{ "id": "english", "version": 1, "url": "https://certo.bio/rubrics/english/1", "json_url": "https://certo.bio/rubrics/english/1.json", "title": "English", "summary": "What the AI interview measures of a candidate's spoken English: seven aspects of speaking, each named after the CEFR scale it rests on (Common European Framework of Reference for Languages, Companion Volume, Council of Europe 2020), judged on five levels from how the candidate speaks in the interview. The result is a CEFR level from A1 to C2. It measures speaking and conversation only: not reading, writing or listening tests, and not knowledge of any subject. An accent is never marked down; intelligibility is what counts. The descriptors and indicators are certo's own words, not the Council of Europe's. Levels 2, 3 and 4 each list concrete indicators the interview marks met, missing or not discussed, with the candidate's own words as proof; a skill the interview could not assess is left unassessed rather than scored low, and a result needs at least 70 percent of the rubric's weight assessed.", "status": "published", "published_at": "2026-10-13T07:59:26.408153+00:00", "retired_at": null, "superseded_by": null, "subject": "language", "occupation": null, "language": "en", "criteria": [ { "key": "spoken_production", "title": "Speaking at length", "cefr_scale": "Overall oral production", "weight": 15, "listens_for": "How the candidate describes, explains and argues on their own: a sustained answer about their work, a process or a view.", "levels": [ { "level": 1, "descriptor": "Answers with short, simple phrases about familiar things; longer answers break down." }, { "level": 2, "descriptor": "Describes their work and experiences in a simple, connected way, as a linear sequence of points.", "indicators": [ { "key": "describes_own_work", "text": "Ability to describe their own work in connected sentences", "must_have": true }, { "key": "narrates_an_experience", "text": "Ability to recount an experience as a sequence of events", "must_have": false } ] }, { "level": 3, "descriptor": "Gives clear, detailed descriptions on familiar and job-related topics, supporting points with reasons and examples.", "indicators": [ { "key": "supports_with_examples", "text": "Ability to support a point with reasons and examples", "must_have": false }, { "key": "explains_a_viewpoint", "text": "Ability to explain a viewpoint with advantages and disadvantages", "must_have": false } ] }, { "level": 4, "descriptor": "Gives clear, well-structured explanations of complex subjects, developing sub-themes and rounding off with a conclusion.", "indicators": [ { "key": "structures_complex_answer", "text": "Ability to structure a long answer on a complex subject", "must_have": false }, { "key": "develops_and_concludes", "text": "Ability to develop sub-points and round off with a conclusion", "must_have": false } ] }, { "level": 5, "descriptor": "Every strong indicator, plus speech that is effortless, precise and adapted to the listener throughout." } ] }, { "key": "spoken_interaction", "title": "Conversation and interaction", "cefr_scale": "Overall spoken interaction", "weight": 15, "listens_for": "How the candidate takes part in the exchange: answering what was asked, asking for clarification, following up and keeping the conversation going.", "levels": [ { "level": 1, "descriptor": "Handles very short exchanges; depends on the interviewer repeating or rephrasing." }, { "level": 2, "descriptor": "Takes part in a conversation on familiar topics and answers the question that was asked.", "indicators": [ { "key": "answers_the_question", "text": "Ability to answer the question that was actually asked", "must_have": true }, { "key": "asks_for_clarification", "text": "Ability to ask for clarification or repetition when needed", "must_have": false } ] }, { "level": 3, "descriptor": "Interacts with fluency and spontaneity, responds to follow-ups and keeps the exchange going without strain.", "indicators": [ { "key": "responds_to_follow_ups", "text": "Ability to respond to follow-up questions without losing the thread", "must_have": false }, { "key": "keeps_exchange_going", "text": "Ability to keep the conversation going without strain", "must_have": false } ] }, { "level": 4, "descriptor": "Uses language flexibly in the exchange, relates their answers skilfully to the interviewer's points and handles a challenge.", "indicators": [ { "key": "relates_to_interviewer", "text": "Ability to relate their answer skilfully to the interviewer's point", "must_have": false }, { "key": "handles_challenge", "text": "Ability to handle a challenge or a counter-argument smoothly", "must_have": false } ] }, { "level": 5, "descriptor": "Every strong indicator, plus ease with nuance, idiom and tone, as in their first language." } ] }, { "key": "fluency", "title": "Fluency", "cefr_scale": "Spoken fluency", "weight": 15, "listens_for": "How