IB IGCSE Tutor for Repton School Abu Dhabi Students in Abu Dhabi

Evidence has to be decided in advance or it will not exist. The useful move at the start of an arrangement is to agree three things: what artefact each week will produce, what will be counted on it, and when the review happens. Without that, week six arrives with nothing to look at except a child's opinion and a tutor's reassurance, and neither is a basis for spending another three thousand dirhams.

The artefacts are ordinary and already exist in the school's own cycle. A marked class assessment, a homework task returned with comments, a timed question done in the session, a draft of an internal assessment, a set of past-paper answers with the losses circled. What turns any of these into evidence is counting something consistently: the number of explain questions answered as descriptions, the number of calculations missing units, the number of method lines absent, the criterion band awarded on a draft.

The table sets out a six-week shape for a typical arrangement. It is a format rather than a case study, and no claim is made about what any particular student achieves, because an outcome claim without a sample size and a method is not information. What the table does show is what a family should expect to be able to read at the end of six weeks, and what to ask for if it is not arriving. If week four produces no comparable paper and week six produces no written summary, the arrangement is not evidence-led regardless of how the sessions felt.

Everything else in this market is contestable. A child says the sessions are useful. A tutor says the student is engaging well. A report says progress is satisfactory. None of those three statements can be checked against anything, and all three are routinely true while a grade stays exactly where it was. A marked paper is different: it has a number, the number came from a published scheme, and the losses are individually visible.

Using it properly means marking for categories rather than for a total. A total tells you where the student is; the categories tell you why. Go through the script and label every lost mark as one of four things: content the student did not know, a verb they answered wrongly, a convention they omitted, or time they ran out of. Add them up. Most families find the content column is the smallest of the four, which is both surprising and useful, because the other three are faster to fix.

The second discipline is comparability. Two papers from the same specification and tier, marked the same way, six weeks apart, give a readable comparison. A class test against a past paper does not, and nor does a paper the student had seen before. This is why an early mock matters so much: it is the first genuinely comparable baseline a family gets, and treating it as a rehearsal rather than a diagnostic wastes the most useful document of the year.

A mark that goes up is not automatically progress. Papers differ in difficulty, a student can have a good day, a familiar topic can inflate a result, and a single assessment carries real random variation. Families who treat every mark as a signal end up changing tutors, subjects and strategies on noise, which is both expensive and demoralising for the student.

The way through this is not statistical sophistication; it is counting the right thing. A loss category is far less noisy than a total mark. If a student omitted units in nine calculations in week one and in one calculation in week four, that is a real change regardless of what the overall marks did, because the behaviour is directly observable and not subject to paper difficulty. Loss counts move before totals do, which is also why they are the better early signal.

The table sets out what to treat as signal and what to discount. The practical rule underneath it is that two consistent movements in the same direction, on comparable material, mean something, and one does not. Six weeks is long enough to see the first and short enough to be worth reviewing at. A family unwilling to stop an arrangement at week six on the evidence is not really collecting evidence; they are collecting reassurance. The uncomfortable corollary is that a tutor should be willing to say when the evidence is flat, and the ones who do are the ones worth keeping.

An internal assessment is the richest evidence item in a Diploma or a coursework-bearing A Level, because it is marked against published criteria that both the student and the family can read. Scored at each draft, it produces a record of exactly which criterion moved and when, and that record is far more informative than any grade prediction.

The method is simple and it is the student's own work throughout. Draft one is scored honestly against the descriptors, usually badly, and the lowest criterion is identified. The student revises that criterion and only that criterion. Draft two is scored again. The movement, or the absence of it, is the evidence. Doing this across three drafts typically moves two criteria substantially and makes the remaining weakness obvious, which is a far better position than a single panicked rewrite in the final fortnight.

The trail matters for a second reason that is worth stating plainly. A documented sequence of the student's own drafts, each with the tutor's criterion-level feedback and the student's own revisions, is also the evidence that the work is theirs. That protects the student if a question is ever raised, and it is exactly what a purchased internal assessment cannot produce. A tutor who contributes text destroys the trail as well as breaking the rules, which is why the honest method and the defensible method turn out to be the same method here rather than a compromise between two.

