IB DP German B Tutor in Dubai

Progress in a language is invisible week to week and obvious across a term, which is why it has to be recorded rather than remembered. The table lists the artefacts a German B engagement should generate, who makes each one, what it actually proves and how often it should appear. A family holding these can answer the question honestly without relying on anyone's impression, including the tutor's.

Two of them do most of the work. The first is a criterion-scored written task, weekly, annotated rather than merely marked. Scores alone are weak evidence because task difficulty varies; scores plus annotation plus a named target are strong, because the next task tests the target. The second is the recorded oral taken on an unseen stimulus at a fixed interval, which is the only way to compare speaking performance against itself rather than against a memory of last term.

The error log is the least glamorous artefact and the most diagnostic. It is a running list of every structural error with its family tagged: case after a two-way preposition, verb position in a subordinate clause, past-tense register, adjective ending. After four weeks the distribution is unmistakable, and it is frequently not what either the student or the school believed. A student convinced they are bad at cases quite often turns out to be making almost all their case errors after one specific set of prepositions, which is a fifty-minute fix rather than a term's work.

What is not evidence: a mark out of ten with no criteria behind it, a comment that the student is engaging well, a percentage improvement with no sample size or method, or a predicted grade from anyone other than the school. We do not issue predicted grades, and any provider quoting a success statistic should be asked for the number of students, the dates and the method on the same page as the claim.

A first German B session spent teaching is a wasted session, because nobody yet knows what to teach. The useful first hour is a set of probes designed to isolate error families, and it should produce a written finding rather than a warm feeling. Each probe is short, and together they take under an hour.

The probes are specific. Translate four sentences that each force a different case after a two-way preposition, and the case problem either appears or it does not. Write two sentences joined by a subordinating conjunction, and verb position shows itself immediately. Narrate the same short event twice, once as speech and once as a written report, and the Perfekt and Präteritum choice becomes visible. Describe a person with three adjectives before the noun, and the adjective ending column either is known or is being guessed. Read one short authentic-style text aloud and answer two questions pointing to the supporting words.

The table below sets out what each probe reveals and what we most often find in a Dubai cohort, where students arrive from very different German backgrounds. The fluent heritage speaker and the Year 10 late starter sitting in the same combined class produce almost opposite findings, and a tutor who begins teaching before establishing which one they have will spend six weeks on the wrong thing.

The output of the diagnostic should be written down and given to the family: the error families ranked, an estimated time to change each, and which one will be targeted first. That document is also the baseline against which the six-weekly review compares. Without a baseline, every later claim about progress is unfalsifiable, which is precisely how tutoring in this market usually gets reported.

Paper 1 is a single written task chosen from three options, each naming a text type, an audience and a purpose. In the current guide it is marked on three criteria: language, message and conceptual understanding, carrying twelve, twelve and six marks. Check the exact allocations against your child's subject guide, because they belong to the IB, but the shape is what matters here. Grammar and vocabulary account for less than half the total.

This is why criterion-level marking changes a family's understanding of their own child. A student who scores well on language and badly on message is not weak at German; they are writing the wrong text. A student who scores respectably on both and loses the conceptual marks has not registered the audience and purpose they were given. Two students with the same overall mark can need entirely different teaching, and a single number at the top of a script hides that completely.

The practical cycle is three tasks long. Task one is marked across all three criteria with no target set, to establish the baseline. The lowest criterion becomes the target, and task two is written to a brief that tests it specifically. Task three checks whether the gain held when attention moved on, because a change that survives inattention is a change and anything else is compliance. Families should expect to see those three scripts together, not one at a time.

For German specifically, the language criterion rewards accurate structure more than ambitious vocabulary. A task written in controlled German with correct cases and verb positions scores better than one reaching for impressive nouns and losing control of the clause. Telling a student to write less ambitiously is counterintuitive advice that frequently gains three or four marks immediately, and it is the clearest example of why criterion-level evidence beats general encouragement.

Almost every structural error in a German B script belongs to one of four families, and each has a diagnostic question that identifies it in under a minute. This is the difference between German and most other B subjects: the errors are systematic, so they are fixable systematically rather than by general exposure.

Case selection is the first family, and it is usually narrower than students fear. Most case errors in Dubai scripts cluster after the prepositions that take either accusative or dative depending on whether movement is involved, rather than being spread evenly across all four cases. The diagnostic question is simply whether the error rate changes when those prepositions are removed from the task. If it does, the fix is a focused two to four weeks on motion versus position, not a term of grammar.

Verb position is the second. A student who knows perfectly well that the verb goes to the end of a subordinate clause will still leave it in second position when writing at speed, because the rule has not yet become automatic. The diagnostic is accuracy under timing against accuracy without it; a large gap means automation work, which is short sentences at volume rather than explanation. Explanation has already happened, often for years.

