1.3 Scoring: The Seven Predikat Bands

The Seven Predikat Bands

UKBI does not report a percentage correct, a pass/fail decision, or a CEFR level. Instead, it reports an overall score on a 251–800 scale, which maps to one of seven predikat (descriptor) bands. The band is the result that institutions actually use.

Three structural features deserve attention:

1. The scale starts at 251, not 0. Every candidate who completes the test receives a predikat; there is no "fail" result, only a lower band. This is why UKBI is described as having no single pass/fail threshold.
2. The bands are not equally wide. Terbatas covers 75 points (251–325); Madya covers 96 points (482–577); Sangat Unggul covers 84. The asymmetry reflects the empirical distribution of test-taker ability — more granularity where most candidates fall.
3. Band 5 (Unggul) is the practical reference point. Many institutional requirements treat Unggul as a threshold of professional competence.

No Universal Pass/Fail

Badan Bahasa does not publish a single passing score. The band required for a given purpose is set by the institution that asks for the result. A university admission office might accept Madya; a teacher licensing pathway might ask for Unggul; a senior civil-service promotion might want Sangat Unggul. Always confirm the required band with the requesting institution before registering — aiming one band higher than the minimum is the safest strategy.

A Note on the "Unggul" Teacher Minimum

In practice, an Unggul (578–640) or higher result is commonly cited as the minimum predikat for Indonesian teachers. This reflects common institutional and local-government practice, not a Badan Bahasa mandate. Badan Bahasa itself publishes the bands and their descriptors but does not legislate which band a teacher must achieve. The binding requirement, if any, is set by the Dinas Pendidikan, the teacher's employer, or the relevant licensing authority. Treat "Unggul or higher" as a realistic target for teacher candidates, but verify the exact requirement with your own institution.

How the Adaptive Score Is Computed

UKBI's adaptive sections use Item Response Theory (IRT). The key thing to understand is that your score is not a count of correct answers. It is an estimate of your ability (θ, theta) computed from the difficulty of the items you answered correctly.

A simplified picture of how IRT works in UKBI:

1. The system starts with a mid-difficulty item unit (around the Madya level).
2. If you answer correctly, your ability estimate goes up, and the next unit is drawn from a harder pool.
3. If you answer incorrectly, your estimate goes down, and the next unit is easier.
4. After each unit, the system narrows the confidence interval around your ability estimate.
5. By the end of up to 8 units, your ability estimate is matched to a predikat band.

Practical consequences

• Early items carry more weight. A correct answer on the first unit moves you to a harder pool, where each correct answer is worth more ability-estimate gain. A wrong answer on the first unit pushes you toward an easier pool where even correct answers contribute less.
• You cannot game the system. Because the next unit depends on your current estimate, random guessing on early items tends to lock you into a lower-difficulty track for the whole section.
• The maximum is up to 8 units, not always 8. The system can stop early if your ability estimate is statistically stable — strong candidates may see fewer, harder units; weak candidates may see more, easier units.
• More correct is not always a higher band. Two candidates with the same raw count of correct answers can land in different bands if one answered harder items correctly.

How Production Sections Combine

Menulis (Seksi IV) and Berbicara (Seksi V) are scored against rubrics covering accuracy (PUEBI conformity), coherence, register, and (for speaking) pronunciation and fluency. These production scores are combined with the MCQ ability estimate to produce the overall predikat. A strong MCQ performance undermined by weak writing can pull the overall band down — and vice versa. Treat Menulis and Berbicara as part of the scored test, not afterthoughts.

Setting a Target Band

When you start preparing, decide your target band early. A reasonable framework:

• Madya (482–577) — sufficient for general professional use and many institutional sittings.
• Unggul (578–640) — the practical teacher target and a safe bet for regulated professions.
• Sangat Unggul (641–724) — appropriate for senior roles, language teachers, and journalism.
• Istimewa (725–800) — rarely required; an aspirational target for language specialists.

Use the official simulasi at simulasiukbi.kemendikdasmen.go.id to estimate where you currently fall, then add one band as your preparation target so that an off-day still leaves you above the requirement.

Common Scoring Misconceptions

• "I need 70% correct to pass." No — UKBI has no percentage-based pass mark. Your band depends on item difficulty, not raw count.
• "A low predikat means I failed." No — a low band is a proficiency descriptor, not a fail. Retaking is allowed and the higher result is typically the one used.
• "All sections weigh equally." No — the MCQ sections (I–III) dominate the score; Menulis and Berbicara are production modifiers that can move you within or across a band boundary.