A written record after every session is not an administrative nicety; it is the mechanism by which a family holds an arrangement accountable. It needs four things and nothing more: what was covered, what loss was being targeted, what was set for the week, and what the tutor still considers weak. Four sentences is enough. A paragraph of encouragement is not a record and should be identified as what it is.

What to do if it does not arrive is the useful half of this. Ask for it once, in writing, specifying the four items. A tutor working from marked scripts will produce it without difficulty, because they already know all four answers. A tutor who cannot produce it is either not reading the student's work or is working without a plan, and in both cases the arrangement is less than a family is paying for. That is a decision available at week two rather than week twenty.

The same principle applies to the school side. A family is entitled to ask the subject teacher what specifically changed in the student's work, rather than whether progress is being made. The second question gets a courteous non-answer; the first gets a usable one. Between the school's answer and the tutor's weekly record, a parent in Abu Dhabi can assemble a genuinely informative picture without waiting for a report cycle. Neither request is unreasonable and neither takes anybody more than a few minutes to answer, which is why the refusal rather than the answer is the informative outcome.

An evidence-led arrangement depends on a slot that actually holds every week, which is where the logistics of this part of Abu Dhabi matter. Al Reem's exit onto the main crossing in the late afternoon is slow, and a tutor driving in to a five o'clock appointment is committing a large part of their evening to one hour of teaching. That is why the local in-home band is quoted at AED 180 to 500 an hour, against a market band of AED 150 to 450 for qualified to specialist work online.

The cost that matters more for evidence, though, is the lost session. Families with travelling tutors commonly lose two or three sessions in a twenty-eight-session term to traffic, weather or an overrun. Each loss breaks the chain of comparable weekly artefacts, which is exactly the thing the whole method depends on. A video session fails only when the student or the tutor is genuinely unavailable, and it can move by an hour rather than being cancelled.

The third cost is the tutor pool. Travel restricts the choice to whoever drives to this part of the island at that hour, and for a particular specification or component that radius may contain nobody with marking experience. IB Gram does not send tutors to homes anywhere in the UAE; home visits are a service offered only in India. What a family here buys instead is a specialist matched nationally, teaching on Gulf Standard Time, with the weekly record intact.

Exam calendars are usually presented as dates to fear. Read as evidence deadlines they are more useful, because each milestone is a point at which a family either has a readable document or does not. The entry deadline is the last moment at which the evidence can change a decision about tier or subject. The mock is the first comparable baseline. The coursework submission is the end of the draft trail.

The calendar facts for a UAE candidate are specific. Cambridge International runs a June series and a November series; the February and March series is offered to India only, so it is not an option here, and guidance suggesting a spring sitting describes another country's timetable. Cambridge allocates every centre to one of six administrative zones and states that using the correct zone version of the timetable is essential, warning that it cannot accept responsibility for a candidate who misses an exam after using the wrong one. Pearson Edexcel International GCSE runs in May and June and again in November.

The dates worth citing precisely are historical, because those can be checked. Pearson's November 2026 final timetable ran from late October to 19 November 2026; June 2026 results reached centres on 19 August 2026 and students the day after; November 2025 results arrived with students on 22 January 2026. Future windows are named by month only here, deliberately, since provisional timetables are revised and a pinned future date makes a page wrong without anyone editing it.

Assessed work has to be the student's own, and the consequences of a breach fall on the student rather than on anyone selling help. That is the principle. The practical version, on an evidence-led page, is that the draft trail which proves progress is the same document that proves authorship, so the honest method and the defensible method are identical.

What a tutor can do is the whole of the service. Read the published criteria with the student and explain what each descriptor awards. Press on a research question until it is answerable within the limit. Teach the statistical test, the laboratory technique or the referencing convention the piece requires. Score a draft the student wrote and name the criterion costing marks. Rehearse the supervisor conversation. Every one of those leaves the student's work untouched and the trail intact.

What a tutor cannot do is contribute to the submission: no text, no redraft into final form, no data presented as collected, no edit that leaves the work substantially someone else's. One operation ranking in this market sells completed internal assessments at a published price. Beyond the rule, that purchase destroys the only evidence the student has that the work is theirs, which is a bad trade at any price. The self-test is whether the student could explain every sentence of their own submission, unprompted, to the person supervising it. If they could not, the line has been crossed whoever crossed it, and the repair is to return to the last draft that was genuinely theirs and continue from there.