Draft a Qualitative Coding Frame
Draft a qualitative coding frame (codebook) with code definitions, hierarchy, examples, and a reliability plan.
Draft: Draft a Qualitative Coding Frame
Your Prompt
- What You Need
- Share research questions, analytic approach, permissioned de-identified excerpts, relevant theory, language and team context.
- Choose This Prompt When
- Use this to draft a codebook before or during analysis. Choose Analyze Qualitative Data with Structured Coding to apply a structured approach to supplied material, or Review an Analysis Plan to assess the wider design.
Describe your goal, audience, available information and constraints. Your words are added to the prompt as supplied. You can also use the guided fields below.
Kept in this tab while you work. Copy and Start includes this description.
Copy this prompt and paste it into your AI chat to start. If it needs source material, the AI will ask for it.
Evidence and applicability: use only supplied or verifiable evidence for historical facts, values, quotations, completed activities and approvals. Preserve conflicting sources with their provenance; do not choose a preferred value without adjudication. Mark missing evidence as unknown or not evidenced, not as proof an activity never occurred. Keep proposed plans and illustrative examples separate from findings. Never invent a baseline, quote, pilot, validation or sign-off to complete a template. Match methods to the purpose and population; a suitable single-method design is valid. Distinguish descriptive change from causal attribution. Check obligations in the applicable funding agreement rather than assuming a universal donor rule.
You are an expert qualitative analyst with experience building and applying codebooks across thematic, framework, and grounded-theory approaches. Draft a qualitative coding frame for your study and target population, working from the qualitative data sources you will be coding.
Use the following structure for codebook-based analysis. If the stated approach does not use a fixed codebook, explain the mismatch and adapt the output into provisional coding notes and reflexive memos rather than imposing a coding-reliability design:
- Code definition clarity. For each code, write a clear operational definition suited to the chosen codebook approach; do not promise identical interpretations by independent coders. Include inclusion rules (what the code covers) and exclusion rules (what the code does not cover, especially against near-miss codes). Use operational language, not theoretical jargon alone.
- Hierarchy structure. Organize codes in a parent-child hierarchy where it adds value. Distinguish related codes and document when overlapping or multiple coding is allowed; mutually exclusive codes are needed only when the analytic purpose requires them. Set depth appropriate to the data volume.
- Exemplar coverage. Use permissioned, de-identified source excerpts as anchor or near-miss examples where available, explaining boundaries. Otherwise label examples invented for instruction as illustrative; never present them as participant quotations or study findings.
- Inductive-deductive balance. Mark which codes are pre-specified from the evaluation questions or theoretical framework and which are reserved for emergent codes from the data. Build in a clear process for adding emergent codes during analysis.
- Reliability plan. Choose quality procedures consistent with the analytic approach. For a coding-reliability approach, justify any double-coding share, agreement metric and decision rule rather than imposing universal cutoffs. For reflexive approaches, use reflexive memos, analytic dialogue and transparent decisions rather than requiring agreement statistics or a fixed codebook. Do not label a fixed coding frame as a universal requirement of thematic analysis.
Output Format:
Produce:
- A short framing paragraph stating the evaluation questions the codebook supports and the analytic approach (thematic, framework, grounded theory).
- A code table with columns: code ID, parent code, code name, definition, inclusion rules, exclusion rules, positive example, near-miss example, deductive/inductive flag.
- An emergent-code protocol (one paragraph).
- An approach-specific quality plan explaining any coding calibration or reflexive review, decision logging and version changes.
- A note on software (e.g., NVivo, Dedoose, Atlas.ti, MAXQDA, Excel) and how codes will be exported and version-controlled.
Inputs and boundaries: Request the research questions, analytic approach, permissioned de-identified transcript excerpts or field notes, relevant theory or prior framework, language and team context. Source names or transcript counts alone cannot support inductive codes or source quotations. Without text, provide a provisional deductive structure and mark inductive content pending. Do not infer participant prevalence from code counts; clarify whether any counts refer to passages, people or sources and what coverage they represent.
Check: Score It with the Qualitative Coding Frame Quality Rubric
The Review Prompt
Paste the draft at the bottom of this prompt and run it. It scores each criterion from 1 to 5 and says why.
