Decision Making: Rationality, Bias and Heuristics

Understanding the concept

Managerial decisions are made with limited time, incomplete information and competing goals. A rational model offers a disciplined sequence, but bounded rationality explains why actual decisions often simplify the search. The aim is not perfect calculation; it is a transparent, revisable judgement.

Decision-making models range from the fully rational model, which assumes complete information and optimization, to bounded rationality, where managers satisfice under time, cognitive and information constraints, to intuitive decision-making based on pattern recognition from experience. Heuristics speed up decisions but can introduce systematic biases (anchoring, availability, confirmation bias, escalation of commitment), which are especially risky in high-stakes capital decisions.

Heuristics can reduce effort but create predictable blind spots. Anchoring may give early figures excessive influence; escalation of commitment may keep a failing project alive because resources have already been spent. Intuition can reflect experience, but its reliability depends on the setting and quality of feedback. A weighted matrix makes priorities visible, yet its result remains conditional on the weights, ratings and available options. Numerical precision does not remove uncertainty.

Core components and theoretical lenses

Use the following distinctions to unpack the chapter's central idea. Each identifies a different aspect of the problem; their relationships matter as much as the labels themselves.

1. 01

   Rational decision model: define problem, generate all alternatives, evaluate systematically, choose optimum

2. 02

   Bounded rationality and satisficing under constraints

3. 03

   Intuitive judgment based on pattern recognition and experience

4. 04

   Common heuristics and biases: anchoring, availability, confirmation bias, sunk-cost/escalation of commitment

5. 05

   Risk preference and framing effects on choice

Conceptual constructs

A construct names an idea used to interpret behaviour. It is not directly interchangeable with a survey score or a single incident. In this chapter, the constructs below organize the diagnosis; examine how they could explain the observations before deciding what to change.

• Bounded rationality
• Heuristics and biases
• Escalation of commitment
• Risk preference
• Framing effect

Observable and measurable indicators

Indicators connect an abstract explanation to something that can be observed. Check the operational definition, reporting period and comparison group. A survey average can conceal differences between teams, and an administrative record can reflect willingness to report as well as underlying behaviour.

• Number of alternatives formally evaluated and documented before a capital decision (proxy for rationality vs shortcutting)
• Time and budget already sunk into a prior related project, tracked to check for escalation risk
• Diversity of information sources consulted (internal vs external, disconfirming vs confirming) before recommendation
• Presence of a formal pre-mortem or devil's advocate step in the decision process
• Variance between individual risk-preference statements and the option ultimately chosen by the group

Reading the case evidence

Separate what the case reports from what you infer. A figure can establish the scale of a problem without establishing its cause. Note what information is absent and how that gap affects your confidence in a recommendation.

Decision alternatives and trade-offs

Alternatives represent different priorities and assumptions. The advantages and limitations below are the existing module's analytical prompts. Compare them against the same evidence rather than choosing first and finding a theory afterward.

Evidence-based discussion

For capacity expansion, separate sunk costs from future consequences and test how rankings change under plausible assumptions. Compare new capacity, outsourcing, automation and remaining unchanged against the same criteria. Explain which risks are tolerable and which information could reverse the recommendation. The Decision Lab is a record of reasoning, not an instruction to obey the highest total.

Theory-linked interpretation from the module

The escalation of commitment literature warns that decision-makers often continue a failing or suboptimal course of action to justify prior investment and avoid appearing indecisive, which is precisely the psychological pressure described in the board's split — this cautions against Option A being chosen for that reason alone. Bounded rationality reminds us that the original new-plant decision was made under the information available eight months ago, and rational decision-making requires re-evaluating with current data (outsourcing lead time, automation option, revised demand forecast) rather than anchoring on the original plan. Option D operationalizes this by structuring a fresh, criteria-based comparison and a pre-mortem to counteract confirmation bias, and even if the board ultimately still prefers a plant-based solution, doing so through Option D produces a justification rooted in current value rather than sunk cost.

Your managerial judgement

Which option should the board choose for meeting the 40% capacity target, and how would you use decision-making theory to justify avoiding or embracing escalation of committed resources?

Write a recommendation that identifies a preferred alternative, uses the reported evidence, acknowledges uncertainty and specifies a review point. Another choice can be defensible when its assumptions and trade-offs are made explicit.

Possible consequences

The module offers the following projections for discussion. They depend on implementation and context; they are not guaranteed effects or verified forecasts.

• Short-term: continuing the new plant (Option A) locks in another 14 months of capital exposure before any capacity gain is realized
• Short-term: pivoting to outsourcing (Option B) captures demand fastest but exposes quality risk immediately
• Long-term: a structured re-evaluation (Option D) improves the quality and defensibility of large capital decisions going forward, not just this one
• Long-term: choosing automation (Option C) may require a follow-up decision later to close the remaining 5% capacity gap
• Long-term: writing off the sunk INR 3 crore, if the data supports it, protects the firm from a larger future loss, consistent with rational capital allocation principles

Reflection questions

1. What psychological and organizational pressures make it hard for the board to abandon the new-plant plan despite new data?
2. How would you distinguish between a legitimate reason to continue an investment and pure escalation of commitment?
3. What role should a formal pre-mortem or devil's advocate play in high-stakes capital decisions in your own organization?

Practice: Know → Analyse → Decide

These are the original module questions. Reading or printing this textbook does not change quiz scores, completion or certificate eligibility.