smoothly the candidate speaks: pausing, false starts and how often they have to search for words.", "levels": [ { "level": 1, "descriptor": "Speech is slow and fragmented, with frequent pauses to search for words." }, { "level": 2, "descriptor": "Keeps going comprehensibly, with noticeable pauses for planning and repair in longer stretches.", "indicators": [ { "key": "keeps_going", "text": "Ability to keep going comprehensibly through a longer answer", "must_have": true } ] }, { "level": 3, "descriptor": "Produces stretches of speech at a fairly even tempo; hesitates when searching for expressions, with few long pauses.", "indicators": [ { "key": "even_tempo", "text": "Ability to speak at a fairly even tempo", "must_have": false }, { "key": "few_long_pauses", "text": "Ability to avoid long pauses when searching for words", "must_have": false } ] }, { "level": 4, "descriptor": "Expresses themselves fluently and spontaneously, almost effortlessly; only a conceptually difficult subject slows them.", "indicators": [ { "key": "almost_effortless", "text": "Ability to speak fluently and spontaneously, almost effortlessly", "must_have": false }, { "key": "recovers_smoothly", "text": "Ability to recover from a slip without the listener noticing", "must_have": false } ] }, { "level": 5, "descriptor": "Every strong indicator, plus a natural, effortless flow with no hesitation on any subject." } ] }, { "key": "range", "title": "Range of vocabulary and expression", "cefr_scale": "General linguistic range", "weight": 15, "listens_for": "Whether the candidate has the words and structures to say what they mean, including about their own field, without circumlocution.", "levels": [ { "level": 1, "descriptor": "Has a basic repertoire for simple, everyday needs." }, { "level": 2, "descriptor": "Has enough language to get by on familiar and work topics, with some circumlocution.", "indicators": [ { "key": "enough_for_work_topics", "text": "Knowledge of enough vocabulary to talk about their work", "must_have": true } ] }, { "level": 3, "descriptor": "Has a sufficient range to give clear descriptions and express viewpoints, using some complex sentence forms.", "indicators": [ { "key": "field_vocabulary", "text": "Knowledge of the vocabulary of their own field", "must_have": false }, { "key": "varies_expression", "text": "Ability to vary how they formulate a point", "must_have": false } ] }, { "level": 4, "descriptor": "Has a broad range that lets them choose a formulation to express themselves precisely without restricting what they say.", "indicators": [ { "key": "precise_formulation", "text": "Ability to choose a formulation that says exactly what they mean", "must_have": false }, { "key": "abstract_topics", "text": "Ability to discuss abstract topics without restricting what they say", "must_have": false } ] }, { "level": 5, "descriptor": "Every strong indicator, plus a command of idiom and nuance that conveys finer shades of meaning." } ] }, { "key": "accuracy", "title": "Grammatical accuracy", "cefr_scale": "Grammatical accuracy", "weight": 15, "listens_for": "How accurately the candidate uses grammar, and whether errors ever get in the way of what they mean.", "levels": [ { "level": 1, "descriptor": "Uses simple structures with basic mistakes that sometimes obscure the meaning." }, { "level": 2, "descriptor": "Uses frequent routines and patterns reasonably accurately; errors occur but the meaning is clear.", "indicators": [ { "key": "meaning_stays_clear", "text": "Ability to keep the meaning clear despite errors", "must_have": true } ] }, { "level": 3, "descriptor": "Shows a relatively high degree of grammatical control; errors do not cause misunderstanding and are often self-corrected.", "indicators": [ { "key": "good_control", "text": "Ability to use grammar without errors that cause misunderstanding", "must_have": false }, { "key": "self_corrects", "text": "Ability to notice and correct their own mistakes", "must_have": false } ] }, { "level": 4, "descriptor": "Consistently maintains a high degree of accuracy; errors are rare and difficult to spot.", "indicators": [ { "key": "consistent_accuracy", "text": "Ability to maintain a high degree of grammatical accuracy throughout", "must_have": false } ] }, { "level": 5, "descriptor": "Every strong indicator, plus consistent control of complex language even when attention is elsewhere." } ] }, { "key": "coherence", "title": "Coherence and cohesion", "cefr_scale": "Coherence and cohesion", "weight": 10, "listens_for": "How the candidate links ideas: whether a longer answer hangs together, and how they signal the connections between points.", "levels": [ { "level": 1, "descriptor": "Links words and groups of words with very basic connectors such as and or then." }, { "level": 2, "descriptor": "Links a series of shorter elements into a connected sequence of points.", "indicators": [ { "key": "links_points", "text": "Ability to link points into a connected sequence", "must_have": true } ] }, { "level": 3, "descriptor": "Uses a range of linking words to mark clearly the relationships between ideas.", "indicators": [ { "key": "marks_relationships", "text": "Ability to mark the relationship between ideas clearly", "must_have": false }, { "key": "uses_varied_connectors", "text": "Ability to use varied linking words beyond and, but and so", "must_have": false } ] }, { "level": 4, "descriptor": "Produces clear, smoothly flowing, well-structured speech, controlling organisational patterns and connectors.", "indicators": [ { "key": "smooth_structure", "text": "Ability to produce well-structured speech that flows smoothly", "must_have": false }, { "key": "signposts_structure", "text": "Ability to signpost the structure of a long answer", "must_have": false } ] }, { "level": 5, "descriptor": "Every strong indicator, plus a coherent, cohesive discourse that uses a full range of organisational patterns." } ] }, { "key": "pronunciation", "title": "Pronunciation and intelligibility", "cefr_scale": "Overall phonological control", "weight": 15, "listens_for": "Whether the candidate is easy to understand. An accent is not marked down: what counts is intelligibility, stress and intonation.", "levels": [ { "level": 1, "descriptor": "Pronunciation can be understood with some effort by a listener used to speakers of their language group." }, { "level": 2, "descriptor": "Pronunciation is generally intelligible, though mispronunciations occur and the listener sometimes asks for repetition.", "indicators": [ { "key": "generally_intelligible", "text": "Ability to be understood without the listener asking for repetition", "must_have": true } ] }, { "level": 3, "descriptor": "Pronunciation is clearly intelligible; stress and intonation support the meaning, even with a noticeable accent.", "indicators": [ { "key": "clearly_intelligible", "text": "Ability to be clearly intelligible throughout the interview", "must_have": false }, { "key": "stress_supports_meaning", "text": "Ability to use stress and intonation to support the meaning", "must_have": false } ] }, { "level": 4, "descriptor": "Varies intonation and places sentence stress correctly to express finer shades of meaning.", "indicators": [ { "key": "intonation_for_nuance", "text": "Ability to vary intonation to express finer shades of meaning", "must_have": false } ] }, { "level": 5, "descriptor": "Every strong indicator, plus articulation that is clear and natural throughout, with full control of prosody." } ] } ], "min_assessed_weight": 70, "level_bands": [ { "band": "A1", "min": 1 }, { "band": "A2", "min": 1.5 }, { "band": "B1", "min": 2 }, { "band": "B2", "min": 3 }, { "band": "C1", "min": 4 }, { "band": "C2", "min": 4.8 } ], "scoring": { "scale": { "levels": [ 1, 2, 3, 4, 5 ], "bars": { "2": "weak", "3": "moderate", "4": "strong" }, "meaning": "Levels 2, 3 and 4 of each criterion are bars -- weak, moderate and strong -- each with written indicators; levels 1 and 5 carry a descriptor only. The interview marks every indicator met, missing or not_discussed, with proof (the candidate's words, or the AI's reading of them)." }, "criterion": { "method": "highest_bar_cleared", "bar_cleared": "no indicator on the bar is missing, every must-have indicator is met, and at least one indicator is met", "rule": "The level is 1 + the bars cleared in a row from 2: the bars stack, so a missing must-have caps the criterion below its bar.", "top": "Level 5 is bar 4 with every indicator met, plus exceptional evidence.", "unassessed": "When the weak bar failed only for want of evidence (nothing on it missing), the criterion is unassessed: null, never level 1. A missing weak indicator is level 1." }, "overall": { "method": "weighted_mean", "formula": "sum(weight × level) / sum(weight), over the assessed criteria only", "weights": "whole percents that add up to 100", "precision": 1, "rounding": "half_up", "range": [ 1, 5 ], "minimum_assessed_weight": "A result needs at least min_assessed_weight percent of the weight assessed; with less there is no result.", "note": "The weights are whole percents, so both sums are integers and the division is rounded half up without floating point: 265 / 80 is 3.3." }, "band": { "method": "last_band_reached", "rule": "The bands are in ascending order of min, the first at 1.0; the assessed band (a seniority, or a CEFR level for a language) is the last band whose min the overall result reaches (min ≤ overall).", "note": "A rubric need not reach every band of its scale; one it does not list cannot be assessed against it." } }, "attribution": { "text": "Criteria after the CEFR Companion Volume (Council of Europe, 2020); descriptors and indicators are certo's own words.", "source": "https://www.coe.int/en/web/common-european-framework-reference-languages" }, "publisher": "https://certo.bio" }