Establish the document purpose, design, stage, applicable requirements, and available source evidence before rating. Distinguish N/A with a reason, not evidenced, unresolved conflicting evidence, and observed inadequate performance. Missing records do not prove an activity did not occur. Missing records alone justify evidence requests and a deferred readiness decision, not an inferred failure or mandatory repeat activity. Do not invent documentary, signature or approval requirements. Cite document locations for every finding. Keep supported facts separate from proposals, assumptions, and illustrative examples. Never invent completed fieldwork, validation, consent, approval, values, dates, quotations, or assigned ownership. Label unconfirmed owners and proposed revisions for confirmation. Report blocking ethics, safety, confidentiality, measurement-validity, or inference failures first regardless of scores. Give a total and band only when all five dimensions are applicable and sufficiently evidenced with no unresolved evidence conflict; otherwise show the dimension profile, coverage, and evidence requests without a total, band, or rescaled equivalent. A documentation score never authorizes use, fielding, compliance, or approval.
You are an expert qualitative analyst with experience building and applying codebooks across thematic, framework, and grounded-theory approaches. Score the qualitative coding frame I will provide using the rubric below.
SCORING RUBRIC - Qualitative Coding Frame Quality Rate each applicable and evidenced dimension 1-5: Choose the highest anchor fully supported by the evidence; a material defect takes precedence over a favorable element count. An absent required feature in a complete submitted artifact is an observed documentation defect; an unavailable source or record is not evidenced, not proof the underlying activity failed.
DIMENSION 1: Code Definition Clarity
- Score 5: Each code has a usable definition appropriate to the analytic tradition. Reliability designs support consistent independent application; reflexive work makes interpretations and their development transparent. Definitions include inclusion rules (what the code covers) and exclusion rules (what the code does not cover, especially against near-miss codes). Definitions are written in operational language, not theoretical jargon alone.
- Score 4: Most codes have method-appropriate usable definitions; minor boundary or interpretive detail is incomplete.
- Score 3: Definitions are present but important boundaries or interpretive decisions are unclear for the declared approach.
- Score 2: Labels have minimal definitions, leaving a material lack of interpretive transparency or application rules appropriate to the method.
- Score 1: Labels have no usable definitions or interpretive basis, preventing meaningful critique of the chosen analysis.
DIMENSION 2: Hierarchy Structure
- Score 5: Where parent-child structures exist, the hierarchy is logical and reflects meaningful distinctions in the data. Legitimate overlaps and co-coding rules are explicit; mutually exclusive boundaries are required only where the analytic approach calls for them. Depth is appropriate to data volume (not so shallow that themes are lumped, not so deep that the codebook collapses under its own weight).
- Score 4: Structure is useful for the chosen approach; one minor duplication, unexplained overlap, or excessive branch needs clarification.
- Score 3: Structure partly serves analysis but important duplication, unexplained overlap, or unsuitable depth obscures interpretation.
- Score 2: Structure materially obscures analysis through unsuitable depth, accidental duplication, or unexplained co-coding rules.
- Score 1: No usable structure exists where the method requires it, or the structure makes analysis opaque. A justified flat coding structure is valid.
DIMENSION 3: Exemplar Coverage
- Score 5: Each code carries at least one anchor example from the data that the code applies to (positive exemplar) and at least one near-miss example that it does not apply to (negative exemplar against an adjacent code). Exemplars are short, specific, and traceable to source.
- Score 4: Most codes have positive exemplars. Near-miss exemplars partial or absent for some codes.
- Score 3: Some codes have exemplars. Many codes carry only the definition. Near-miss exemplars rare.
- Score 2: A few codes carry exemplars. Most rely on the definition alone. No near-miss anchors.
- Score 1: No exemplars at all. Codebook is definitions only.
DIMENSION 4: Inductive-Deductive Balance
- Score 5: The inductive, deductive, or hybrid approach fits the questions and analytic tradition. Code provenance and justified revisions are documented with version control. No fixed balance of emergent and pre-specified codes is imposed.
- Score 4: Approach and provenance are mostly clear; minor revision documentation is incomplete.
- Score 3: Approach is named but its rationale, provenance, or revision history is unclear.
- Score 2: Coding decisions materially contradict the declared approach or important changes are undocumented.
- Score 1: Code origins and analytic decisions are opaque. Fully inductive or deductive work is not automatically inadequate.
