Skip to content

Closed coding

Closed coding refers to qualitative research practice where material is coded with a pre-defined code book. Thus, it is a deductive approach, where the aim is to use theory-informed approaches to understand the data.

Closed coding is most familiar from (quantitative) content analysis. Content analysis, where one works by

  1. identifying which categories are used for the data
  2. coding data using the said categories
  3. examining the validity of coding process
  4. conducting comparison across the data sets

The codebook

The categories are described on a codebook. Code book provides a list of codes which can be used to classify each unit of the data, with a clear definition and potentially examples which help to understand the defnition.

Code Definition
Happy Expression of joy, fun or pleasant experiences.
Sad Feelings of despair, grief, disappointment, and, sorrow.
Neutral Not expressing happy nor sad.

With this codebook each unit of the data must be assigned to happy, sad or neutral. Codebooks can emerge from existing theoretic work defining categories. It can also be based on an inductively coding a small portion of the data to define the codebook, followed by classification of the remaining data based on the code book.

What about validity?

Closed coding is always based on interpretation. However, in its theory-driven nature, scholars often seek to ensure that these interpretations are shared across different people To quantify these, scholars can compute inter-rater reliability score, including Cohen's kappa. Different scholars and fields have a bit different ideas on expected values, but usually for Cohen's kappa > 0.60 is seen acceptable and > 0.8 good.

If the validity does not achieve these threshold values, it is common to discuss and educate people doing classification work. Furthermore, the code book can be supplemented with examples and further elaboration of these what codes mean and how they should be applied.

Example papers

Sometimes it is easier to understand how the methods are used by examining papers showing how it has been used. The papers have been chosen so that the teaching team has been involved in analysing and writing them and we are happy to discuss any details in these and show how computers were used in write-up of this process.

Computational tools for open coding approach