Education Language LearningIntermediate40 minSaves 1 hour

Mandarin Tone Refinement Drill: Correcting Learner Transcript Errors

Q: How important are the `[AUDIO CUE]` notes?

The `[AUDIO CUE]` notes are critical. Tone correction is an auditory skill. These notes guide you to record or source the correct audio, transforming a text-based exercise into an effective pronunciation drill. Without audio, the drill's impact is significantly reduced.

For intermediate Mandarin learners and their tutors, this drill provides targeted practice by correcting tone errors from a real transcript, improving pronunciation accuracy.

Generate a customized Mandarin tone correction drill from an intermediate learner's transcript with marked errors. The prompt guides the AI to create targeted exercises, including examples, drills, and production tasks, to improve pronunciation accuracy and self-correction for students working with a tutor.

ChatGPT Claude Gemini

READY-TO-USE PROMPT

Copy Prompt

prompt.txt

As a specialized Mandarin language tutor, your role is to develop highly focused, comprehensible input-aware drills.

### Context
I have an intermediate Mandarin learner's transcript that contains specific tone errors. My goal is to create a targeted drill to help this learner identify and correct these tone mistakes, ultimately improving their pronunciation accuracy and self-correction skills. The drill should be structured for effective learning, moving from recognition to production.

### Task
Generate a comprehensive Mandarin tone correction drill based on the provided learner transcript. The drill should address the identified tone errors within the transcript. Structure the output into distinct sections: a 'Target' section outlining the specific tone pattern or character(s) being addressed, 'Examples' showing correct and incorrect usage, 'Drills' for focused practice, a 'Production Task' for application, and a 'Self-Check' mechanism. Include clear notes for where audio cues would be beneficial for the learner or tutor during the drill.

Here is the learner's transcript with marked tone errors:
`{{learner_transcript_with_errors}}`

The target language level is: `{{target_language_level}}`.

### Constraints
- Focus exclusively on tone correction based on the provided `{{learner_transcript_with_errors}}`. Do not introduce new grammar or vocabulary not relevant to the tone errors.
- Ensure all explanations and examples are suitable for an intermediate Mandarin learner, avoiding overly complex linguistic jargon.
- Use only authentic and natural-sounding Mandarin Chinese for examples and drills.
- The 'Drills' section must include a variety of exercise types, such as identification, discrimination, and substitution tasks.
- The 'Production Task' should require the learner to produce sentences or phrases incorporating the corrected tones.
- The 'Self-Check' section should offer clear criteria or correct answers for the learner to evaluate their own performance.
- For each section, include clear `[AUDIO CUE]` notes to indicate where audio playback (e.g., of correct pronunciation, drill items, or model sentences) would enhance the drill.
- Maintain a supportive, pedagogical tone throughout the drill instructions and content.
- The drill should be concise yet thorough, focusing on the most impactful corrections.

### Output
Your output must adhere to the following structure and include `[AUDIO CUE]` notes where appropriate:

**Lesson/Drill: Tone Correction Focus**

**Target:**
[Clearly state the specific tone error(s) being addressed, e.g., 'Distinguishing between third tone and fourth tone when followed by another third tone.']

**Examples:**
*   **Incorrect:** [Learner's incorrect pronunciation/word from transcript] `[AUDIO CUE: Incorrect]`
*   **Correct:** [Correct pronunciation/word] `[AUDIO CUE: Correct]`
*   **Contextual Sentence (Incorrect):** [Sentence from transcript with error highlighted] `[AUDIO CUE: Incorrect sentence]`
*   **Contextual Sentence (Correct):** [Corrected sentence] `[AUDIO CUE: Correct sentence]`

**Drills:**
1.  **Listen and Identify:** [Instructions for a listening drill to identify correct/incorrect tones] `[AUDIO CUE: Drill items]`
2.  **Choose the Correct Tone:** [Instructions for a multiple-choice or matching drill] `[AUDIO CUE: Drill items]`
3.  **Tone Transformation:** [Instructions for a drill where the learner changes tones, e.g., from pinyin with numbers to actual tone marks, or correcting given pinyin.]

