Worked example 6: English story (60 s)

Recording AA_5d2c0d0d-… · grade 5 · age 11 · enumerator #02 · window 840–852 s of the session, 60-s clock from the child's first word at 844.4 s

How to read this page. Section 1 is the printed text the child had in front of them, numbered in reading order. Section 2 is the 60-second clip, timed from the child's first word. Section 3 is what the speech-to-text service typed from that clip; the microphone hears the enumerator too, so prompts like "shabash" can appear between the child's words. Section 4 has one row per scorer with a colour chip for every printed word: the enumerator row is the live tablet marking, the official score; the baseline row is the first automatic method, which matches transcript words to printed words by spelling; each model row is that model's own judgement from the audio or from the transcript. Red where the enumerator has green means the scorer flagged too much; green where the enumerator has red means it waved a mistake through; grey means skipped or never reached; faded means no verdict. "Attempted" is the furthest word reached. Section 5's "difference from enumerator" is each scorer's correct count minus the enumerator's: near zero is good, a large positive number is too lenient, a large negative number too harsh. Section 6 is the second human's re-listen.

1. What the child was supposed to read

1. imran 2. woke 3. up 4. early 5. for 6. school 7. today 8. his 9. class 10. was 11. planting 12. trees 13. he 14. took 15. a 16. small 17. plant 18. in 19. a 20. bag 21. the 22. teacher 23. asked 24. imran 25. to 26. show 27. the 28. class 29. how 30. to 31. plant 32. it 33. imran 34. dug 35. a 36. hole 37. put 38. the 39. plant 40. in 41. the 42. soil 43. and 44. put 45. water 46. in 47. it 48. the 49. teacher 50. was 51. happy 52. with 53. his 54. work 55. his 56. friends 57. clapped 58. for 59. him

The study export holds only right/wrong flags, not the text itself, so this text was reconstructed from 30 children's reads with an LLM; a word or two of it may itself be wrong, which would show up as the same "mistake" in every scorer's row.

2. The child's audio

Play the 60-second clip (Drive, Taleemabad accounts)
8 kHz classroom recording from the May 2026 study, upsampled. Folder: worked-examples clips

3. What the speech-to-text service heard

imran was asked to get ready for school

whisper-1 forced to English · 8 words inside the 60-s window. The enumerator's own words (prompts, "shabash") are included when they fall inside the window.

4. Every scorer's verdict on every word

read correctly wrong skipped or not reached no verdict

Enumerator (tablet, live)

imranwokeupearlyforschooltodayhisclasswasplantingtreeshetookasmallplantinabagtheteacheraskedimrantoshowtheclasshowtoplantitimrandugaholeputtheplantinthesoilandputwaterinittheteacherwashappywithhisworkhisfriendsclappedforhim

1 correct of 6 attempted in 60 s (words after the 60-second point are grey: the export records 0 for both "correct" and "never reached", so they are only known up to the attempted count)

Baseline: speech-to-text + spelling alignment

5 correct of 59 attempted; 0 extra words the child said are ignored

Gemini 3.1 Pro · audio

5 correct of 6 attempted · model note: The child reads the first 6 words, mispronouncing 'early' as 'really', and then the teacher stops the reading to switch to a math assessment…

Gemini 3.8 Flash · audio

6 correct of 6 attempted · model note: The student read the first six words correctly before the teacher stopped the reading task to switch to a math activity.

Gemini 3.5 Flash-Lite · audio

4 correct of 6 attempted · model note: The child read only the first few words and then stopped/switched to math task.

Gemini 2.5 Flash · audio

6 correct of 6 attempted · model note: The child read the first 6 words correctly and then stopped reading. The remaining words were skipped.

GPT audio mini · audio

0 correct of 0 attempted · model note: Child did not read the printed text; the child engaged in a different activity.

MiMo 2.5 · audio

8 correct of 8 attempted · model note: Child read the first 8 words ("Imran woke up early for school today his") with accented but recognizable pronunciations, then stopped; teach…

Claude Sonnet 5 · transcript

1 correct of 8 attempted · model note: Only the first word matched the printed text exactly; subsequent spoken words diverged from the expected sequence and the rest of the passag…

GPT-5 mini · transcript

1 correct of 8 attempted · model note: Child spoke 8 words; only the first printed word ('imran') matched; printed words 2–8 were read as different words and words 9–59 were skipp…

Gemini 3 Flash · transcript

1 correct of 6 attempted · model note: The child substituted most of the first sentence with different words and stopped reading after word 6.

DeepSeek 3.2 · transcript

7 correct of 26 attempted · model note: The child read only the first part of the text, stopping after 'to'. The transcript 'imran was asked to get ready for school' does not match…

5. The counts

Scorer	Words correct	Words attempted (furthest word reached)	Difference from enumerator (correct)	Difference (attempted)
Enumerator	1	6	+0	+0
Baseline: speech-to-text + alignment	5	59	+4	+53
Gemini 3.1 Pro · audio	5	6	+4	+0
Gemini 3.8 Flash · audio	6	6	+5	+0
Gemini 3.5 Flash-Lite · audio	4	6	+3	+0
Gemini 2.5 Flash · audio	6	6	+5	+0
GPT audio mini · audio	0	0	-1	-6
MiMo 2.5 · audio	8	8	+7	+2
Claude Sonnet 5 · transcript	1	8	+0	+2
GPT-5 mini · transcript	1	8	+0	+2
Gemini 3 Flash · transcript	1	6	+0	+0
DeepSeek 3.2 · transcript	7	26	+6	+20

6. The second human

The second human (fareeda) re-listened: on the correct count the reviewer agreed with the enumerator; on the attempted count the reviewer agreed with the enumerator; on orf_reading_sentences_eng the reviewer agreed with the enumerator; the enumerator followed the protocol.