Cognitive Bias in Idea Generation: How Bias Shapes Strategic Thinking

Last Updated June 4, 2026

Cognitive bias in idea generation refers to the systematic ways in which mental shortcuts, prior beliefs, emotional preferences, social dynamics, and institutional routines shape the production of ideas before formal evaluation begins. In strategic ideation, cognition does not operate on a neutral landscape. The generation of ideas is structured from the outset by selective attention, memory constraints, associative processes, implicit expectations, dominant frames, and social signals about what is realistic, safe, useful, or acceptable.

This matters because much of the literature on bias focuses on judgment and decision-making after options have already appeared. That emphasis understates a more fundamental strategic problem. Bias exerts some of its strongest influence upstream, during the formation of the idea space itself. If certain possibilities never emerge, no amount of downstream analysis can recover them. Strategic failure therefore often begins not only in poor evaluation, but in constrained ideation.

At its deepest level, this means that creativity is never fully open. The possibilities a team can see are filtered in advance by cognitive accessibility, familiarity, institutional memory, emotional comfort, hierarchy, category habits, and prior commitments. What feels like free ideation is often a highly structured search process whose constraints remain invisible to the people inside it.

Once this is recognized, the task of improving strategy changes. It is not enough to evaluate ideas more rigorously after they have been generated. Organizations must also design ideation systems that widen what can appear in the first place. Bias mitigation is therefore not only a psychological concern. It is part of strategic infrastructure.

This article examines cognitive bias in idea generation as a core problem in strategic ideation. It explores how bias operates at the level of possibility, how attention and memory shape the idea space, how specific biases narrow search, how group and institutional dynamics reproduce inherited assumptions, how pseudo-innovation emerges, and how disciplined ideation systems can expand the range of thinkable strategic options.

Bias Operates at the Level of Possibility

Strategic ideation is often imagined as an open field of creativity. In reality, it is structured before it begins. Cognitive processes determine what is noticed, what is retrieved from memory, which associations are activated, which analogies feel plausible, which constraints seem real, and which futures appear realistic. These processes act as filters that define the boundaries of the idea space.

This has a critical implication: bias does not merely distort choices. It constrains possibility. Decision-makers do not evaluate all options equally because they do not generate all options equally. They generate a subset shaped by cognitive accessibility, familiarity, prior framing, emotional salience, and institutional cues. What is absent from thought never reaches evaluation at all.

This upstream effect is one reason weak strategies can appear rational. A leadership team may apply disciplined analysis to the ideas it has in front of it, but the option set itself may already be narrow. The problem is not necessarily bad reasoning over available alternatives. It may be that the alternatives were biased before judgment began.

Bias at the level of possibility also explains why some organizations keep producing familiar strategies despite repeated calls for innovation. Their ideation processes may invite creativity while preserving the same frames, metrics, examples, language, and authority structures. The result is a controlled form of variation: enough novelty to feel fresh, but not enough to challenge the underlying model.

This connects directly to Heuristics in Strategic Ideation, where shortcuts guide search, and to Mental Models in Strategic Thinking, where representation determines how reality becomes thinkable. Bias is not separate from these processes. It is one of the ways they operate.

Bias matters strategically because it does not wait until judgment. It helps determine what counts as thinkable from the outset.

Back to top ↑

The Cognitive Architecture of Ideation

Idea generation emerges from interacting cognitive systems rather than from one unified faculty called creativity. Several processes matter especially: attention, memory retrieval, association, simulation, emotional valuation, and social interpretation. These systems help the mind function under constraint, but they also shape what kinds of ideas appear first, most clearly, or most persuasively.

Attention determines which parts of the environment enter awareness. In strategy, attention is rarely neutral. It is directed by metrics, reports, leadership concerns, recent events, market narratives, technical constraints, stakeholder pressure, and organizational anxiety. What attention repeatedly selects becomes the raw material for ideation.

