Eight out of ten in dictation, silence in pair work
Class V has 38 children and 40 minutes of English daily. Dictation of ten words gives a class average of 8/10, and rhyme recitation is accurate and loud. Last week the teacher set a paired task: ask your partner three questions about their family. Of 38 children, 31 produced only single words ('mother', 'three', 'brother') or silence, four produced full sentences, three were absent. When one girl said 'I have two sister', the teacher immediately repeated 'two sisters', made the class chorus it three times, and moved on; two children stopped volunteering for the rest of the week. The term test in twelve days is entirely written - fill in the blanks, matching, one paragraph. Unit 4's role-play is marked 'optional' in the teacher's handbook and has been skipped. There are no audio recordings; the teacher's spoken English is fluent and she is the only model the children hear.
1. Explain why children who score 8/10 in dictation cannot hold a two-line conversation.
Read the worked answer
Dictation tests sound-to-spelling under controlled conditions: the child listens to a word she already knows, in isolation, and writes it. Speaking requires retrieving words, assembling syntax, pronouncing, monitoring the listener and coping with a reply - all in real time, with no preparation. These are different competences, and the class has been practising only the first.
Three further factors are visible in the setup. First, there is almost no pushed output: the children have never been required to produce a message that someone needs to understand. Second, the correction episode shows what happens when they try. The girl produced a comprehensible message - 'two sister' - and received a public form correction plus three choruses. The message was ignored and the error was made social. Two children stopped volunteering, which is a rational response to that. Third, input is one voice only, so their ear is tuned to a single accent and a single speed.
Notice also that 31 of 38 produced single words, not wrong sentences. They have vocabulary and no frame. The missing piece is formulaic chunks and a reason to speak, not more grammar rules.
2. Design a seven-minute daily routine for the next twelve days that fits alongside test preparation.
Read the worked answer
Minute 1 - Question of the day on the board with a sentence frame: 'What do you eat for breakfast? I eat ______.' Teacher models once, then the whole class says it once.
Minutes 2–5 - Pair talk. Each pair asks and answers two questions from the frame, then swaps partners once. Teacher circulates with a notebook, says nothing about errors, and writes down three things she hears.
Minutes 6–7 - Two or three pairs report to the class; the teacher responds to the content ('You eat poha every day - interesting') and then recasts one form naturally: 'Ah, two sisters.' No chorus, no demand to repeat.
Why this shape: the frame removes the assembly burden so the child can attend to meaning and fluency; the fixed routine lowers anxiety; the teacher's notebook replaces public correction; and content-first responses keep the message central. Twelve days gives twelve short repetitions, which is enough to make a few chunks automatic.
Because the written test is only twelve days away, do not abandon it: use the same frame sentences as fill-in-the-blank items, so speaking practice and test preparation feed each other instead of competing.
3. How should errors be handled, and how can speaking be assessed inside a written-test regime?
Read the worked answer
Handle errors by type and by stage. During fluency work, do not interrupt: note errors and respond to meaning. Afterwards, choose one frequent error - not the rarest - and give a two-minute focus the following day with a clear contrast, for example 'one sister / two sisters' with three examples, then move on. Recast rather than demand repetition; a child who hears 'two sisters' in a natural reply has received the correction without losing face. Save explicit correction for accuracy-focused writing tasks, where the child has time to think.
Assess speaking with a simple observation checklist, not a test. Take five children a day, three criteria - did the child answer the question, was it comprehensible, did the child attempt a full sentence - and mark ✓ or ✗. Thirty-eight children are covered in eight days. Keep the record formative and report it separately from the written paper, so no child is failed on an oral test the syllabus never prepared them for.
Finally, reinstate the Unit 4 role-play after the test. Twelve days of frames will not build fluency; a term of short, daily, low-anxiety pair talk will.