Understanding qualitative data

Qualitative data comes from open-ended questions or other methods where participants can respond freely, leading to varied and detailed answers. This type of data focuses on descriptions, opinions, and experiences rather than numbers. As a result, it often requires structured methods to identify patterns and make sense of the information.

Analysing qualitative data helps researchers uncover deeper insights into human behaviour and social phenomena. One common approach is thematic analysis, which involves spotting recurring ideas or patterns in the data. This method is flexible and widely used in psychology to organise complex information into meaningful categories.

What is thematic analysis?

Thematic analysis is a method for identifying, analysing, and reporting patterns within qualitative data. It helps researchers describe phenomena related to their research question by grouping similar ideas into themes.

Key terms in thematic analysis

• Themes - These are patterns or recurring ideas that appear across the data and relate directly to the research question. They act as categories that capture the essence of what participants are saying.
• Datasets - This refers to the collection of qualitative information being analysed. Datasets can vary in size, from a single short response to many hundreds of pages of text, such as interview transcripts or survey answers.

Thematic analysis allows researchers to handle different dataset sizes by adapting their strategies. For smaller datasets, the process might be straightforward, while larger ones require more systematic organisation.

The purpose and benefits of thematic analysis

Thematic analysis simplifies complex qualitative data by reducing it into key patterns while preserving the original meaning. This occurs because researchers become deeply familiar with the data through repeated reviews.

Advantages of thematic analysis

• It identifies important patterns that explain a phenomenon.
• Themes represent the data fairly, ensuring the analysis stays true to participants' views.
• The process is iterative, meaning researchers can revisit and refine their ideas multiple times.
• Analysed data can sometimes be converted into quantitative data (numbers) for easier comparison, such as counting how often a theme appears, without losing the qualitative depth.

This conversion helps manage large amounts of information and spot trends more clearly.

Approaches to thematic analysis

Thematic analysis can follow different paths depending on how researchers develop their themes. There are two main approaches: inductive and deductive.

Inductive approach

In an inductive approach, themes emerge directly from the data itself, without preconceived ideas. Researchers let patterns form naturally as they review the information. This is useful for exploratory research where the goal is to discover new insights.

Deductive approach

A deductive approach starts with existing theories or hypotheses, using them to guide theme development. Researchers look for data that fits these predefined categories. This is often applied when testing specific ideas from previous studies.

Both approaches involve going back and forth with the data, but the inductive method is more open-ended, while the deductive one is more structured.

Stages of thematic analysis

Thematic analysis follows a systematic process to ensure thorough and reliable results. It begins after data collection and builds understanding step by step.

The process of thematic analysis

1. Familiarisation - Researchers read and re-read the data to gain a deep understanding. This helps spot initial patterns.
2. Initial coding - Break the data into smaller parts by assigning codes (labels) to sections that relate to the research question.
3. Theme development - Group similar codes into broader themes. Refine these by checking if they accurately reflect the data.
4. Review and refinement - Compare themes against the original data, making adjustments as needed. This might involve recoding sections.
5. Finalisation - Define and name the themes clearly, ensuring they answer the research question.

The process is cyclical, not linear, so researchers may loop back to earlier stages until satisfied.

Coding in thematic analysis

Coding is a core part of thematic analysis, acting as a way to organise and simplify the data. It involves labelling sections of text to highlight meaningful patterns related to the research question.

Features of the coding process

• Cyclical nature - Coding evolves through repeated reviews, with researchers going back and forth between data and emerging ideas.
• Inductive evolution - Codes often develop inductively, emerging from the data rather than being set in advance.
• Data reduction - By using broad codes, researchers make large datasets more manageable without losing key details.
• Reflective journal - Researchers keep a record of the entire process, including raw data, coding decisions, thoughts during analysis, and how themes develop. This journal ensures transparency and helps track changes.

Coding can be done by the main researcher or an independent analyst who is unaware of the study's aims to reduce bias.

Ensuring reliability in coding

To make thematic analysis trustworthy, researchers often involve multiple people in coding and check for consistency.

Methods for improving reliability

• Multiple coders - Utilising several individuals to code the data allows comparison of results.
• Inter-rater reliability - This measures how much agreement there is between different coders. High agreement suggests the coding is consistent and objective.