Memory retrieval activates available examples, precedents, analogies, failures, successes, and institutional stories. The ideas easiest to retrieve often feel more relevant than they are. Teams may overuse recent cases, memorable competitors, high-status exemplars, or prior internal initiatives simply because they are cognitively available.

Association links concepts through similarity, proximity, metaphor, emotional resonance, and learned patterns. Association allows creativity, but it also tends to remain near familiar clusters unless deliberately disrupted through lateral thinking, distant analogy, random entry, stakeholder rotation, or first-principles work.

Simulation allows decision-makers to imagine possible outcomes. Yet simulation is shaped by current beliefs, emotional comfort, and available mental models. Ideas that are easy to imagine may feel more feasible, while ideas that require unfamiliar institutional arrangements, new stakeholders, or different metrics may feel unrealistic even when they are strategically stronger.

These systems operate under constraint. Cognitive load, time pressure, incomplete information, reputational risk, and social expectation force reliance on simplifying mechanisms. Much ideation is therefore driven by fast, intuitive processes rather than slow, comprehensive reasoning. These processes are efficient but selective. They privilege ideas that are cognitively fluent: easy to generate, easy to understand, easy to justify, and easy to communicate. Fluency, however, is not a reliable indicator of strategic value.

The ideas that arrive most smoothly are not necessarily the ideas most worth pursuing.

Back to top ↑

Together, these biases act less like isolated errors and more like a shaping force that bends idea generation toward the familiar, the available, and the socially safe.

Back to top ↑

Bias as Adaptive Constraint

Cognitive bias is often treated as error. This interpretation is incomplete. Bias is also an adaptive response to constraint. Without shortcuts, ideation would become computationally intractable. Real thinkers cannot search infinitely, remember everything, evaluate all possible analogies, or suspend all prior belief while generating ideas.

Herbert Simon’s theory of bounded rationality is central here. Decision-makers operate under limits of information, time, attention, and cognitive capacity. They satisfice rather than optimize. This principle applies to ideation as much as to decision-making. Individuals and teams generate ideas that are sufficient under the circumstances rather than exhaustive across the full possibility space.

Bias, in this sense, is partly the cost of remaining able to think and act at all under scarcity. Availability bias reduces search cost. Anchoring creates an initial structure. Framing simplifies ambiguity. Familiar categories help teams coordinate. Heuristics compress complexity. Social conformity reduces conflict and speeds alignment. These processes can be useful.

The problem is not the existence of bias, but its invisibility and miscalibration. When shortcuts operate without awareness, they become structural constraints on strategic thinking. What begins as an adaptive compression of complexity can harden into a narrowing of imagination. A shortcut that once saved time becomes a rule that prevents discovery.

This is why bias mitigation should not be framed as a demand that people become perfectly rational. That goal is impossible. The better goal is process design. Strategic ideation should use the mind’s shortcuts where they help and interrupt them where they narrow the search too early.

Bias is dangerous not because the mind simplifies, but because simplification often hides itself as objectivity.

Back to top ↑

Bias and the Narrowing of Search

From a search perspective, ideation involves exploring a high-dimensional space of possibilities. Bias functions as a pruning mechanism, reducing the number of paths considered. Some pruning is necessary. Without it, ideation would never progress. But when pruning becomes too aggressive, search collapses into local search, exploring only slight variations of familiar ideas.

This creates a fundamental tradeoff between efficiency and completeness. Efficiency reduces search cost and moves the team toward a workable option. Completeness increases variation and exposes unfamiliar possibilities. Strategic ideation must balance both. Too little pruning produces noise. Too much pruning produces premature convergence.

The timing of narrowing matters. Early narrowing is especially damaging because it restricts the idea space before sufficient variation has emerged. Once a team converges around a narrow frame, later analysis may become increasingly rigorous while becoming less strategically useful. The process becomes disciplined inside a diminished search space.

This dynamic aligns with Divergent vs Convergent Thinking. Divergence is meant to expand the option space before convergence selects. Bias distorts this sequence by causing convergence to happen informally before the group recognizes it. Anchors, familiar examples, leadership cues, and social conformity can all narrow exploration while the team still believes it is brainstorming.

