A SMALL LAB FOR A BIG QUESTION
More options can expand possibility while shifting comparison work onto the chooser. Yet a larger, well-organized set can reveal a better fit than a narrow shortlist. Change the presentation, keep the options constant, and inspect what each design makes visible—or hides.
More options are not automatically a burden. When preferences are clear and the catalogue is legible, variety can improve fit.
A specialist may need twenty well-labelled options to find the one compatible feature that a tidy shortlist omitted.
A display changes what is easy to notice. It does not, by itself, establish what a person will choose.
EXPERIMENT 001 / CHANGE THE DISPLAY, NOT THE WORLD
This is a deterministic teaching model, not a psychological test. The attributes are invented ordinal values, not observed data. Use it to inspect presentation trade-offs, not to predict fatigue or choose a winner.
A group wants a recurring reading ritual without turning it into another obligation.
Members have different schedules, reading speeds, language preferences, and comfort with speaking in a group.
Select up to three options to keep in your private comparison tray. Values are illustrative ranks from 1–5; no universal score is calculated.
Scanning and noticing unusual items.
A clean, consistent comparison across every attribute.
No fatigue, regret, satisfaction or outcome is measured here.
Choose two options to see their attribute differences and missing context.
Bookmarks are temporary page state only. Nothing is sent or saved to an account.
Nothing here yet. Begin by bookmarking an option that is plausible for your context—not necessarily perfect.
Choose a format that works for the next two meetings; review after the trial.
Choose the easiest format that gives quieter members a real route to participate.
Local-only by design: your current selections stay in this browser tab. There is no analytics, account, tracking pixel, or remote submission. Exported notes are created on your device.
THE CATALOGUE / TWELVE DECISION ROOMS
The same display can be generous in one setting and burdensome in another. Pick a room to change the constraints, overlooked value, reversal route, and counterexample. These are invented scenarios, not reports about real people.
THE EVIDENCE DRAWER
The literature does not support a universal threshold at which choice becomes bad. Findings vary with task complexity, preference uncertainty, goals, and how outcomes are measured.
Iyengar and Lepper's 2000 paper reported three field and laboratory studies in which smaller assortments (six) sometimes produced more engagement than larger assortments (24 or 30). The result challenges the claim that more choice is always better; it does not establish a universal law.
CMS documented a redesign intended to make plan information easier to read and compare, including side-by-side comparison of up to three plans and a guided enrollment process. The announcement records design decisions and consumer testing; it is not independent proof that the redesign reduced overload or improved outcomes.
THE ADVERSARIAL EXAMPLE
A specialist knows what they need, has stable preferences, and can scan meaningful labels. Their sought-after option appears at position 17, outside the three-item shortlist. The catalogue wins because the long tail contains a real fit—not because more options are always better. A good interface makes that long tail navigable.
Clear, stable preferences weaken the claim that a larger set is inherently a burden. Sorting and useful labels may let someone locate a rare fit.
FIELD NOTES / THREE SHORT ESSAYS
Three complementary ideas: when more is genuinely better, why friction can be information rather than failure, and how reversibility creates room to act without perfect certainty.
A large set expands the reachable possibilities. It is particularly valuable when people's needs differ, the best match is unusual, or the chooser is still discovering what matters. A small menu can be a form of exclusion if it hides rare needs behind a supposedly sensible default. The design problem is not to minimize options at all costs; it is to help people find the relevant parts of a large set without requiring them to inspect everything equally.
A wider choice is useful only if the added possibilities are legible enough to be found and evaluated. Search, filters, categories, clear labels, and a path to see the full list can preserve variety while reducing unnecessary comparison. If the labels are inaccurate or the categories reflect the designer's convenience rather than the user's task, the interface merely moves the work around.
Slow choice can mean many things: incompatible requirements, missing information, high stakes, low confidence, fear of consequences, or simply care. Conversely, a quick choice may reflect expertise, a good default, indifference, exhaustion, or an avoidable nudge. Speed alone cannot tell these stories apart.
