The 240-Word Essay That Argues Well and Reads Badly
You teach a 14-week MET preparation course. A student, Ana, writes a 240-word opinion essay in 30 minutes on whether her city should close its center to cars. Her argument is genuinely good: a clear thesis, two reasons, a concrete example about the bike lanes on her street. The problems sit at sentence level - seven missing or wrong articles, five subject-verb agreement errors, three run-on sentences - plus a conclusion she never wrote because time ran out. Her vocabulary is accurate but plain, and her only linking words are 'first', 'second' and 'finally'. She needs a B2-level writing score in six weeks and has two hours of homework per week. She is frustrated because her teacher keeps writing 'grammar!' in the margin.
1. You can only target two features in the next two weeks. Which two, and what makes them the right choice?
Read the worked answer
Pick by frequency × fixability × rubric weight, not by what is most annoying to read.
1. The missing conclusion (and paragraph completeness). It costs points twice - under task fulfilment and under organization - and it is close to free to fix. A conclusion is a three-sentence move she can learn in ten minutes; the payoff per hour is the highest of anything on the list.
2. Articles and subject-verb agreement, taught as a self-editing routine, not as grammar theory. These are her two most frequent error types, both are rule-based, and both are catchable in a three-minute final read if she has a checklist. Seven article errors in 240 words is roughly one every 34 words - enough to pull down the language score noticeably.
What I would not target yet: plain vocabulary and linking words. 'First / second / finally' is adequate cohesion for a B2 essay; examiners tolerate simple lexis if it is accurate. Spending her two hours a week on synonym lists buys almost nothing. And I would replace the marginal 'grammar!' with an error tally, so she can see the count fall.
2. Is the missing conclusion a scoring problem or a time-management problem? What evidence would settle it?
Read the worked answer
These need different remedies, so diagnose before teaching.
Evidence to collect: compare one timed essay with one untimed essay on a similar prompt. Also ask her to log her 30 minutes (planning / drafting / reviewing).
- If the untimed essay has a conclusion, she knows the move and the problem is pacing. Remedy: a fixed clock - 3 minutes planning, 22 drafting, 5 reviewing, with a hard stop at 25 no matter where she is, then write the conclusion first in the review window.
- If neither essay has one, she does not hold the genre schema. Remedy: teach the move explicitly - restate the thesis in new words, summarize the two reasons, add one implication. Drill it from six supplied thesis statements.
- If her log shows she planned for 12 minutes, the pacing problem is upstream of writing and the clock fix is still the answer.
Note the trap: under exam conditions an absent conclusion costs organization points regardless of cause. So even if it is purely pacing, she must be trained out of it - you are treating the cause and the symptom at once.
3. Design a 4-hour, two-week homework plan and say what you would cut.
Read the worked answer
Four 60-minute sessions, each with a measurable output.
Session 1 (60 min): Rewrite the same essay from her notes, with a 3-minute final read against a two-item checklist (articles, agreement). Then count errors per 100 words, before and after. Target: seven article errors down to three. Session 2 (60 min): Three conclusion drills, ten minutes each, from supplied thesis statements; then write one conclusion for her own essay from Session 1. Session 3 (60 min): Full timed essay, 30 minutes, with the planning template and a written timing log. Score only organization and language. Session 4 (60 min): Edit a peer essay with the same checklist, then re-edit her Session 3 essay. Finding errors in someone else's text is what makes the checklist automatic.
Cut: the vocabulary notebook and any linking-word worksheet. They cost time and do not move the two targets. Measure by error count per 100 words and by whether a conclusion exists, not by a holistic grade - she needs to see the number move before the score does.