Many weak strategic outputs are not the result of bad reasoning over a rich set of possibilities, but of good reasoning applied to a search space that was narrowed too soon.

Back to top ↑

Institutions do not merely evaluate ideas with bias. They often manufacture idea spaces already shaped by inherited assumptions.

Back to top ↑

The Illusion of Innovation

Bias can produce apparent novelty without structural change. Ideas may differ superficially while preserving the same underlying assumptions. A new platform may reproduce the same incentive logic. A new service channel may preserve the same administrative burden. A new product feature may sit atop the same weak mental model of the user. A new strategy may adopt new language while retaining old power relations.

This leads to pseudo-innovation: visible variation within a fixed conceptual frame. Organizations believe they are innovating while reproducing existing models under a different surface form. This is particularly common in mature industries and large institutions, where dominant narratives shape both perception and ideation.

Bias does not always block change outright. Sometimes it permits only the kind of change that leaves deeper assumptions intact. An organization may welcome innovation that improves efficiency but resist innovation that questions value. It may welcome new tools but resist new governance. It may welcome stakeholder research but resist shifting decision authority. It may welcome data but resist changing what is measured.

This is why novelty is not enough. The strategic question is whether an idea changes the underlying frame, mechanism, incentive, relationship, capability, or system behavior. An idea can look new and still be structurally conservative. Conversely, an idea can look modest but produce deep change if it alters a key feedback loop, rule, boundary, or mental model.

The most difficult biases to detect are often those that still allow movement, provided the movement stays within familiar conceptual walls.

Back to top ↑

Bias, Power, and Whose Ideas Count

Bias in idea generation is not only cognitive. It is also social and political. Some ideas become thinkable because they align with authority, funding, professional status, institutional legitimacy, or dominant cultural assumptions. Other ideas are treated as unrealistic, emotional, disruptive, radical, impractical, or outside scope because they come from less powerful stakeholders or challenge existing arrangements.

This means bias determines not only what ideas appear, but whose ideas count. In organizations, senior leaders often define realism. In policy settings, expert categories may override lived experience. In technology settings, technical possibility may dominate social consequence. In sustainability contexts, affluent institutional perspectives may obscure burdens carried by marginalized communities, workers, ecosystems, or future generations.

Bias mitigation therefore requires more than cognitive debiasing. It requires participation, voice, and accountability. The question is not only “Which mental shortcut distorted this idea?” It is also “Who was not present when the idea space was formed?” “Whose knowledge was treated as anecdotal rather than strategic?” “Which constraints were defined by those who bear none of the burden?”

Strategic ideation becomes stronger when it treats marginalized, frontline, affected, and historically excluded perspectives as sources of structural insight, not as after-the-fact feedback. These perspectives often reveal assumptions that insiders cannot see because insiders benefit from the current frame or are habituated to it.

The politics of idea generation matters because bias often determines which futures are dismissed before they are allowed to become options.

Back to top ↑

Bias mitigation works best when it changes the structure of ideation, not only the attitude of the participants.

Back to top ↑

Bias, Cognition, and Strategic Judgment

Idea generation and decision-making cannot be separated. The quality of strategic judgment depends on the quality of the ideas entering evaluation. If ideation is constrained, analysis operates within a limited domain. Even rigorous evaluation cannot compensate for missing alternatives.

This has implications for governance. Ideation processes must be treated as part of strategic infrastructure. Bias at this stage affects outcomes across the entire decision system. A board, leadership team, public agency, product group, research organization, or planning unit may believe it is making rational choices while never realizing how narrow the option set was before deliberation even began.

Strategic judgment should therefore ask two questions. First, are we evaluating the available ideas well? Second, did we generate a sufficiently diverse, relevant, and structurally meaningful set of ideas in the first place? Most organizations emphasize the first question. Stronger strategic systems institutionalize the second.

