How to Create a Codebook for Qualitative Research (with Template)

What is a codebook in qualitative research?

A codebook (also coding manual, coding frame or category system) is a structured document that lists all the codes you use to analyse your qualitative data and explains exactly how each code should be applied. It turns your interpretive decisions into explicit rules that you, a second coder or an examiner can follow.

Codebooks are central to qualitative content analysis, framework analysis, template analysis and coding reliability approaches to thematic analysis. In reflexive thematic analysis, Braun and Clarke (2022) do not require a formal codebook, because codes are expected to evolve organically; even there, a simple list of codes with short descriptions helps you keep track and document your process.

Running example: you interviewed 12 first-generation students about belonging at university. After coding four interviews you notice that you used the codes “peer support” and “study group” inconsistently. A codebook with clear criteria solves exactly this problem.

What goes into a codebook entry?

Most templates for team-based qualitative research go back to a format proposed by MacQueen and colleagues in the late 1990s, which has become a de facto standard. A complete entry contains the elements in the table below. For a dissertation, the first five are essential; the others are helpful for difficult codes.

Step by step: building your codebook

Building a codebook is an iterative process. Do not expect to get it right on the first attempt; examiners actually want to see that you revised it based on the data.

• 1 Start with your research question and interview guide. Derive a few broad parent codes (deductive starting points), for example “Sources of support”, “Barriers to belonging”, “Turning points”.
• 2 Code two or three transcripts openly and add inductive codes where the data show something new.
• 3 Write a draft entry for each code: label, definition, inclusion and exclusion criteria, one example.
• 4 Apply the draft to two further transcripts. Note every segment where you hesitated between codes.
• 5 Revise: sharpen definitions, merge overlapping codes, split codes that are too broad, add decision rules.
• 6 Optional but valuable: ask a fellow student to code one transcript with your codebook and discuss disagreements.
• 7 Freeze the codebook for the main coding and recode the early transcripts with the final version.
• 8 Document versions and the reasons for changes in your memos.

Deductive, inductive and hybrid codebooks

A deductive codebook is developed before coding from theory, previous studies or a framework. It is typical of directed content analysis and framework analysis and makes your study comparable with existing research. An inductive codebook is developed from the data and documented afterwards. A hybrid codebook, the most common in dissertations, starts with deductive parent codes and adds inductive child codes.

Whatever you choose, label the origin of each code in the codebook (for example “D” for deductive and “I” for inductive). This lets examiners see at a glance which parts of your findings extend existing theory.

Rules that make a codebook work

Good codebooks share a few properties. Clarity: definitions use plain language and avoid circular wording (“peer support is support from peers”). Distinctiveness: codes on the same level do not overlap, or overlaps are resolved by explicit rules. Coverage: every relevant segment can be coded, if necessary using an “other” code that you review regularly. Manageable size: for a master’s project, 20 to 50 codes in a two-level hierarchy are typical; more than 100 codes in the final version usually means you have not yet moved to second cycle coding.

Also decide on coding rules that apply to all codes: can a segment receive several codes? How much context do you code? What happens with the interviewer’s questions?

• Code the participant’s answer, not the interviewer’s question
• Code enough context that the segment is understandable on its own
• Multiple coding is allowed across parent codes but not within mutually exclusive subcodes
• If two codes fit equally well, apply the more specific child code

Codebook format: table, spreadsheet or tool

For a dissertation, a simple table in your thesis document or a spreadsheet with one row per code is enough. QDA software usually stores code descriptions directly with the codes. Our free coding tool at /codieren lets you enter a definition, an anchor example and a coding rule for every category, organise parent and child codes and export the complete codebook together with your coded segments and a code matrix. This export can go straight into your appendix.

Your department or supervisor may expect a particular format, for example a codebook following Schreier or a specific template; follow your university’s guidelines (you can upload them to Professor Digital so the feedback uses them).

Presenting the codebook in your dissertation

In the methods chapter, describe how you developed the codebook (deductive sources, inductive additions, number of revisions, pilot coding) and summarise its structure, for example: “The final codebook comprised 6 parent codes and 31 child codes.” Show one complete entry as an example.

In the appendix, provide the full codebook. In the findings chapter, use the codebook structure only if your method is descriptive (as in qualitative content analysis). In thematic analysis, organise findings by themes, not by codes.