Moderating a Disputed Oral Score
At a Maltese secondary school, two teachers co-rate a Year 11 mock SEC Arabic oral. The candidate, a 16-year-old fluent speaker of Levantine Arabic, talks for 90 seconds about her summer plans. She uses بدي instead of أريد, drops case endings, code-switches twice into Maltese (imma, grazzi), and answers only half of the follow-up question about travel costs. Rater A awards 78/100, citing natural fluency. Rater B awards 55/100, arguing that non-standard forms must be penalised. The speaking rubric weights communicative effectiveness 40%, range 20%, accuracy 20%, pronunciation 20%. The syllabus asks for “simple standard Arabic” but nowhere lists dialect as an automatic error. A 10-minute moderation meeting is scheduled; the recording and rubric are available. The coordinator must resolve the discrepancy, document a defensible score, and give the candidate feedback that improves performance without shaming her dialect.
1. Which rater’s interpretation is better supported by the rubric? Justify your decision criterion by criterion, and state the score band you would record.
Read the worked answer
Neither rater is simply right; each overweights one criterion. Work through the four rubric weights.
Communicative effectiveness (40%) is the largest component and asks whether the candidate fulfils the task. She describes her plans for most of the 90 seconds but answers only half of the follow-up about costs. That is partial fulfilment, not failure: perhaps 60–65% of this criterion. Rater A’s 78 assumes full task completion; Rater B’s 55 ignores that she communicated a coherent message.
Range (20%) is moderate: she uses present and future reference and some connectors, but relies on familiar phrases. About 55–60%.
Accuracy (20%) is where بدي, dropped case endings and code-switching matter. The syllabus requests simple standard Arabic, so these reduce accuracy, but the rubric does not say “dialect = zero.” About 45–50%.
Pronunciation (20%) is intelligible and fluent; occasional Maltese-influenced vowels do not impede understanding. About 65–70%.
Weighted: 0.40×62 + 0.20×57 + 0.20×47 + 0.20×67 ≈ 59. That places her in a mid band, not 78 and not 55. I would record around 59–62 and explain the weighting in the moderation note, so the score is defensible on the rubric rather than on either teacher’s instinct.
2. Draft the written feedback you would give the candidate, including at least two concrete language targets and one way to keep her fluency strength.
Read the worked answer
Feedback should separate what worked from what the rubric rewards, and should treat dialect as a register choice rather than a moral failing.
Strengths to name: she sustained 90 seconds without long pauses and made her plans clear; her pronunciation was easy to follow. Keep this by continuing to record herself speaking for 90 seconds, but with a checklist.
Target 1 - answer the whole follow-up. The follow-up asked about costs. Practise a three-part answer: direct answer, one reason, one detail (e.g. «تكلفة التذكرة مرتفعة لأنني أسافر في الصيف؛ سأبحث عن عرض»). This raises communicative effectiveness, the 40% criterion.
Target 2 - convert high-frequency dialect forms into simple standard alternatives. Build a personal substitution list: بدي → أريد, بدّي روح → أريد أن أذهب, منيح → جيد. Do not ban her dialect in conversation; ask her to deliver the same content twice, once freely and once in standard Arabic, so she hears the difference.
Target 3 - reduce code-switching under exam conditions. Replace imma with لكن and grazzi with شكراً. Two switches in 90 seconds is manageable with a rehearsed opening and closing.
Keep the strength: her fluency is an asset; the goal is to add a standard register, not to slow her down.