This also changes the role of evidence. Evidence should not only be used to compare options. It should also be used to challenge the idea-generation process itself. What evidence did the current frame exclude? Which stakeholders were not consulted? Which analogies were not considered? Which assumptions shaped the search? Which possibilities were removed before evaluation?

Strategic judgment is only as strong as the possibility space it is permitted to examine.

Back to top ↑

Bias in Complex Systems

Complex systems intensify both the necessity and the risk of cognitive bias. As environments become more uncertain and interconnected, exhaustive reasoning becomes less feasible. Heuristics and biases therefore play a larger role. Teams rely more heavily on pattern recognition, familiar frames, intuitive analogies, and simplified causal stories because the alternative is paralysis.

At the same time, complex systems are less forgiving of simplification. Linear assumptions, static models, and narrow frames can produce significant errors when feedback loops, delayed effects, thresholds, adaptation, emergence, and second-order consequences matter. A shortcut that works in a simple environment may become dangerous in a complex one.

This is why bias awareness must be integrated with Systems Thinking in Ideation and Complex Systems and Strategic Uncertainty. Understanding how systems behave is essential for evaluating whether cognitive shortcuts are appropriate. A team must ask whether its ideas account for feedback, incentives, interdependence, path dependence, stakeholder adaptation, and unintended consequences.

Bias also affects which systems are seen as systems. A team may notice technical interdependencies but miss social ones. It may model financial flows but ignore ecological flows. It may see user behavior but ignore administrative burden. It may see immediate costs but ignore long-term fragility. In complex systems, biased attention determines not only which ideas appear, but which relationships are considered real.

In complex environments, the problem is not that minds simplify. It is that simple models are often trusted longer than the system can justify.

Back to top ↑

The danger of biased ideation is that it often feels orderly, efficient, and aligned while excluding the ideas most needed for strategic change.

Back to top ↑

A bias audit treats ideation as a designed system rather than a spontaneous creative event.

Back to top ↑

Mathematical Lens: Biased Search and Constrained Possibility

A stylized idea space can be represented as:

Biased search can be represented conceptually as:

Premature convergence can be represented as:

Idea-space diversity can be represented as:

Bias-adjusted ideation quality can be represented as:

The mathematical lens shows that biased ideation is not simply imperfect creativity. It is constrained search through a filtered possibility space.

Back to top ↑

Advanced R Workflow: Comparing Bias Profiles in Idea Generation

The R workflow below compares stylized ideation contexts across availability pressure, anchoring intensity, conformity pressure, framing rigidity, institutional lock-in, stakeholder diversity, exploratory variation, and evidence discipline. It is designed as a transparent diagnostic for comparing biased search conditions.

# Install packages if needed.
# install.packages(c("tidyverse"))

library(tidyverse)

# ------------------------------------------------------------
# R Workflow: Comparing Bias Profiles in Idea Generation
# Purpose:
#   Build stylized profiles across ideation contexts using
#   availability pressure, anchoring intensity,
#   conformity pressure, framing rigidity,
#   institutional lock-in, stakeholder diversity,
#   exploratory variation, and evidence discipline.
# ------------------------------------------------------------

contexts <- tibble(
  context = c(
    "Fast Familiarity-Biased Context",
    "Balanced Reflective Context",
    "Hierarchy-Driven Group Context",
    "Variation-Oriented Design Context",
    "Institutionally Locked Context"
  ),
  availability_pressure = c(0.84, 0.46, 0.58, 0.32, 0.70),
  anchoring_intensity = c(0.71, 0.38, 0.82, 0.29, 0.76),
  conformity_pressure = c(0.63, 0.34, 0.88, 0.27, 0.72),
  framing_rigidity = c(0.77, 0.41, 0.72, 0.31, 0.84),
  institutional_lock_in = c(0.66, 0.42, 0.78, 0.30, 0.90),
  stakeholder_diversity = c(0.34, 0.68, 0.30, 0.82, 0.28),
  exploratory_variation = c(0.28, 0.73, 0.22, 0.89, 0.24),
  evidence_discipline = c(0.48, 0.74, 0.52, 0.76, 0.46)
)

