Length, weight, and temperature use different measurement conventions

What it is

A locale-sensitive unit preference means that one physical quantity may be expressed in different units or mixed units depending on region, purpose, and magnitude. This is more specific than “metric versus imperial.” Person height, road distance, rainfall, cooking mass, and weather temperature may follow different conventions in the same region. A single use case may also switch from miles to feet, or metres to centimetres, as the magnitude changes. Unicode LDML therefore models preference as a joint choice over quantity category, usage, region, and magnitude-threshold range.

Why it happens

A measurement combines a number, a unit, and a physical dimension; the displayed unit is only one representation of that semantic quantity. CLDR unitPreferenceData resolves rules by quantity category and usage, applies regional fallback, and uses geq thresholds to choose a unit appropriate to the magnitude. The result can be one unit or a mixed unit such as feet and inches. Locale units data then supplies long, short, and narrow names and their grammatical forms. Temperature exposes another complication: Celsius–Fahrenheit conversion is affine, while a temperature interval does not use the same offset as an absolute temperature. A single “metric/imperial” flag inferred from interface language discards the variables that actually determine a useful expression.

Studying it

Inventory measurement fields by physical dimension and usage, then resolve the target regions' unit sequences, thresholds, and mixed units from the current CLDR data as a baseline. In user studies, ask participants from the target context to estimate, compare, enter, and act on measurements rather than merely state a preference. Useful independent variables include usage, magnitude, region, display width, and unit width; outcomes include comprehension accuracy, spontaneous conversion, completion time, and confidence. Record participants' everyday and professional unit experience, since nationality and interface language are weak proxies for familiarity.

Where it stops holding

CLDR provides maintainable regional defaults, not the final word on law, domain practice, or individual preference. Aviation, medicine, science, engineering, and navigation can use professional units across regional boundaries. Markets such as the UK mix conventions by use case. Everyday “weight” often refers to mass, but force and mass are not interchangeable in calculation. Defaults also evolve between data releases, so a country must not be permanently hard-coded to one system, and findings from one participant sample do not describe an entire region.

Applying it

• Store canonical quantities with a physical dimension, and derive display units from them; do not treat a preformatted number-plus-suffix string as source data.
• Declare a category and a specific usage for each field, resolve the preferred unit from region and magnitude, then use available width only to select its long, short, or narrow representation.
• Use CLDR unit identifiers, conversion relations, preference thresholds, and localized units patterns. Support both single and mixed units, and distinguish absolute temperature from temperature difference.
• Build golden cases for road distance, person height, weather, cooking, and small magnitudes in each target region; verify selected units, threshold transitions, plurals, spacing, and narrow-screen output.

Related

• Same group: S2.04.2 Conversion rounding can create inconsistencies · S2.04.3 Users need to be able to switch units · S2.04.4 A missing unit is a safety hazard, not a typography issue
• Adjacent: S2.02 Numbers, decimals, and digit grouping · S4.02 Separating region from language settings
• Search terms: CLDR unitPreferenceData · locale-sensitive units · mixed units