing in groups can lead to premature consensus (e.g., Enron’s "mark-to-market" accounting).
12. Power dynamics: "How does the Vroom-Yetton-Jago model relate to leadership styles?"-Autocratic (AI) aligns with directive leadership; consultative (CI) with participative.

Quick Practice Scenario

Scenario 1: Your team is debating whether to launch a new product feature. The data is mixed—some users love it, others hate it. The CEO wants a decision by tomorrow. Which decision-making model(s) should you use, and why?

Answer: Use bounded rationality (satisfice due to time pressure) + intuitive (if the team has domain expertise). Example: Spotify’s "Discover Weekly"—limited data on user preferences led to a "good enough" algorithm tweak, refined later.

Scenario 2: A hospital administrator must decide whether to invest in a new MRI machine. The budget is tight, but patient wait times are increasing. Which model is most appropriate, and what steps should they take?

Answer: Use the rational model (structured problem, high stakes). Steps: (1) Define criteria (cost, patient outcomes, ROI), (2) Weight criteria, (3) Generate alternatives (lease vs. buy, different vendors), (4) Evaluate, (5) Decide. Example: Mayo Clinic’s data-driven equipment purchases.

Last-Minute Cram Sheet

1. Rational Model: 6 steps (define-criteria-weight-alternatives-evaluate-decide). Rarely achievable in real life.
2. Bounded Rationality: Satisficing + heuristics. Why? Cognitive limits, time, uncertainty. Example: Zappos’ fast hiring.
3. Intuitive: Expert pattern recognition. When? High pressure, ambiguity. Example: ER doctors, Southwest Airlines.
4. Prospect Theory: People hate losses > they love gains. Framing matters! Example: Google’s 20% time.
5. Garbage Can Model: Decisions = problems + solutions + participants + opportunities. Example: Twitter/X’s chaotic changes.
6. Vroom-Yetton-Jago: Autocratic (AI/AII), Consultative (CI/CII), Group (GII). Match to situation!
7. Groupthink-Groupshift: Groupthink = pressure for unanimity (e.g., Challenger); groupshift = extreme positions (e.g., WallStreetBets).
8. Pre-mortem: "Imagine this failed—why?" Counters overconfidence. Example: Google’s moonshot projects.
9. Devil’s Advocate: Assign someone to challenge assumptions. Example: Intel’s "constructive confrontation."
10. Blended Models: Most real decisions mix rational + bounded + intuitive. Example: Netflix’s streaming pivot.