contexts <- contexts %>%
  mutate(
    bias_pressure =
      0.18 * availability_pressure +
      0.18 * anchoring_intensity +
      0.18 * conformity_pressure +
      0.18 * framing_rigidity +
      0.16 * institutional_lock_in -
      0.06 * stakeholder_diversity -
      0.06 * evidence_discipline,
    bias_adjusted_ideation_profile =
      -0.14 * availability_pressure -
      0.14 * anchoring_intensity -
      0.16 * conformity_pressure -
      0.14 * framing_rigidity -
      0.12 * institutional_lock_in +
      0.16 * stakeholder_diversity +
      0.20 * exploratory_variation +
      0.14 * evidence_discipline,
    premature_convergence_risk =
      anchoring_intensity * conformity_pressure * framing_rigidity,
    diagnosis = case_when(
      premature_convergence_risk >= 0.45 ~ "premature_convergence_risk",
      institutional_lock_in >= 0.80 ~ "institutional_bias_risk",
      stakeholder_diversity < 0.35 ~ "stakeholder_visibility_gap",
      bias_adjusted_ideation_profile >= 0.35 ~ "stronger_ideation_conditions",
      TRUE ~ "requires_bias_review"
    )
  )

print(contexts)

contexts_long <- contexts %>%
  pivot_longer(
    cols = c(
      availability_pressure,
      anchoring_intensity,
      conformity_pressure,
      framing_rigidity,
      institutional_lock_in,
      stakeholder_diversity,
      exploratory_variation,
      evidence_discipline
    ),
    names_to = "dimension",
    values_to = "value"
  )

ggplot(contexts_long, aes(x = dimension, y = value, fill = context)) +
  geom_col(position = "dodge") +
  labs(
    title = "Stylized Bias Dimensions in Idea Generation",
    x = "Dimension",
    y = "Value",
    fill = "Context"
  ) +
  theme_minimal(base_size = 12) +
  coord_flip()

ggplot(contexts, aes(x = reorder(context, bias_adjusted_ideation_profile), y = bias_adjusted_ideation_profile)) +
  geom_col() +
  coord_flip() +
  labs(
    title = "Bias-Adjusted Ideation Profile",
    x = "Context",
    y = "Profile Score"
  ) +
  theme_minimal(base_size = 12)

write_csv(contexts, "cognitive_bias_ideation_profiles.csv")

This workflow can be extended with real workshop data, independent idea submissions, stakeholder participation records, frame-rotation outputs, and portfolio review data. Its purpose is not to mechanize creativity, but to make biased search conditions visible enough for better strategic design.

Back to top ↑

Advanced Python Workflow: Simulating Biased Search and Idea-Space Narrowing

The Python workflow below simulates stylized ideation contexts over repeated steps, showing how anchoring, conformity, and framing rigidity narrow the explored space, while exploratory variation and stakeholder diversity preserve broader search.

# Install packages if needed:
# pip install pandas numpy matplotlib

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt

# ------------------------------------------------------------
# Python Workflow: Simulating Biased Search
# Purpose:
#   Compare ideation systems whose search breadth depends on
#   anchoring, conformity, framing rigidity, institutional lock-in,
#   stakeholder diversity, and exploratory variation.
# ------------------------------------------------------------

time_steps = np.arange(1, 31)

def simulate_context(
    anchor,
    conformity,
    rigidity,
    institutional_lock_in,
    stakeholder_diversity,
    variation,
    evidence_discipline,
    initial_state=0.30
):
    state = np.zeros(len(time_steps))
    state[0] = initial_state

    for t in range(1, len(time_steps)):
        expansion = (
            0.16 * variation +
            0.10 * stakeholder_diversity +
            0.08 * evidence_discipline
        )

        narrowing = (
            0.12 * anchor +
            0.12 * conformity +
            0.12 * rigidity +
            0.10 * institutional_lock_in
        )

        state[t] = state[t - 1] + expansion / 5 - narrowing / 8
        state[t] = np.clip(state[t], 0, 1.6)

