Decision-Making & Cognitive Biases

How people actually choose — for pricing pages, defaults, comparison UIs, and choice architecture. Pair with Persuasion, Ethics & Dark Patterns: every effect below can serve users or exploit them.

Try it — the decoy effect. Toggle the middle tier. With the decoy present, "Pro" reads like the obvious choice; without it, "Pro" has to argue on its own merits. (Frederick et al. 2014's caveat still applies: the effect is strongest in clean, comparable option sets like this one.)

Dual-process foundation

Kahneman (2011): fast, automatic System 1 does most everyday choosing; slow, effortful System 2 engages reluctantly. Interfaces are used in System 1 mode — design for the intuitive read, and deliberately slow users down (friction, confirmation) only for consequential, irreversible decisions.

The heavyweight effects

Defaults / status quo bias — People overwhelmingly keep preset options (Samuelson & Zeckhauser 1988; Johnson & Goldstein 2003 showed organ-donation consent varying enormously with opt-in vs. opt-out defaults). The default IS the design decision: make it the option that serves the user's typical interest, and never bury harmful defaults.
Loss aversion — Losses loom larger than equivalent gains (prospect theory, Kahneman & Tversky 1979). Framing matters: "keep your progress" outpulls "gain a feature." Caveat: magnitude is debated for small stakes (Gal & Rucker 2018) — assume it for meaningful losses, not trivial ones. Ethical use: warn about real losses (unsaved work). Abuse: fake scarcity.
Anchoring — First numbers seen bias estimates (Tversky & Kahneman 1974). Price tiers anchor on the highest; "was $X" anchors value.
Framing — Logically equivalent statements choose differently ("90% success" vs "10% failure"). Choose frames that inform, not distort.
Choice overload — More options can reduce purchase and satisfaction in some conditions (Iyengar & Lepper 2000 jam study). Meta-analysis (Scheibehenne et al. 2010) shows the effect is context-dependent — it bites when options are hard to compare, users lack expertise, or there's no dominant option. Remedies: curated defaults, "most popular" markers, comparison tables, filters, tiering (good/better/best).
Decoy / asymmetric dominance — An inferior third option shifts choice toward the option that dominates it (Huber, Payne & Puto 1982) — the mechanism behind classic 3-tier pricing. Caveat: robust with simple numeric attributes, but Frederick, Lee & Baskin (2014) found it often vanishes (or reverses) with realistic products — A/B test your decoy; don't assume it.
Scarcity — Scarce items are valued more (Worchel, Lee & Adewole 1975; Cialdini's scarcity principle), and limited stock/time cues measurably drive purchases — which is exactly why fabricated scarcity is among the most common deceptive patterns online (Mathur et al. 2019). Show real inventory and real deadlines only; detected fakery destroys trust.
Hyperbolic discounting — Present rewards are overweighted vs. future ones (Ainslie 1975; Laibson 1997) — why "start free today" works and why commitment devices (annual plans, scheduled sends) help users follow through on their own goals.
Sunk-cost & endowment effects — Users overvalue what they've invested in or "own" (Thaler 1980) — customization and saved data increase retention legitimately; holding data hostage crosses the line.
Social proof — Others' behavior guides choice under uncertainty (Cialdini 1984). Real reviews, usage counts, and testimonials inform; fabricated activity ("3 people are looking at this") deceives.
Peak–end rule — Experiences are remembered by their peak moment and their ending (Kahneman et al. 1993; Fredrickson & Kahneman) — invest in the flow's best moment and its final screen (confirmation, success state); a rough checkout ending taints a great shopping session.

Choice architecture principles (Thaler & Sunstein, Nudge)

Good defaults, expect error, give feedback, map choices to outcomes users understand, structure complex choices. Their own standard: nudges must be transparent and easy to opt out of ("libertarian paternalism") — a useful internal test for any persuasive pattern.

Friction as a design tool (and sludge)

Friction is a dial, not a defect. Deliberate, well-placed friction slows System 1 for consequential/irreversible actions: type-the-name-to-delete, cooling-off delays on large transfers, "are you replying to everyone?" checks, prompts to read an article before sharing it. The inverse — sludge — is friction that blocks the user's interest: cancellation mazes, rebate hoops, buried opt-outs (Thaler 2018; Sunstein 2021). Test: friction that protects the user's own stated goal is design; friction that protects only your metric is sludge (see Persuasion, Ethics & Dark Patterns).

Replication-status notes (be honest about the evidence)

Sturdy under large replication efforts: anchoring, framing, default effects, sunk-cost framing (e.g., Many Labs; Klein et al. 2014).
Context-dependent, test before relying on: choice overload (above), decoy effect (above), loss aversion at small stakes (above).
Do not build on: subtle/incidental social and behavioral priming (money priming, elderly-walking priming, most "prime the user with imagery X" claims) — headline effects failed high-powered replication (Doyen et al. 2012; Klein et al. 2014; Vadillo, Hardwicke & Shanks 2016). Perceptual/repetition priming (recognizing recently-seen items faster) remains robust and is a different phenomenon.

Sources

Kahneman, D. (2011). Thinking, Fast and Slow. FSG.
Kahneman, D. & Tversky, A. (1979). "Prospect theory." Econometrica, 47(2); Tversky & Kahneman (1974). "Judgment under uncertainty." Science.
Johnson, E. J. & Goldstein, D. (2003). "Do defaults save lives?" Science, 302; Samuelson & Zeckhauser (1988). J. Risk & Uncertainty.
Iyengar, S. & Lepper, M. (2000). JPSP, 79(6); Scheibehenne, Greifeneder & Todd (2010). "Can there ever be too many options?" JCR, 37(3).
Huber, Payne & Puto (1982). JCR, 9(1); Frederick, Lee & Baskin (2014). "The Limits of Attraction." JMR, 51(4). Thaler, R. (1980). JEBO.
Worchel, Lee & Adewole (1975). JPSP, 32(5); Mathur, A. et al. (2019). "Dark Patterns at Scale." Proc. CSCW.
Gal, D. & Rucker, D. (2018). "The Loss of Loss Aversion." J. Consumer Psychology, 28(3). Ainslie (1975). Psych. Bulletin; Laibson (1997). QJE, 112(2).
Thaler, R. (2018). "Nudge, not sludge." Science, 361; Sunstein, C. (2021). Sludge. MIT Press.
Klein, R. A. et al. (2014). "Many Labs" replication project. Social Psychology, 45(3); Doyen et al. (2012). PLoS ONE, 7(1); Vadillo, Hardwicke & Shanks (2016). Psych. Bulletin, 142(5).
Cialdini, R. (1984; rev. 2021). Influence. Harper.
Kahneman, Fredrickson, Schreiber & Redelmeier (1993). "When more pain is preferred to less." Psychological Science, 4(6).
Thaler, R. & Sunstein, C. (2008). Nudge. Yale UP.