⚠️Work in progress! Help me!⚠️

Content analysis with inductive coding

6.2. Content analysis with inductive coding#

Warning

This page needs expanding on the background of the method. For now it jumps directly into an example

Using ChatGPT, we paste the data and we want to make sure that the context is clear. It is good to ask for some questions. For example

You You are a helpful research assistant expert in content analysis for qualitative research. I have collected survey data and I am pasting here as a table in text format. – Here goes the table above – Do you think these data in the table could be used for inductive coding?

ChatGPT Yes, the data in the tables can certainly be used for inductive coding. Inductive coding is a qualitative data analysis method where themes and categories are derived directly from the data itself, without any preconceived structure or framework. This method is particularly suited for exploring varied and nuanced perspectives, as provided in your synthetic interview data.

Here’s how the data you’ve gathered aligns with the process of inductive coding:

…omitted…

We are now sure that ChatGPT knows what we are talking about. We can now ask for the analysis to start.

You Now perform inductive coding on the table for the 10 subjects.

ChatGPT Inductive coding involves reviewing the qualitative data, identifying patterns, and developing categories based on these observations. Here, I’ll perform a basic inductive coding process using the responses from the 10 subjects provided in the tables regarding their experiences watching “Forrest Gump” with an eye-tracking device.

Step-by-Step Inductive Coding Process:

  1. Initial Reading: Review all data to gain a holistic understanding.

  2. Coding: Identify specific words, phrases, or sentiments that stand out and assign initial codes.

  3. Searching for Themes: Group similar codes to form potential themes.

  4. Reviewing Themes: Check if themes work in relation to the coded extracts and the entire data set.

  5. Defining and Naming Themes: Develop specific definitions for each theme.

Coding the Data: Based on the participants’ answers about their emotional experience with the movie and their experience with the eye-tracking device, I’ll generate initial codes and propose potential themes.

Emotional Responses to the Movie (Question 1):

  • Humor and Surprised Engagement: Emily, Alex (first-time watchers who expected boredom but found humor and depth).

  • Emotional Depth and Repeat Value: John, Lisa (repeat watchers who focus on emotional aspects and find new depth in repeat viewings).

  • Historical Interest: Michael (appreciates the historical context, finds the plot predictable).

  • Complex Narrative Engagement: Olivia (first-time watcher analyzing character development).

  • Tragicomic Perception: Ethan (finds the blend of tragedy and comedy appealing).

Experience Using the Eye-Tracking Device (Question 2):

  • Adaptability and Minor Distraction: John, Darren, Ethan (noticed the device initially but adapted quickly).

  • Initial Intrusion but Forgotten: Sarah (found it intrusive but got used to it).

  • Constant Awareness: Olivia (remained constantly aware, which affected viewing).

  • Comfort Issues but Interesting: Alex, Grace (uncomfortable at first but found the technological aspect intriguing).

  • Comfort and Distraction Issues: Emily (comfortable but distracted during humorous scenes).

Proposed Themes: From these initial codes, we can develop the following themes:

For Emotional Responses to “Forrest Gump”:

  • Varied Emotional Engagement: Different levels of engagement (humor, emotional depth, historical context) reflect personal and prior exposure influences.

  • Complexity in Simplicity: Viewers find layers of complexity within what appears to be a straightforward story, often colored by their own life experiences and expectations.

For Experience with the Eye-Tracking Device:

  • Technological Adaptation and Distraction: Users’ adaptation to the technology varies, with some finding it a minor distraction that they quickly overcome, while others remain aware of its presence throughout.

  • Technological Novelty and Comfort: The novelty of the technology provides an additional layer of interest, although comfort issues are common but typically transient.

Conclusion: These themes provide a structured way to understand the complex data set. They capture the main elements of the viewers’ experiences with the movie and the eye-tracking device, and can be used to further explore the impact of such technologies on cinematic experiences, as well as variations in individual viewer engagement with media. This exercise in inductive coding is fundamental for qualitative research, helping to structure unstructured data into coherent themes that can inform deeper analysis and understanding.

Did ChatGPT do a good job? ChatGPT is the junior assistant who did some task for you, but you are the expert to evaluate its output. If you think this is reasonable, if you think this is what you have also done, then you could ask to store the result in a table format for further analysis.

You Would it be appropriate to now create a table with one row for each of the 10 subjects and mark the results of the inductive coding process? Or is the goal of inductive coding to be more at the summary level of all subjects like you wrote here above?

ChatGPT …snip from the result… ** Example Table (Simplified for Illustration):**

Name Surname

Initial Codes

Themes

John Doe

Emotional depth, adaptability

Varied Emotional Engagement, Technological Adaptation and Distraction

Emily Stone

Humor, comfort issues

Varied Emotional Engagement, Technological Novelty and Comfort

Michael Rivera

Historical interest, adaptability

Varied Emotional Engagement, Technological Adaptation and Distraction

Lisa Chang

Emotional depth, adaptability

Varied Emotional Engagement, Technological Adaptation

Alex Johnson

Surprised engagement, novelty interest

Varied Emotional Engagement, Technological

Sarah Bennett

Emotional depth, initial intrusion

Varied Emotional Engagement, Technological Adaptation and Distraction

Darren Smith

Repeat value, adaptability

Varied Emotional Engagement, Technological Adaptation and Distraction

Olivia Martinez

Narrative engagement, constant awareness

Varied Emotional Engagement, Technological Adaptation and Distraction

Ethan Wright

Tragicomic perception, adaptability

Varied Emotional Engagement, Technological Adaptation and Distraction

Grace Kim Narrative engagement, novelty interest

Varied Emotional Engagement, Technological Novelty and Comfort

Now we could refine the themes by interacting more with ChatGPT, or continue offline for example by combining results without AI with the output generated by ChatGPT.