    return state

familiarity_biased = simulate_context(
    anchor=0.71,
    conformity=0.63,
    rigidity=0.77,
    institutional_lock_in=0.66,
    stakeholder_diversity=0.34,
    variation=0.28,
    evidence_discipline=0.48
)

balanced_reflective = simulate_context(
    anchor=0.38,
    conformity=0.34,
    rigidity=0.41,
    institutional_lock_in=0.42,
    stakeholder_diversity=0.68,
    variation=0.73,
    evidence_discipline=0.74
)

variation_oriented = simulate_context(
    anchor=0.29,
    conformity=0.27,
    rigidity=0.31,
    institutional_lock_in=0.30,
    stakeholder_diversity=0.82,
    variation=0.89,
    evidence_discipline=0.76
)

institutionally_locked = simulate_context(
    anchor=0.76,
    conformity=0.72,
    rigidity=0.84,
    institutional_lock_in=0.90,
    stakeholder_diversity=0.28,
    variation=0.24,
    evidence_discipline=0.46
)

df = pd.DataFrame({
    "time": time_steps,
    "Fast Familiarity-Biased Context": familiarity_biased,
    "Balanced Reflective Context": balanced_reflective,
    "Variation-Oriented Design Context": variation_oriented,
    "Institutionally Locked Context": institutionally_locked
})

print(df.head())

plt.figure(figsize=(10, 6))

for col in df.columns[1:]:
    plt.plot(df["time"], df[col], label=col)

plt.xlabel("Ideation Step")
plt.ylabel("Search Breadth")
plt.title("Biased Search and Idea-Space Narrowing")
plt.legend()
plt.tight_layout()
plt.show()

df.to_csv("cognitive_bias_search_simulation.csv", index=False)

This simulation can be developed into a more serious workflow by using real ideation logs, idea diversity coding, stakeholder participation measures, frame-rotation outputs, and decision-memory records. The central lesson remains: search breadth improves when ideation systems deliberately counteract anchoring, conformity, framing rigidity, and institutional lock-in.

Back to top ↑

GitHub Repository

The companion repository for this article will provide advanced strategist-facing workflows for bias-profile diagnostics, idea-space diversity scoring, anchoring and conformity review, frame-rotation analysis, assumption mapping, stakeholder visibility checks, premature-convergence risk, and decision-memory records.

The repository structure is designed to support professional strategic analysis rather than generic coding demonstrations. The python/ folder can model availability pressure, anchoring intensity, conformity pressure, framing rigidity, institutional lock-in, stakeholder diversity, exploratory variation, evidence discipline, premature-convergence risk, and idea-space diversity. The r/ folder can compare bias profiles, visualize bias-adjusted ideation conditions, and flag contexts requiring review. The julia/ folder can support scenario-based sensitivity analysis for biased search and variation preservation. The sql/ folder can define schemas for ideation sessions, ideas, frames, biases, assumptions, stakeholders, evaluation rounds, decision memory, and intervention design.

Additional folders can support command-line diagnostics, lower-level scoring utilities, and reproducible documentation. The rust/ folder can provide a command-line bias diagnostics scaffold. The go/ folder can provide an idea-space diversity review utility. The cpp, fortran, and c folders can provide efficient scoring examples and low-level utilities. The docs, data, outputs, and notebooks folders can support article notes, modeling principles, synthetic datasets, generated outputs, and notebook placeholders.

This code should be understood as a transparent learning and modeling scaffold. It is intended for synthetic-data research, methods demonstration, institutional learning, strategic analysis, and reproducible workflow development. It is not a substitute for stakeholder engagement, ethical review, domain expertise, accountable governance, or participatory judgment.

Back to top ↑

Conclusion

Cognitive bias in idea generation demonstrates that strategic creativity is never unconstrained. It is shaped by cognitive processes, social dynamics, institutional routines, power relations, and inherited frames that influence which ideas emerge and which are excluded. The problem is not simply that bias leads to weak judgments after ideation. It is that bias often structures ideation itself.

