The core laws of UX and UI design

None of what follows is a style opinion. Each finding comes from psychology or usability research, and most have held up for decades of retesting. Use the list as a diagnostic: point it at a design and ask which of these it is breaking, rather than treating it as a rulebook that replaces judgment. Applied together, these findings explain why some interfaces feel effortless and others feel like work before a single word of copy has been read.

Deciding: Hick's Law and Miller's Law

Hick's Law says that the time it takes to make a decision grows with the number and complexity of the choices on offer.¹ It comes from two linked studies: W. E. Hick's 1952 paper measuring reaction time against the number of stimulus-response options,² and Ray Hyman's 1953 follow-up confirming the relationship, which is why some writeups call it the Hick-Hyman law.³ A menu with four dishes gets ordered from faster than one with forty, even when the forty include the same four. Apply Hick's Law by cutting the choices shown at any one moment to the essential few, grouping what remains under clear labels, and moving rarely used options to a secondary screen instead of laying every option on the first one.

Miller's Law comes from George Miller's 1956 paper, which found that working memory holds roughly seven items at once, plus or minus two.⁴ The exact number gets quoted more than it deserves; later research puts the real figure closer to four for unrelated, meaningless items. What survives is not the number, it is the mechanism: chunking. People recall three groups of four digits far more reliably than twelve loose digits. A sixteen-digit card number is unreadable as one string and instantly parseable split into four groups of four. Apply Miller's Law by breaking long forms, phone numbers, and navigation menus into small labeled clusters instead of one undifferentiated list.

Expectation: Jakob's Law

Jakob's Law, named for usability researcher Jakob Nielsen, observes that people spend most of their time on other products, so they arrive at yours already expecting it to behave the same way.¹ A cart icon in the top right, a logo in the top left that links home, a search bar instead of a search paragraph: none of these are technical requirements, they are conventions every visitor already carries with them. A checkout flow that reorders the standard address fields forces every visitor to relearn something they already knew, for no benefit. Apply Jakob's Law by defaulting to the standard pattern for anything people are only passing through, such as navigation, forms, and account settings, and saving genuine novelty for the parts of the product that are actually new.

Targeting: Fitts's Law

Fitts's Law predicts that the time it takes to move a pointer to a target depends on the target's size and its distance: bigger and closer is faster, smaller and farther is slower.⁵ Paul Fitts published the underlying model in 1954 after timing how quickly people could tap targets of varying sizes at varying distances, and the relationship has replicated since with a mouse, a stylus, and a fingertip alike. A primary action button spanning 200 pixels is a faster, more reliable target than an 80-pixel link buried mid-paragraph. Screen edges and corners are the fastest targets of all, because a pointer physically cannot travel past them, which is why operating systems put menus at the very top of the screen and docks at the very edge. Apply Fitts's Law by sizing primary actions generously, keeping related controls near each other, and pushing frequent actions toward edges and corners rather than the open middle of a screen.

Attention: the Von Restorff effect

The Von Restorff effect, also called the isolation effect, is the finding that when several items look alike, the one that differs from the rest is the one people notice and remember.¹ It comes from memory research by psychiatrist Hedwig von Restorff in the 1930s: a list of similar items with one outlier produces sharply better recall for that outlier than for any item in an otherwise uniform list. The consequence for interfaces is counterintuitive: if everything is bold, colored, or urgent, nothing reads as bold, colored, or urgent, because there is no longer an outlier to notice. Apply it by giving one element per screen, usually the single most important action, a visual treatment nothing else on that screen shares, and resisting the urge to add a second one.

Grouping: the Gestalt principles

The Gestalt principles describe how the eye assembles separate marks into groups before any conscious reasoning happens. Three of them do most of the work in interface design.¹

• Proximity: items placed close together read as one group, and items placed farther apart read as unrelated, regardless of any other styling applied to them. Two form fields six pixels apart read as a pair; the same two fields with sixty pixels between them read as separate questions.
• Common region: items that share an enclosed boundary, a card, a shaded panel, a border, read as one group even when they are not otherwise close together. A pricing card with a visible border reads as one unit; the same numbers as a bare row do not.
• Similarity: items that share a visual property, color, shape, size, read as related, even when scattered across a layout. A row of icons in one accent color reads as one navigation system; a stray icon in a different color reads as something else entirely.

The practical rule is easy to state and easy to violate: group related items with space and a shared container, not with a dividing line. A thin rule cutting through a rounded card fights the grouping the card's own border already implies, which is a common reason a layout that looks fine in isolation reads as subtly broken once it is full of real content.

Scanning: the F-pattern

Eye-tracking research from the Nielsen Norman Group found that people do not read web pages top to bottom, they scan them, frequently in a pattern shaped roughly like the letter F: a horizontal sweep near the top, a shorter horizontal sweep further down, then a vertical scan down the left edge.⁶ Across the underlying studies, participants read on average only a modest fraction of the words on a page during a typical visit. The pattern is a symptom, not a preference: people scan because reading everything costs more time than most pages are worth. Apply it by front-loading headings, the first few words of each sentence, and the left edge of any list with the words that carry the meaning, since that is disproportionately what gets seen.

Revealing: progressive disclosure

Progressive disclosure is the design response to Hick's and Miller's laws: instead of showing every option and every field at once, show what the current task needs and keep the rest one clear, discoverable step away.⁷ It lowers what a screen demands of working memory without removing any capability from the product; the option is still there, just not competing for attention until it is relevant. An "Advanced" section tucked behind a single toggle, a form that only asks for a shipping address after billing is confirmed, and a short tooltip that expands into a longer explanation on request are the same mechanism at three different scales. Apply it by asking, for every field or control on a screen, whether most users need it right now; if not, it belongs one click away, not on the surface.

Perception: the aesthetic-usability effect

The aesthetic-usability effect is the finding that people perceive attractive designs as easier to use than plain ones, and forgive small problems in a polished interface that they would flag immediately in a rough one. Masaaki Kurosu and Kaori Kashimura first measured it in a 1995 study of ATM interface layouts: ratings of a design's apparent ease of use correlated more strongly with how attractive people found it than with how usable it actually was.⁸ The Nielsen Norman Group's review of the research adds an important limit: the effect has a floor, and it masks minor friction, not serious, task-blocking usability problems.⁹ A well-designed form with one slightly unclear label will usually get a pass; a well-designed form that cannot be submitted will not. Apply it by treating visual craft, spacing, alignment, consistent type, restrained color, as functional work that buys real patience and trust, not decoration added once the "real" work is done, and never as cover for an interface that is genuinely broken underneath.

Using them together

These findings are not independent switches; they pull on the same few resources. Hick's Law and progressive disclosure both fight against showing too much at once. Miller's Law and the Gestalt principles both fight against forcing people to hold loose, ungrouped information in their head. Fitts's Law and the F-pattern both describe how the eye and the hand move faster along paths they already expect, which is also the whole argument behind Jakob's Law. The Von Restorff and aesthetic-usability effects both work at the level of perception rather than logic: what looks considered earns more trust and more attention than what is merely correct underneath. An interface that feels obvious is usually one where several of these are quietly satisfied at once, not one that leans hard on any single trick.