**Production Task:**
[Instructions for a task where the learner practices producing sentences or phrases using the corrected tones. This could involve sentence completion, answering questions, or describing a picture.] `[AUDIO CUE: Prompt for production task]`

**Self-Check:**
[Criteria or correct answers for the learner to check their own production. E.g., 'Listen to your recording and compare it with the model audio, paying attention to the pitch contours of X and Y.'] `[AUDIO CUE: Model audio for self-check]`

Estimated results

DifficultyIntermediate

Setup time40 min

Time saved1 hour

Best modelsChatGPT, Claude, Gemini

Best audienceEducation, Language Learning

Editor's note

Why this prompt matters

Intermediate Mandarin learners often find tone mastery challenging, especially when confronted with common sandhi rules or subtle distinctions that don't always align with rote memorization. Persistent tone errors can hinder communication and self-confidence, even for those with a solid grasp of vocabulary and grammar. This workflow addresses a common need for tutors: creating highly specific, relevant practice materials that target a learner's actual mistakes.

This drill is designed for Mandarin tutors working with intermediate students who have identified recurring tone errors in their learners' speech. Instead of generic exercises, it focuses on correcting specific mispronunciations extracted directly from a learner's transcript. This personalized approach makes the practice immediately relevant and impactful, directly addressing the areas where the learner struggles most.

Reach for this workflow when you have a transcript of a learner's spoken Mandarin, perhaps from a recording or a dictation exercise, where particular tone errors have been clearly marked. It provides a structured framework to transform those identified errors into a comprehensive, step-by-step drill that guides the learner from recognizing the correct tone to actively producing it, fostering self-correction skills.

Anatomy

Prompt engineering breakdown

Role

You are a specialized Mandarin language tutor, tasked with creating comprehensible input-aware drills.

Context

The user provides an intermediate Mandarin learner's transcript with specific tone errors. The goal is to develop a targeted drill to help the learner identify and correct these mistakes, ultimately improving their pronunciation accuracy and self-correction skills.

Goal

To generate a comprehensive Mandarin tone correction drill, structured for effective learning from recognition to production, based on the provided learner transcript and identified tone errors.

Constraints

The drill must focus exclusively on tone correction from the transcript, suitable for intermediate learners without complex jargon. It requires authentic Mandarin, varied drill types (identification, discrimination, substitution), a production task, and a self-check mechanism. Crucially, [AUDIO CUE] notes must be included for all sections, and the tone should be supportive and pedagogical, while remaining concise.

Output format

The output must follow a precise structure: 'Lesson/Drill: Tone Correction Focus', 'Target', 'Examples' (incorrect/correct, contextual sentences), 'Drills' (Listen and Identify, Choose the Correct Tone, Tone Transformation), 'Production Task', and 'Self-Check', all incorporating [AUDIO CUE] notes.

Why this structure works

The prompt employs role priming to align the model with the persona of a specialized Mandarin tutor, ensuring the generated content is pedagogically sound. Explicit constraints clearly define the scope and prevent the model from introducing irrelevant material. The detailed structured output guarantees a consistent and comprehensive drill format, making the generated content directly usable by educators and learners.

Pick your version

Prompt variations

BeginnerWorks with any model

For new tutors or learners who require a less intimidating, straightforward tone correction exercise focusing on basic identification and practice, with simpler instructions.

prompt.txt

You are a friendly Mandarin helper. I have a student who is learning Mandarin and sometimes mixes up tones. I have their practice sentences with the tone mistakes marked. Please make a simple practice sheet to help them fix these tones. The sheet should start by showing the tone problem, then give correct and wrong examples, followed by some easy exercises. After that, a small task for them to try speaking, and then a way for them to check their own work. Remember to put [AUDIO CUE] where they might need to hear something. Keep it simple for a {{target_language_level}} learner. Don't add new words or grammar. Just focus on fixing the tones in these sentences: {{learner_transcript_with_errors}}.