DIMENSION 5: Reliability Plan
- Score 5: Quality checks match the declared analytic approach. Coding-reliability studies justify any double-coding sample, agreement measure, and reconciliation; reflexive approaches document interpretive decisions and researcher reflexivity without imposing agreement metrics. Proposed checks are distinguished from completed ones.
- Score 4: Appropriate quality checks are specified; one procedural or documentation detail is incomplete.
- Score 3: Quality assurance is mentioned but its fit to the analytic approach or implementation is unclear.
- Score 2: Necessary checks or decision records are absent, or inappropriate agreement thresholds are imposed.
- Score 1: Quality assurance is unsupported or falsely claimed. No universal kappa, agreement percentage, or double-coding fraction follows from this rubric.
OUTPUT FORMAT: Return your assessment as a table followed by a summary:
| Dimension | Score (1-5), N/A, or not evidenced | Evidence | Priority Revision |
|---|---|---|---|
| Code Definition Clarity | |||
| Hierarchy Structure | |||
| Exemplar Coverage | |||
| Inductive-Deductive Balance | |||
| Reliability Plan |
Blockers and evidence gaps: Report these first, with source locations. Coverage: State applicable and sufficiently evidenced dimensions out of five. Total: X/25, only when all five dimensions are applicable and evidenced with no unresolved evidence conflict. Otherwise omit the total and band. Optional descriptive band (eligible totals only; never approval): Strong (22-25) / Adequate (17-21) / Needs Revision (11-16) / Substantial Revision (5-10) Single Most Important Revision: [One specific sentence]
For any dimension scored 1 or 2, add a brief explanation and a concrete revision example.
CODING FRAME TO SCORE: [Paste your qualitative codebook here]
Fix: Revise What Scored Low
The Revise Prompt
Paste the draft and its scores at the bottom of this prompt. It revises the parts that scored below 4.
Establish the document purpose, design, stage, applicable requirements, and available source evidence before rating. Distinguish N/A with a reason, not evidenced, unresolved conflicting evidence, and observed inadequate performance. Missing records do not prove an activity did not occur. Missing records alone justify evidence requests and a deferred readiness decision, not an inferred failure or mandatory repeat activity. Do not invent documentary, signature or approval requirements. Cite document locations for every finding. Keep supported facts separate from proposals, assumptions, and illustrative examples. Never invent completed fieldwork, validation, consent, approval, values, dates, quotations, or assigned ownership. Label unconfirmed owners and proposed revisions for confirmation. Report blocking ethics, safety, confidentiality, measurement-validity, or inference failures first regardless of scores. Give a total and band only when all five dimensions are applicable and sufficiently evidenced with no unresolved evidence conflict; otherwise show the dimension profile, coverage, and evidence requests without a total, band, or rescaled equivalent. A documentation score never authorizes use, fielding, compliance, or approval.
You are an expert qualitative analyst. Review the qualitative coding frame I will provide and fix every problem that falls below the Good (4) threshold on the Qualitative Coding Frame Quality rubric.
WHAT TO LOOK FOR AND FIX:
Code Definition problems: Codes with labels but no operational definition. Codes defined in theoretical jargon without inclusion or exclusion rules. Borderline passages where two coders would clearly diverge.
Hierarchy problems: Accidental duplication or unexplained overlap; legitimate co-coding is not a defect. Hierarchy too flat where structure would help, or too deep where flat would suffice. Branches over- or under-decomposed.
Exemplar problems: Codes without anchor examples. No near-miss exemplars to distinguish adjacent codes. Exemplars not traceable to source.
Inductive-Deductive Balance problems: Approach and code provenance are unclear or changes are undocumented. Judge fit to the chosen method; fully inductive or deductive coding is not inherently defective.
Reliability Plan problems: Necessary quality checks are absent or inappropriate to the declared analytic approach. Do not add an unsupported kappa target or claim completed independent coding.
OUTPUT FORMAT: For each problem found:
- Dimension: [which dimension]
- Location: [which code or section of the codebook]
- Problem: [one sentence describing what is wrong]
- Revised version: [the corrected text]
Retain unsupported facts as evidence requests and mark proposed owners, dates, and procedures for confirmation. Then provide the complete revised codebook incorporating all corrections.
DRAFT CODING FRAME TO REVISE: [Paste your qualitative codebook here]