The six-mark 'own words' question at district moderation
A district moderation meeting has 40 NSSCAS Oshindonga First Language scripts on the table. Question 2(b) is worth 6 marks and reads: 'What does the writer suggest about young people's attitude to cattle farming? Support your answer with evidence from the passage, in your own words.' The scheme allots 2 marks for a stated attitude, 2 for relevant evidence and 2 for explaining how that evidence supports the attitude. The passage is a newspaper opinion piece in which an elderly herder complains that his sons have left for Windhoek; the writer adds, ironically, that the family's cattle 'now have better manners than the boys'. Three scripts are in dispute. A states that young people dislike cattle and then copies three sentences from the passage. B claims the writer is angry but quotes nothing. C paraphrases the sons' departure, quotes the ironic sentence and explains the sarcasm. The two markers have so far agreed on only 24 of the 40 scripts.
1. Allocate a mark out of 6 to scripts A, B and C, justifying each decision against the wording of the scheme.
Read the worked answer
A - 4/6. Attitude stated independently ('young people dislike cattle'): 2. Evidence is relevant even though it is lifted: 2. Explanation: 0, because A never says how the copied sentences prove the attitude. The disputed phrase 'in your own words' must be ruled on first: as moderator I would rule that it governs the statement of attitude and the explanation, not the quoting of evidence - otherwise the question would forbid the very evidence it demands. But copying is not a route to full marks: it forfeits the explanation strand automatically. Do not award 6 and do not award 0.
B - 1/6. 'The writer is angry' answers a different question (the writer's feeling, not the young people's attitude to cattle farming). Because nothing is offered on the actual focus, the attitude mark cannot be earned; the single mark reflects a partially relevant response only if the scheme permits a 0–1 band for tangential answers. In a strict reading it is 0.
C - 6/6. Attitude paraphrased (2), accurate quotation (2), and the sarcasm is explained so the link between evidence and attitude is explicit (2). This is the script that shows what the top band looks like: selection, re-wording and commentary.
Consistency point: after this ruling, remark the 16 disagreed scripts against the same three-strand logic before returning any marks.
2. Two markers disagree over a script that infers the sons feel 'trapped by tradition', with no textual support. How do you resolve it, and what single line would you add to the scheme?
Read the worked answer
Resolution. The principle to apply is that a plausible inference must never outscore a supported one. Hold the script against the passage: the inference is only creditable if a specific word, phrase or incident can be pointed to by the candidate. Here nothing in the passage says or implies 'trapped by tradition' - the sons leave, which suggests the opposite. The third marker reads the relevant lines aloud; if two of the three can locate support the candidate has cited, the explanation marks stand. If not, the attitude mark may still be awarded for a defensible reading, but the evidence and explanation marks are withheld. Result: 2/6 at most.
Line to add: 'A claim that cannot be tied to a specific word, phrase or incident in the passage scores 0 in the evidence and explanation strands, however plausible it sounds; examiners must be able to point to the support the candidate has used.'
Add a second sentence to protect the other direction: 'Candidates who quote correctly but never state an attitude should not receive the evidence marks, because evidence must support a claim.'
3. Eleven scripts all scored exactly 2/6. What does this pattern tell you, and what do you do before the marks are released?
Read the worked answer
A cluster of eleven identical scores is more likely to be a marker effect than eleven identical scripts. Two/6 is the natural landing point for answers that identify a topic but give no evidence, so first check whether the marker is stopping at the attitude strand - a common drift when a scheme is written in strands. Read three of the eleven aloud and score them blind.
If they genuinely are unsupported assertions, the marks stand, and the pattern is diagnostic teaching information: the class has learned to state opinions without anchoring them. If several contain quotable evidence, the marker has been applying the scheme as a single 6-mark holistic judgement rather than strand by strand - retrain and remark.
Before release: (1) re-moderate all 40 strand by strand; (2) record the final agreed mark for each disputed script; (3) issue a one-page note showing A and C annotated, so the standard is visible to both markers and to the class. Do not adjust individual marks to make the distribution look normal; adjust the application of the scheme.