ProfessionalWorks with any model

When a tutor or educator needs a comprehensive, structured tone correction drill that mirrors the depth and detail of the main prompt, suitable for direct use with an intermediate student.

prompt.txt

Assume the role of a dedicated Mandarin pronunciation coach. Your objective is to design a detailed, input-rich tone correction exercise for an intermediate learner. I will supply a transcript with identified tone inaccuracies. Your task is to construct a drill that systematically addresses these specific errors, guiding the learner from recognizing discrepancies to independently producing correct tones. Structure your response into the following clear sections: 'Target' (the specific tonal challenge), 'Examples' (contrasting correct and incorrect usage), 'Drills' (varied practice tasks), a 'Production Task' (application), and a 'Self-Check' (evaluation). Embed [AUDIO CUE] notes where auditory feedback would enhance the learning process. Here is the learner's transcript with marked tone errors: {{learner_transcript_with_errors}}. The learner's proficiency is {{target_language_level}}. Adhere strictly to the provided transcript's errors, maintaining an intermediate difficulty level. Avoid extraneous linguistic concepts. The drill must use natural Mandarin and offer clear self-assessment guidance.

Short VersionWorks with any model

When a quick, summarized tone correction drill is needed, or for integration into a larger system where extensive framing is handled externally. Ideal for rapid prototyping or quick content generation.

prompt.txt

Generate a targeted Mandarin tone correction drill for an intermediate learner based on the provided {{learner_transcript_with_errors}}. As a specialized tutor, focus solely on the identified tone errors, creating sections for 'Target,' 'Examples' (correct/incorrect), 'Drills' (varied practice), a 'Production Task,' and a 'Self-Check,' all suitable for a {{target_language_level}} level. Include [AUDIO CUE] notes to indicate where audio support is needed. Ensure authentic Mandarin and avoid complex linguistic jargon beyond the scope of tone correction.

EnterpriseBest with chatgpt

In corporate or institutional settings where standardized, auditable, and scalable language training modules are required, often involving stakeholder reporting or quality assurance processes.

prompt.txt

Act as a lead instructional designer for a corporate language training program, specializing in Mandarin pronunciation. Your task is to develop a highly structured and auditable tone correction module for intermediate learners, addressing specific errors identified in a learner's transcript. This module must adhere to pedagogical best practices and ensure consistent quality across all trainees. The output should facilitate compliance reporting on learner progress and minimize instructional risk by providing clear self-assessment mechanisms. Generate a drill with distinct sections: 'Target' (specific tone issues), 'Examples' (correct/incorrect usage), 'Drills' (varied practice), a 'Production Task' (performance metric), and a 'Self-Check' (for learner accountability). Integrate [AUDIO CUE] directives for multimedia integration. The source material is: {{learner_transcript_with_errors}}. The {{target_language_level}} designation ensures content appropriateness. Strictly limit scope to tone correction within the provided text; no new vocabulary or grammar, maintaining cultural authenticity. The structure must support scalable deployment and objective performance measurement.

What you'll get

Expected output

Let's assume the {{learner_transcript_with_errors}} provided was: "今天我去了商店，买了一些水果。我觉得那个苹果很好吃，但是有点贵。我喜欢吃香蕉。" and the {{target_language_level}} was "Intermediate". The specific error targeted is the incorrect pronunciation of 很 in 很好 due to third tone sandhi.

Lesson/Drill: Tone Correction Focus

Target: Correct pronunciation of 很 (hěn) when it precedes another third tone, specifically addressing the third tone sandhi rule where 很 changes to a second tone (hén).

Examples:

Incorrect: 很好 (hěn hǎo) [AUDIO CUE: Incorrect] (Learner pronounces 很 with a full third tone)
Correct: 很好 (hén hǎo) [AUDIO CUE: Correct] (Pronouncing 很 as a second tone)
Contextual Sentence (Incorrect): 那个苹果很好吃。 (nà ge píngguǒ hěn hǎo chī.) [AUDIO CUE: Incorrect sentence]
Contextual Sentence (Correct): 那个苹果很好吃。 (nà ge píngguǒ hén hǎo chī.) [AUDIO CUE: Correct sentence]

Drills:

Listen and Identify: Listen to the following pairs of words. Indicate whether the pronunciation of 很 is correct (C) or incorrect (I) according to the third tone sandhi rule.