Past-tense register is the third and the one schools flag least. German draws a distinction between the spoken and written past that English does not, and a student who writes a formal report in the conversational past loses language and conceptual marks simultaneously. The fourth family is adjective endings, where students pick from the wrong column because they are reading the article rather than the case and gender. The table pairs each family with its symptom, its diagnostic question and the work that changes it.

Evidence only means something on a schedule, so here is one term laid out. The cycle assumes one weekly session with marking between, a recorded oral every third week, and a written review at week six and week twelve. It is deliberately unremarkable: the value is that each week produces something a parent can look at.

Weeks one to three establish the baseline and target the first error family. The diagnostic in week one, the first marked task in week two, the first recorded oral in week three. By the end of week three a family should have a written statement of the error families in priority order, two criterion-scored scripts and one recording. If any of those three is missing, the engagement is already not producing evidence.

Weeks four to eight are the working middle, and they are where most tutoring quietly loses its thread. The protection is the retention check: the error family targeted in week four is retested in week seven, on a task that is not about it. If the gain held, the next family begins. If it did not, the same family continues, and the review says so rather than moving on to keep the sessions feeling fresh.

Weeks nine to twelve add exam conditions, because accuracy that survives timing is the only accuracy that counts in May. The second recorded oral is compared directly with the first from week three, which is the single most convincing artefact in the whole set. The week twelve review compares like with like: same criteria, same stimulus type, different task. Nothing in it predicts a grade, because predicting is the school's job and a prediction would be the least useful line on the page.

The individual oral is internally conducted and recorded by the school, and it carries a quarter of the grade. At Standard Level it starts from a visual stimulus linked to one of the prescribed themes; at Higher Level it starts from an extract from one of the two literary works the school chose. Because the real thing is a recording, a rehearsal recording is an unusually faithful simulation, which makes it the best evidence in the set.

The mechanics matter. Rehearsals use a stimulus the student has genuinely not seen, under the real preparation conditions, and are marked against the same criteria every time. Changing the stimulus type between recordings destroys comparability, which is how families end up with four recordings that cannot be compared with each other. Keeping the format fixed and the content unseen is what makes the April recording meaningful next to the November one.

The second reason to record is that students do not believe descriptions of their own speech. Tell a student their word order collapses when they are thinking, and they will disagree. Show them ninety seconds of themselves doing it, and the next rehearsal is different. For the German error families in particular, which are structural and therefore audible, watching is worth more than being told, and it shortens the fix by weeks.

At Higher Level the recordings carry an extra job. They show whether the student can discuss a literary extract rather than retell it, which is the commonest Higher Level failure and the one a combined-level Dubai class has least time to address. A monthly extract response, recorded, is the only way to see that range growing. Ask any prospective Higher Level tutor which two works your child's school chose before agreeing anything, because a tutor working from different texts is rehearsing the wrong thing.

Evidence tells you what has happened. It does not tell you what will happen, and a tutor who converts a marked-task trend into a promised grade has stopped doing the thing that made the evidence valuable. We do not issue predicted grades for German B. The school does that, with access to a full cohort and the school's own assessment history, and our review notes are written to be useful alongside that rather than in competition with it.

Nor does evidence justify statistics. A provider in this market claiming that a percentage of their students improved should be asked, on the same page, for the number of students, the dates covered and the method of measurement. Without those three, the figure is decoration. We publish none, which is a deliberate choice rather than an oversight: at the scale these pages operate, an unsourced number repeated across hundreds of URLs is hundreds of liabilities.

Two more limits worth stating plainly. A tutor can mark, annotate and teach; a tutor cannot write or rewrite work your child submits as their own, cannot draft an Extended Essay or a TOK essay, and has no part in a recorded individual oral beyond rehearsal beforehand. IB academic honesty rules place the consequences on the student. And on tutor status: we ask tutors to hold the MoHRE private tuition work permit, introduced by Ministerial Resolution No. 710 of 2023 and granted free of charge for two years, applied for through MoHRE's own channels. Holding it is a legal requirement for the tutor, not an endorsement of IB Gram by anyone.

For completeness on the things families often ask in the same breath: IB Gram operates no premises in Dubai or anywhere in the UAE, holds no KHDA, ADEK, SPEA, Ministry of Education, IB or Cambridge endorsement, and does not offer home visits in the Emirates. Market rates in this city, from published provider rate cards and Bestimate's Tutoring Cost Dubai 2026 table read in October 2026, run roughly AED 150 to AED 300 an hour for a qualified language teacher online and AED 200 to AED 450 and above for an IB-experienced specialist. Our rate is agreed in writing after the free trial lesson.