The 9-Minute Cloze
Six weeks before the C1 Advanced exam, Marta, 17, scores 4/8 on Part 1 multiple-choice cloze in three consecutive mocks. Her errors are not spread evenly: four are phrasal verbs or preposition collocations, one is a linker, one is a lexical distractor. She spends 9 of her 75 minutes on Part 1, reads the options before the text, and rarely re-reads the complete sentence. She can state the rules for linkers but cannot recall phrasal verbs unless she sees them. Her teacher has 12 contact hours left and must also cover Parts 5–7. Marta's target is a grade B.
1. What is the most defensible diagnosis of Marta's bottleneck, and what evidence in the setup supports it?
Read the worked answer
The stronger diagnosis is retrieval of lexical chunks, not a general timing problem or a grammar gap. Evidence:
- Four of six errors are phrasal verbs or preposition collocations. That is a cluster, not random noise.
- She can explain linker rules, so metalinguistic knowledge is not the main deficit.
- Reading options first encourages local matching: she looks for a word she recognises rather than constructing meaning from the whole sentence.
- Nine minutes is not excessive for eight items at C1; time is unlikely to be the primary cause.
The judgement call is to avoid treating every Part 1 error as 'vocabulary to memorise'. The pattern points to context-sensitive chunk retrieval under time pressure, so practice should require her to choose between near-synonymous chunks inside full sentences, not just review lists.
2. Propose a three-week Part 1 plan that fits the 12-hour constraint, and explain why it should not be 'do more full mocks'.
Read the worked answer
A defensible plan uses short, high-frequency cycles rather than more full papers:
- Weeks 1–2 (two 25-minute slots per contact hour): 8-item cloze sets built around one error family at a time - phrasal verb + particle, dependent preposition, linker. Marta must cover the options, predict meaning, then choose.
- After each set: she writes the whole sentence from memory with the correct chunk; this tests retrieval, not recognition.
- Week 3: mixed timed sets of 8 items in 8–9 minutes, followed by sentence-level error classification: 'particle', 'preposition', 'linker', 'lexical distractor'.
- Homework: 10 minutes, three times a week, recycling the same chunks in new sentences.
More full mocks would mainly re-measure the same weakness and consume the 12 hours needed for Parts 5–7. The plan targets the cluster while keeping Part 1 practice short enough to fit alongside reading practice.