The practical objective is not to eliminate bias. That is impossible. The objective is to understand, expose, interrupt, and redesign the conditions under which bias narrows the idea space. By separating generation from evaluation, using multiple frames, encouraging independent ideation, introducing distant analogies, surfacing assumptions, protecting dissent, rotating stakeholder vantage points, and preserving decision memory, organizations can improve the quality of their strategic thinking.

In complex environments, this becomes essential. Strategy depends not only on how decisions are made, but on how possibilities are first allowed to appear. A rigorous evaluation process applied to a biased and narrow idea pool can still produce weak strategy. The deeper work is to improve the architecture of ideation itself.

Bias-aware strategic ideation therefore requires humility. It recognizes that what seems obvious may be merely available, what seems realistic may be socially safe, what seems innovative may preserve old assumptions, and what seems impossible may simply be outside the current frame.

Strategic imagination improves when organizations stop treating the first visible idea space as the whole field of possibility.

Back to top ↑

Related articles

• Strategic Ideation
• Mental Models in Strategic Thinking
• Heuristics in Strategic Ideation
• Lateral Thinking in Strategy
• Analogical Thinking and Idea Transfer
• Divergent vs Convergent Thinking
• Problem Framing and Problem Definition
• Systems Thinking in Ideation
• Complex Systems and Strategic Uncertainty
• Risk, Tradeoffs, and Strategic Choices

Further reading

• Gigerenzer, G. (2007) Gut Feelings: The Intelligence of the Unconscious. New York: Viking.
• Gigerenzer, G. and Gaissmaier, W. (2011) ‘Heuristic decision making’, Annual Review of Psychology, 62, pp. 451–482. Available at: https://doi.org/10.1146/annurev-psych-120709-145346
• Kahneman, D. (2011) Thinking, Fast and Slow. New York: Farrar, Straus and Giroux.
• Kahneman, D. and Tversky, A. (1974) ‘Judgment under uncertainty: Heuristics and biases’, Science, 185(4157), pp. 1124–1131. Available at: https://www.science.org/doi/10.1126/science.185.4157.1124
• Simon, H.A. (1996) The Sciences of the Artificial. 3rd edn. Cambridge, MA: MIT Press. Available at: https://mitpress.mit.edu/9780262690232/the-sciences-of-the-artificial/
• Stanford Encyclopedia of Philosophy (no date) Bounded Rationality. Available at: https://plato.stanford.edu/entries/bounded-rationality/
• OECD (no date) Behavioural Insights. Available at: https://www.oecd.org/en/topics/behavioural-insights.html
• Stanovich, K.E. (2011) Rationality and the Reflective Mind. Oxford: Oxford University Press.
• Thaler, R.H. and Sunstein, C.R. (2021) Nudge: The Final Edition. New Haven: Yale University Press.

Back to top ↑

References

• Gigerenzer, G. (2007) Gut Feelings: The Intelligence of the Unconscious. New York: Viking.
• Gigerenzer, G. and Gaissmaier, W. (2011) ‘Heuristic decision making’, Annual Review of Psychology, 62, pp. 451–482. Available at: https://doi.org/10.1146/annurev-psych-120709-145346
• Kahneman, D. (2011) Thinking, Fast and Slow. New York: Farrar, Straus and Giroux.
• Kahneman, D. and Tversky, A. (1974) ‘Judgment under uncertainty: Heuristics and biases’, Science, 185(4157), pp. 1124–1131. Available at: https://www.science.org/doi/10.1126/science.185.4157.1124
• Simon, H.A. (1996) The Sciences of the Artificial. 3rd edn. Cambridge, MA: MIT Press. Available at: https://mitpress.mit.edu/9780262690232/the-sciences-of-the-artificial/
• Stanovich, K.E. (2011) Rationality and the Reflective Mind. Oxford: Oxford University Press.
• Thaler, R.H. and Sunstein, C.R. (2021) Nudge: The Final Edition. New Haven: Yale University Press.

Back to top ↑