Hick's classic experiments linked choice uncertainty to reaction time under defined laboratory conditions. That is not a general law of personal decision-making. A consumer decision includes values, money, identity, social obligations, future consequences and sometimes incomplete or misleading information. Treating every pause as a defect can make a system faster for the designer while making it less accountable to the person choosing.
Some choices are not equally irreversible. A pilot, refundable booking, borrowed tool, draft, temporary schedule or two-week trial can reduce the cost of acting under uncertainty. Reversibility is not always available and it can shift risk onto someone else; a cancellation policy may burden the host, worker or service provider. Still, when genuine, a small reversible commitment changes the decision from “identify the perfect future” to “choose a defensible next step and learn.”
A good-enough rule is not a slogan about settling. It is a rule made explicit: what must be true, what can be traded off, who needs to agree, what would cause a review, and when the choice can be changed. The rule should be chosen by the people affected, not smuggled in as an algorithmic answer.
METHOD PLATE / 02
Option count and presentation are inputs to this exhibit. The model then describes an interface property: what becomes easy to scan, compare, or discover. It does not transform those inputs into a measured psychological outcome.
FIG. 02 — Input → presentation rule → visible trade-off ledger. A dotted path marks the inference the model deliberately refuses to make.
QUESTIONS FROM THE LAB
An honest exhibit needs boundaries. These answers distinguish the mechanism implemented here from claims that would require evidence beyond this page.
Not reliably in every setting. Reviews find substantial variation across studies. Choice-set size interacts with task difficulty, preference uncertainty, complexity, and the goal a chooser brings to the task. This exhibit does not measure fatigue.
To isolate presentation. In this demonstration, the underlying synthetic profiles remain the same while layout, comparison support and order change. Real interfaces can also alter which options are generated or excluded; that is a separate and important design choice.
They are useful contrast settings for a repeatable exercise, not empirically validated thresholds. The values allow you to inspect a narrow window, a middle window and a wider catalogue without pretending that a specific count is universally ideal.
No. A table supports comparison on known attributes, but it can flatten qualitative differences, make missing information look like a zero, and privilege the criteria selected by the designer. A narrative, demonstration or full catalogue may be more useful for other tasks.
A rule set before the choice that identifies acceptable conditions and a review point. For example: “Any venue with step-free access, an affordable fee and a cancellable booking is acceptable for a one-session pilot.” This is a user-owned rule, not an algorithmic winner.
Start with a clear question and a pre-specified outcome: task completion, choice deferral, errors, comprehension, confidence or later satisfaction, for example. They are not interchangeable by default. Define the population and context, compare equivalent tasks, and report uncertainty and counterevidence.
SOURCE SHELF / CHECK THE CLAIM
These sources support specific historical, methodological, or conceptual claims. They do not all support the same conclusion, and none validates this synthetic interaction as a psychological measure.
Three field and laboratory studies examining when larger assortments may be less motivating than smaller ones.
Reports a near-zero mean effect across 63 conditions with considerable variability across studies.
Reviews moderators including task difficulty, set complexity, preference uncertainty and decision goals.
Primary agency documentation of a guided, mobile-friendly plan comparison redesign; not an independent outcome evaluation.
A contrasting choice-architecture case about automatic enrolment and the influence of defaults in one employer's plan.
Classic constrained choice-reaction-time experiments; not a general measure of decision fatigue.
Primary research on stimulus information and reaction time in experimental tasks.
A foundational account of choice processes in relation to the structure of an environment.
Explores fast-and-frugal decision rules and the conditions in which simple heuristics may work well.
A major account of choice architecture and defaults, including a section addressing objections.
Bibliographic records and linked pages were checked on 28 October 2026. This check verifies the identity of sources and reachable records, not independent replication of every underlying result.
TAKE THE EXHIBIT OFFLINE
Seventeen pages of mechanisms, evidence, case studies, a fully worked interaction transcript, a counterexample, twelve decision rooms and printable worksheets. Designed to remain useful without the website.
Download the Field Guide