Reconciling two markers on an L3 writing script
In the S7 L3 German mock examination, the writing task asks candidates to write a 250-word argumentative text on whether the school day should start at 9:00. The agreed grid awards 10 points for content, 5 for organisation and 10 for language. The language descriptors are: 9–10 wide range, rare errors; 7–8 good range, errors do not impede communication; 5–6 adequate range, frequent errors occasionally impede; 3–4 limited range, errors often impede. Candidate M argues a clear thesis, uses two well-developed reasons and a counterargument, and organises the text with paragraphs and linking words. However, the text contains 14 language errors: four subject–verb agreement errors, three tense errors in subordinate clauses, two case errors after prepositions, and five spelling errors. Two complex sentences require re-reading. Marker A proposes 8/10 for language because the meaning is usually clear; Marker B proposes 5/10 because the errors are frequent and basic. The school requires markers to reconcile differences of more than one band before entering the mark.
1. Which language band best fits Candidate M, and which single mark within that band? Justify by matching the script evidence to the descriptors.
Read the worked answer
The best fit is Band 5–6, at the upper end (6/10). The script is not in Band 3–4 because the candidate communicates a clear argument and the errors do not 'often impede'; the reader can reconstruct the meaning throughout. It is not in Band 7–8 because that band requires errors that do not impede communication, whereas here two complex sentences require re-reading and the errors include basic agreement, tense and case items. The profile is 'adequate range, frequent errors occasionally impede', which is precisely the 5–6 descriptor. Within the band, 6 rather than 5 is warranted because the errors cluster in complex structures and do not destroy the argument; the content and organisation are strong. The number of errors (14) matters less than their effect, but the presence of basic errors in simple clauses prevents a 7. The marker should cite two examples, e.g. an agreement error and a subordinate-clause tense error, on the script.
2. The two markers are three points apart. What should they do under the school's reconciliation rule, and what total should be entered?
Read the worked answer
The difference (8 vs 5) crosses more than one band and triggers reconciliation. The markers should not simply average to 6.5, because the grid is banded and an average can fall between descriptors. They should re-read the script independently against the descriptors, then meet and identify the evidence for each band. If they still disagree, the school's procedure should apply: a third marker or a moderation meeting. The likely agreed language mark is 6/10 for the reasons above. The total is then content 10 + organisation 5 + language 6 = 21/25, i.e. 84%. If the school converts 25-point totals to a 10-point subject mark, the conversion table should be applied rather than a mental calculation. The key is that the final mark must be defensible from the descriptors, not a compromise between two numbers. The reconciliation record should note the band chosen, the evidence, and the final total.
3. What feedback would move this candidate from 6 to 8 in language without undermining the strong content?
Read the worked answer
The feedback should be diagnostic and limited to two or three priorities. First, the basic agreement and case errors in simple clauses are the main reason the script is not in Band 7–8; targeted drills and a proofreading checklist would help. Second, the tense errors in subordinate clauses need work on verb-final word order and tense sequencing. Third, the two sentences that require re-reading should be rewritten by the candidate to practise controlling complexity. The feedback should not tell the candidate to simplify content; the content is already at maximum. It should show that range and accuracy are both assessed: the candidate already has range, so the next step is accuracy in complex structures. A short annotated copy with two model corrections and a follow-up task (rewrite three sentences) is more useful than a generic 'revise grammar' comment. The candidate should also be told that 8 requires errors that do not impede, so the goal is not error-free writing but stable control.