Rater Judgement on Three Flagged Sentences
A rater is scoring a 250-word placement essay written in 30 minutes by a candidate applying to a foundation programme. Eleven sentences, three of them flagged. (1) 'Neither the residents nor the council have agreed on a plan.' (2) 'Walking through the market, the prices seemed high to me.' (3) 'The manager said that he will review the schedule the next day.' The rubric awards grammar out of 6: 6 = no distracting errors, 4 = occasional errors that do not impede meaning, 2 = frequent errors that obscure meaning. The candidate's argument is recoverable throughout. The rater must assign a band and must decide whether sentence 3 is an error at all, given that the reporting verb is past but the review may still lie in the future.
1. Correct sentences 1 and 2, name the rule each breaks, and say which error is more serious for a rater.
Read the worked answer
Sentence 1: 'Neither the residents nor the council has agreed on a plan.' With neither … nor, the verb agrees with the nearer subject - council, singular. Plural agreement is increasingly heard in speech, but a placement test scores the standard written rule.
Sentence 2: 'Walking through the market, I found the prices high.' The participle clause needs a subject that can walk; as written, the prices are walking. That is a dangling modifier.
Which is worse: the dangling modifier. The agreement slip is a local form error that a reader corrects automatically and reads past. The dangling modifier momentarily assigns the wrong agent and forces a backtrack. Applying 'errors that do not impede meaning', the rater should treat sentence 1 as minor and sentence 2 as the one that costs, because it distorts the relationship the writer intended.
2. Is sentence 3 an error? Decide, and state what evidence in the essay would settle it.
Read the worked answer
Probably not - and the essay itself can decide. The default is backshift: after a past reporting verb the reported clause moves back a tense (will → would). But backshift is suspended when the reported situation still holds at the time of reporting. If the review has not yet happened, will is defensible and would is also acceptable.
Two pieces of context settle it. First, time reference: if the essay elsewhere anchors the review to a date still in the future, will is correct; if the review has already occurred, would is required. Second, consistency with neighbouring sentences, which is what a rater actually scores.
A rater who marks sentence 3 wrong without checking context is applying a rule of thumb as if it were absolute. Flag it for verification rather than counting it automatically.
3. Assign a grammar band and defend it with numbers rather than impressions.
Read the worked answer
Band 4. The arithmetic: 3 flagged sentences out of 11 is about 27% - well under half, so not 'frequent'. Two of the three (agreement, backshift) are surface-level, and the third costs clarity for one clause, not the whole paragraph.
The rubric's operative contrast is distracting versus obscuring. 6 requires no distracting errors and this essay has some, so 6 is out. 2 requires errors that obscure meaning, and the argument is recoverable throughout, so 2 is out. 4 - occasional errors that do not impede meaning - matches.
Defend the band by citing sentence types, not counts alone: the same 27% would drop to 2 if both dangling modifiers landed in the thesis, because the damage would fall where the reader most needs clarity.