The Research Task With a Suspected Translation
A Russian continuers student, Dmitri, submits a school-assessed written research task in Russian on a prescribed topic. The Russian is markedly more sophisticated than his classroom work: near-native idiom, rare vocabulary, and no errors across 600 words. The teacher is suspicious. School policy allows the teacher to question the student about the work but requires evidence before any penalty. Dmitri says he used online translation tools "for some sentences" and a dictionary for vocabulary, which the task sheet permits for individual words but not for whole sentences. The teacher must decide whether to investigate further, what evidence would be needed, and how to mark the task fairly to Dmitri and to classmates who wrote unaided.
1. What steps should the teacher take to investigate, and what counts as sufficient evidence?
Read the worked answer
Investigate proportionately, and define "sufficient" before acting.
Steps:
- Compare against baseline: Dmitri's classroom writing and previous assessments provide a reference. A sudden, unexplained jump in idiom and range is a signal, not proof.
- Oral verification: ask Dmitri, in Russian, to explain and expand on specific sentences and vocabulary from his own text. A student who wrote the work can discuss it; one who machine-translated often cannot explain idiom or paraphrase their own argument.
- Check the text for machine-translation fingerprints - unnatural register shifts, over-formal phrasing - but treat these as indicators, not verdicts, since fluent learners can produce them.
What counts as sufficient evidence: a documented pattern, ideally the oral verification failing on content the student supposedly authored, plus the baseline comparison. Suspicion alone, or "it sounds too good," is not sufficient. The teacher should write down the evidence and the student's explanation before deciding.
Guardrail: the investigation must be recorded, and Dmitri must have a chance to respond to the specific concerns - not a general accusation.
2. Suppose the evidence is inconclusive. How should the teacher mark the task, and how should the outcome be recorded?
Read the worked answer
If evidence is inconclusive, mark the work as submitted, and say so in the record.
- Marking: the task sheet permits dictionaries for individual words but not whole-sentence translation. If the teacher cannot establish that whole sentences were translated, the work stands as submitted and is marked on its merits. Penalising on suspicion would be unfair and would not survive moderation.
- Recording: the teacher should note the concern, the investigation steps taken, the student's explanation, and the conclusion that evidence was insufficient. This protects the teacher if the issue recurs and gives the moderator full context.
- Fairness to classmates: the concern that unaided students are disadvantaged is real, but it is addressed by evidence-based process, not by guessing. If a pattern emerges across a cohort, the response is a design change - e.g., in-class supervised drafting, or an oral defence component built into the task - not retroactive penalties.
- Feedback to Dmitri: if the work stands, feedback should still address the register gap honestly, noting that future tasks may include oral verification so that sophisticated language can be confidently attributed.
Principle: the burden of proof sits with the school. Inconclusive evidence means the mark stands, with the concern documented rather than acted on.