External examination: interpreting a source text under time pressure
In the external examination, students receive a short Korean source text about 청년 실업 (youth unemployment) that cites two statistics: 2023 youth unemployment at 6.5% and a survey where 68% of respondents said they would accept a lower-paying job for security. The question asks students to 'analyse the attitudes presented and evaluate the extent to which the statistics support the writer's argument.' A student writes an answer that restates both statistics, says the writer 'seems worried,' and concludes 'therefore the government should create jobs.' The response is 180 words in Korean. The marker must place it against a descriptor band.
1. Diagnose the main weakness in the student's response relative to the command words.
Read the worked answer
The response restates rather than analyses. The command words are 'analyse' and 'evaluate the extent', which require:
- Identifying attitudes - not just 'the writer seems worried', but naming the stance (e.g. the writer frames youth unemployment as structural insecurity, not personal failure) with textual evidence.
- Evaluating the statistics' function - asking whether 6.5% and 68% actually support the argument. Does 68% wanting security prove the writer's claim, or merely describe preference? A strong answer notes the gap between a preference survey and a causal claim.
- Reaching a reasoned judgement - 'the government should create jobs' is the student's own opinion, not an evaluation of the writer's argument.
The response sits in a low band because it demonstrates comprehension but not analysis or evaluation.
2. What would a mid-band versus high-band answer look like on the 'evaluate the extent' element?
Read the worked answer
Mid-band: identifies the writer's stance, quotes one statistic, and offers a simple judgement such as 'the statistics mostly support the argument because many young people want secure jobs.' This shows evaluation but stays surface-level and does not interrogate the evidence.
High-band: distinguishes what each statistic can and cannot show. For example: the 6.5% figure establishes a problem's scale but says nothing about causes; the 68% figure shows preference for security but cannot demonstrate that insecurity causes unemployment. A high-band answer then judges the extent - 'the statistics support the description of the problem but not the writer's causal claim, so the argument is only partly supported' - and supports this with textual reference. The difference is interrogating evidence, not merely citing it.
3. The answer is only 180 words. How should length factor into the band placement?
Read the worked answer
Length is not a criterion in itself; it is evidence of development. A short answer can reach a high band if it is dense and analytical, and a long answer can stay low if repetitive. Here, however, the brevity coincides with underdevelopment: no textual evidence beyond restatement, no weighing of the two statistics, no counter-consideration.
The marker should therefore place the response on quality descriptors (analysis, evaluation, evidence use), then note that the 180 words left insufficient space for the required evaluation. Feedback should target what to add - evidence interrogation and a reasoned judgement - rather than 'write more'.