: This note is a Work In Progress (WIP); Beware of typos and other inconsistencies.
THEMATIC ANALYSIS
- Initial data collection: Deeply study collected data.
- Generate codes
- Organize codes by combining similar codes
- Search for themes
- Review themes
- Define and name emerged themes
Note
- Themes are usually developed by yourself rather than emerge.
- Hash focused codes against your research question to get a sense of what you know now about the data and how they (the codes you developed) have answered (part or in whole) your research question(s).
- You want to make sure the reader know what you know and knows all you want them to know about the data you analyzed.
CONTENT ANALYSIS
Used when considering word frequency, word patterns, or sequence of occurrence.
- Collect data
- Generate codes
- Observe and document frequency, etc.
- Generate themes from the observations.
GROUNDED ANALYSIS
Useful for building theories.
- Collect data from multiple sources
- Look for similarities and differences
- Use important peoples, events, or situations to then emerge themes.
Note
- Recommended read: Constructing grounded theory (Charmaz, 2014).
- You’re not looking to confirm or nullify a theory. Instead, you want to explore and understand what could be an explanation or possible theory behind your initial research question from your data.
- You want to deeply focus on the data to allow anything important from it emerge.
- You can adopt a line-by-line coding methodology with every code acting as a summary of some sort for each sentence.
HEURISTIC ANALYSIS
Removes researcher bias and focuses entirely on the participants
GENERAL NOTES
- In Qualitative Data Analysis, coding is a form of fine-grained tagging of text segments.
- Your codes should tell a story of your data and research in general.
- Coding is possible on a line-by-line, paragraph, or file basis.
- It’s helpful to have a “Table of Themes” in your results. The table contains the frequency of themes in your sources and communicates how strong a theme was due to its recurrence in the data.
- It’s best to leave out extracts in the table of themes.
- Create models, diagrams, etc.
- You can share your findings by main themes and subthemes, by groups involved in the study, or by data types; it depends.
- Not everything has to link or be linked to your research. They could just be context required to understand your overall research.
- You can be less deductive by using focused codes after a few coding sessions to analyze your remaining data
- Focused codes are codes reorganized into defined categories that answer your research question (or disprove it)
- Code preferably on a line-by-line basis because a certain code might pop up in multiple places where coding by paragraphs could easily lose that connection.
- “Codes are your tools.” They’re your little units for understanding the data you have collected—they could act as notes to yourself to remember nuances in the data.
- Themes on the other hand are for your audience—they’re to help THEM understand what you now understand.
- Subjectivity is part of the research process for QDA.
- It’s helpful to create a backup folder where initial codes are dumped and another where those codes are grouped. This could serve as an audit trail—a documentation of how qualitative analysis research was conducted—or as backup for when you delete a code/s.
- focused codes can be generated:
- after coding all data files for your study
- after coding a few files where you notice patterns starting to emerge that give insight into your research question.
- Focused codes are any common sense grouping. They help make sense of the hundred of codes by providing some structure. They could be simple categories such as “Good Things 👍”, “Bad Things 👎”, and “Background Information 💭”