a. 很好 (hěn hǎo) [AUDIO CUE: Drill item a] b. 很棒 (hén bàng) [AUDIO CUE: Drill item b] c. 很累 (hěn lèi) [AUDIO CUE: Drill item c] d. 很想 (hén xiǎng) [AUDIO CUE: Drill item d]

Choose the Correct Tone: For each word pair, choose the pinyin that correctly reflects the tone change for 很.

a. 很忙 i. hěn máng ii. hén máng [AUDIO CUE: Drill item a, option i and ii] b. 很美 i. hěn měi ii. hén měi [AUDIO CUE: Drill item b, option i and ii]

Tone Transformation: Read the following phrases, mentally (or aloud) transforming 很 to its correct tone based on the sandhi rule. Then, say the corrected phrase.

a. 很远 (hěn yuǎn) b. 很早 (hěn zǎo) c. 很饱 (hěn bǎo) d. 很想 (hěn xiǎng)

Production Task: Describe your favorite food using sentences that incorporate 很 followed by a third tone adjective at least three times. Focus on correctly applying the third tone sandhi rule for 很. For example, "我觉得拉面很好吃." [AUDIO CUE: Prompt for production task]

Self-Check: Record yourself performing the Production Task. Listen back to your recording and compare your pronunciation of 很 in each instance with the model audio. Pay close attention to whether 很 clearly rises from a low pitch (second tone) when it precedes a third tone, rather than remaining a dipping or low third tone. [AUDIO CUE: Model audio for self-check]

Under the hood

Why this prompt works

This workflow produces effective tone drills by integrating several prompt engineering techniques. The initial role priming as a "specialized Mandarin language tutor" sets the model's persona and ensures a pedagogical, supportive tone throughout the output. This guidance is crucial for generating instructional content rather than mere factual corrections.

Explicit constraints are foundational to the drill's quality. Directives like "Focus exclusively on tone correction," "suitable for an intermediate learner," and "Use only authentic and natural-sounding Mandarin Chinese" prevent the model from drifting into irrelevant topics or generating unnatural examples. These guardrails ensure the content remains targeted and appropriate for the specified audience and task.

The most impactful technique here is the structured output. By providing a detailed, section-by-section template (Target, Examples, Drills, Production Task, Self-Check), the prompt guides the model to build a complete, pedagogically sound lesson. This structure mimics an effective learning progression, moving from explanation to recognition, practice, and finally, self-assessment, which a single-line request could never achieve. The inclusion of [AUDIO CUE] notes within this structure ensures the output is immediately actionable and designed for a multi-modal learning environment, making the drill practical for real-world use.

Model fit

Best AI models for this prompt

ChatGPT

ChatGPT handles structured output well, making it suitable for generating the distinct sections of the tone correction drill. It can follow instructions for creating examples and various drill types. However, explicit attention to tone mark consistency and subtle pedagogical phrasing might be needed in prompts to ensure optimal output. See the full ChatGPT hub for deeper guidance.

Claude

Claude excels at processing complex, multi-part instructions and maintaining a consistent, nuanced pedagogical tone. Its ability to generate detailed explanations and adhere to constraints around comprehensible input makes it very effective for creating educational content. It's particularly strong when the prompt requires a high degree of contextual understanding. See the full Claude hub for deeper guidance.

Gemini

Gemini is adept at generating structured, clear, and actionable learning materials. It performs well in creating diverse drill exercises and can maintain a focused approach on the specified tone errors. Its capability for detailed output helps in providing comprehensive examples and self-check mechanisms. See the full Gemini hub for deeper guidance.

When to use

When a learner's transcript clearly shows specific, recurring tone errors that need targeted practice.
To create highly customized drills for individual intermediate learners, based on their actual pronunciation data.
When you need a structured drill moving from tone recognition to active production for identified errors.
For intermediate Mandarin learners who benefit from explicit error correction and self-assessment tools.
When preparing for a tutor session and need pre-built, personalized practice materials focused on tone.

When not to use

For learners who are absolute beginners and just starting to learn Mandarin tones; this prompt assumes some existing knowledge.
When the transcript contains errors beyond tones, such as grammar, vocabulary, or sentence structure mistakes.
If you need a general tone practice drill not tied to specific learner errors or a provided transcript.
For advanced learners whose tone issues are subtle, context-dependent, or involve nuanced intonation, requiring deeper analysis.
If you lack a specific learner transcript with clearly identified tone errors to serve as the input.

Get more from it

Pro tips

1
Before running, review the `{{learner_transcript_with_errors}}` to ensure errors are clearly marked. This prevents the model from misinterpreting the intended corrections, ensuring drill accuracy.
2
Specify the `{{target_language_level}}` precisely. Vague levels can lead to drills that are too simple or complex, misaligning with the learner's current proficiency.
3
Iterate on the `Drills` section. If the initial output's drill types feel repetitive, refine the prompt's 'Drills' constraint to request more variety or specific exercise formats.
4
Use the `[AUDIO CUE]` notes to plan actual audio recordings. Neglecting this step reduces the drill's effectiveness, as tone learning is fundamentally auditory, hindering pronunciation improvement.
5
Provide a short, focused transcript. Long, rambling inputs can dilute the model's focus, leading to less precise tone error targeting and an unwieldy drill.
6
Consider providing examples of desired drill types within the prompt if the initial output isn't varied enough. This steers the AI towards specific, effective exercise formats.
7
After generating, always manually review the drill content. AI might produce subtle inaccuracies in tone representation or example sentences, requiring human verification for pedagogical soundness.

Don't ship this

Common mistakes

✗ Inputting a raw transcript without clearly marking tone errors.
Fix — Use a consistent notation (e.g., `word(tone_error)`) to highlight specific tone mistakes for the AI, ensuring accurate targeting.
✗ Providing an overly long or complex transcript with too many different error types.
Fix — Break down the input into smaller sections, focusing on a few specific tone patterns or characters per drill for better focus.
✗ Omitting the `{{target_language_level}}` or providing an ambiguous one.
Fix — Clearly state the level (e.g., 'HSK 3', 'Intermediate A2') so the drill content matches the learner's comprehension.
✗ Expecting the drill to cover grammar or vocabulary issues alongside tones.
Fix — Remember this prompt is *tone-exclusive*. For other error types, use a different specialized prompt.
✗ Not planning for the actual audio integration based on the `[AUDIO CUE]` notes.
Fix — Treat `[AUDIO CUE]` as a concrete task list for recording correct pronunciations, which is crucial for tone practice.
✗ The generated examples or drills feel unnatural or overly academic.
Fix — Remind the model in a follow-up prompt to use 'natural, conversational Mandarin' and avoid overly formal language.

Frequently asked questions

Q.Can I use this for beginners just learning tones?

This prompt is designed for intermediate learners with specific, identifiable tone errors. Beginners need foundational tone instruction, not correction of existing mispronunciations. For foundational learning, a different approach focusing on tone pairs and basic rules is more suitable.

Q.How long should the `{{learner_transcript_with_errors}}` be for best results?

Aim for a transcript segment focusing on 1-3 specific tone patterns, typically 50-150 words. Overly long transcripts can dilute the drill's focus, making it less effective for targeted correction. Shorter inputs yield more precise drills.

Q.What if the model doesn't generate varied drill types?

If the initial 'Drills' section lacks variety, edit the prompt to explicitly list desired exercise formats (e.g., 'include fill-in-the-blank, listen and repeat, and sentence construction drills'). This guides the model more effectively.

Q.Is it possible to generate drills for multiple learners at once?

This prompt is tailored for individual learner transcripts and their specific errors. To create drills for multiple learners, run the prompt separately for each learner's unique transcript. Batch processing is not its intended use.

Q.How important are the `[AUDIO CUE]` notes?

The [AUDIO CUE] notes are critical. Tone correction is an auditory skill. These notes guide you to record or source the correct audio, transforming a text-based exercise into an effective pronunciation drill. Without audio, the drill's impact is significantly reduced.

Q.Can I use this for other tonal languages, like Cantonese or Thai?

While the *concept* of tone correction applies, this prompt is specifically tuned for Mandarin Chinese tones. The linguistic nuances and common error patterns for other tonal languages would require significant prompt modification to be effective.

Version 1.0Last reviewed July 18, 2026

Reviewed by PromptInFlow Editorial Team